Prehistory of Te Hua Can

•
.• '
•
Subsistence
v\\\V\\\V\\\
m
M
wWWWV^^
D
B3K
R. S. MacNeish, Douglas S. Byers, and Others

Insatiable curiosity about the origins of New
World civilization and agriculture in the —
New World this means Indian coni led —
archaeologist R. S. MacNeish to undertake a
hunt that eventually led him to a cave in
the arid Tehuacan Valley of east central
Mexico. There, in 1960, he found what he
believed to be his goal. The Tehuacan
Archaeological-Botanical Project, sponsored
by the Peabody Foundation for Archae-
R. S.
ology, Phillips Academy, Andover, Massa-
chusetts, was organized by MacNeish in
order to take full advantage of his initial
discovery.
Under his leadership, field parties spent
four seasons searching through rubbish of
long-dead towns and millennia-old litter of
hunting parties for traces of early com and

its wild ancestors, and for first evidence of
civilization. They succeeded in establishing
an unbroken record spanning 9,000 years
and covering most aspects of man’s life. This
is the longest continuous record of man’s
economic activities yet found anywhere.
A team of scholars, many trained in fields
other than archaeology, have interpreted for
us the significance of this record. Midway in
its course plant breeders set about improv-

ing wild food plants, with the result that
corn, avocados, and other plants of impor-
tance today eventually took on their modem
form.
The greatly increased yield of the im-
proved plants furnished reserves of food to
support artists, astronomers, priests, engi-
neers, city planners, and rulers. Although
com was known to be the foundation
long
of New World civilizations, its origin re-
mained a mystery until the Tehuacan Ar-
chaeological-Botanical Project uncovered
tiny cobs that almost duplicate the hypo-
thetical wild ancestor of com bred by Paul
C. Mangelsdorf.
Volume I, Environment and Subsistence,
is concerned principally with geography,

climate, and the natural resources exploited
(Continued on back flap)

Digitized by the Internet Archive
in 2015
https://archive.org/details/prehistoryoftehu01tehu
THE PREHISTORY OF THE TEHUACAN VALLEY
THE PREHISTORY OF THE TEHUACAN VALLEY
VOLUME ONE
Environment and Subsistence
Edited by Douglas S. Byers
Published for the
ROBERT S. PEABODY FOUNDATION

Phillips Academy Andover
,
UNIVERSITY OF TEXAS PRESS • AUSTIN & LONDON

Published in Great Britain by the
University of Texas Press, Ltd., London
Library of Congress Catalog Card Number 67-17873

Copyright © 1967 by the University of Texas Press
All rights reserved
The preparation and publication of

The Prehistory of the Tehuacan Valley
has been assisted by grants from
the National Science Foundation.
Typesetting by Service Typographers, Indianapolis, Indiana

Printingby The Meriden Gravure Company, Meriden, Connecticut
Binding by Universal Bookbindery, Inc., San Antonio, Texas
EDITOR’S PREFACE
The Prehistory of the Tehuacan Valley will make formation about the development of irrigation that
known the findings of The Tehuacan Archaeological- has been gathered by R. B. Woodbury and his col-
Botanical Project, which, during four field seasons, leagues. In spite of its obvious connection with agri-
1961-1964, sought to discover the beginnings of agricultural activities, this material had to be relegated
culture in the New World and the concomitant rise to another volume.
of autocthonous civilization. This project, supported Volume II is concerned with the nonceramic arti-
hv the National Science Foundation and the Rocke- facts of the ancient inhabitants of the Tehuacan Val-
feller Foundation, was the brain child of R. S. Mac- ley, and with the identification and seriation of types
Neish, Research Associate, R. S. Peabody Foundation, of artifacts to derive a chronology for the valley. It
who envisaged it as an interdisciplinary project. As is concerned with the distribution of such types
also
such, it was the first of its kind to be undertaken in and with the possibilities of such distribution in time
Middle America. and space as a means of reconstructing cultural units
Assembled in a team to collect data or analyze ma- and their relationships, whether locally or on a wider
terial gathered in the course of archaeological excava- scale.
tions were people trained in botany, zoology, geology, Volume III will be concerned with ceramic artifacts.
and anthropology. Any lack of either the “inter-” or the The sherds, vessels, and figurines unearthed in the
“discipline” in the resulting papers may well stem from excavations will be described, discussed, and used in
the pioneering nature of this study. At this stage, it is the reconstruction of the cultural and mercantile his-
easy to see how the project might have been carried out tory of the valley and in the placement of the bur-
to better advantage for all contributing sciences. Al- geoning and principalities with respect to
cities
though preliminary reports of some of the findings contemporary developments in Mexico beyond the rim
have appeared elsewhere, a number of “first” papers of the Tehuacan Valley.
for the region will appear in this series. Volume IV will be concerned with chronology, both
first of the six volumes that present the
This, the relative and absolute. Stratigraphic relationships and
findings, concerned primarily with fields other than
is succeeding types of artifacts will provide the key to
archaeology, with the Tehuacan Valley and its immedi- relative chronology, but radiocarbon dating holds the
ate surroundings, with its natural history and natural key to absolute dating. By combining the two — absolute
resources, with its people, and with the exploitive chronology and relative chronology —an estimate of the
activities of a people faced with the necessity of wring- rate of growth of the culture, progressively increasing
ing the greatest possible advantage from what appear in richness and complexity, can be derived.
to have been the progressively deteriorating resources Volume V will comprise detailed reports of the
of an increasingly more hostile and discouraging en- archaeological survey and of the several excavations,
vironment. In this volume are included data bearing with sections to show stratigraphic relationships in the
on the domestication of native American plants and ground. It will also attempt to reconstruct the way
their subsequent improvement. Findings regarding the of life represented by each occupational unit.
domestication of maize should answer many questions Volume VI, the concluding volume, is intended to
concerning the origin of this important crop plant, but contain a summation of what has been presented in
leave many more be answered. That other
still to the other volumes and to draw conclusions regarding
economic plants were exploited at an even earlier date the processes which transform a simple, gathering so-
w ill come as a surprise to many. ciety into a complex, sophisticated, and urban civiliza-
Lack of space and problems of time have made it tion based on differences of class and caste.
impossible to include in this volume extensive in- We have attempted to instill some unity into these
v
EDITORS PREFACE
volumes. That these attempts have not been com- in large part reshaped, manuscript for several chapters.
pletely successful will be apparent to the casual The success of illustrations for other chapters is entirely
reader. It is difficult to process so many manuscripts and painstaking care.
attributable to his interest
from so many authors in such a way as to eliminate To Nancy H. Flannery, of the Tehuacan Archae-
all differences in style. It is obviously impossible to ological-Botanical Project, we are indebted for various
force agreement on even though some individual
all, pieces of art work in more than one volume. The skill-
contributions depart radically from findings of other ful and capable hands of Ashley Baker, of the Robert
contributors. To attempt to do so would be to smother S. Peabody Foundation for Archaeology, have turned
with a protective editorial blue-gray what may even- to the illustration of the chapter on the Borgia Codex
tually prove to be strokes of genius and insight. as well as to other art work and the making of photo-
The National Science Foundation has supported graphic prints. She has been of inestimable help in
the preparation of these manuscripts, even though it preparing the many complicated tables in Volume I.
has seemed a never-ending job. To the successive It would have been impossible to carry the work to
program directors for Anthropology, Allan H. Smith completion without the discerning and sympathetic
and Richard W. Lieban, and through them, to the efforts of Chase J.
Duffy, who has carried the major
National Science Foundation, we express our grati- editorial load.
tude for their patience and encouragement.
We gratefully acknowledge the very considerable Douglas S. Byers
contributions to this volume made by Frederick John- Director, Robert S. Peabody Foundation
son. He organized a conference among authors out of Andover, Massachusetts
which this volume took shape. He hammered out, and August 1966
CONTENTS
Editor’s Preface v
Introduction 3
Richard S. MacNeish
1. An Interdisciplinary Approach to an Archaeological Problem ... 14
Richard S. MacNeish
2. Field Laboratory and Techniques 25
Richard S. MacNeish
3. The Region and Its People 34
Douglas S. Byers
4. Climate and Hydrology 48
Douglas S. Byers
5. Geologic Studies 66
Jean Brunet
6. The Human Skeletons 91
James E. Anderson
7. Codex Borgia and the Venta Salada Phase 114
Robert Chadwick and Richard S. MacNeish
8. The Vertebrate Fauna and Hunting Patterns . 132
Kent V. Flannery
9. Wild and Cultivated Maize
Prehistoric 178
Paul C. Mangelsdorf, Richard S. MacNeish, and Walton C. Galinat
10. Archaeological Phaseolus from Tehuacan 201
Lawrence Kaplan
11. Curcubits from the Tehuacan Caves 212
Hugh C. Cutler and Thomas W. Whitaker
12. Plant Remains 220
C. Earle Smith, Jr.
13. A Cotton Boll Segment from Coxcatlan Cave 256

Stanley G. Stevens
14. Analysis of the Tehuacan Coprolites 261
Eric O. C allen
15. A Summary of the Subsistence 290
Richard S. MacNeish
Bibliography 311
Index 321
vn
TABLES
1. Maximum, Minimum, and Mean Temperatures 22. Maize Cobs and Kernels by Race .... 196
at Valley Stations, 1952-62 71 23. Bean Remains from All Sites 204
2. Burials from El Riego Cave, West Niche 92 24. Characteristics of Remains of Fhaseolus 209
....
. . .
3. Burials from El Riego Cave, East Niche . 92 25. Cucurbit Remains from All Sites 218
4. Burials from Coxcatlan Cave 92 26. Remains of Food and Fiber Plants from All
5. Burials from the Ajalpan Site 92 Sites following 232
6. Burials from Quachilco 92 27. Maximum Width of Septum and Segment and
7. Burials from Purron Cave 93 s/r Ratio of Cotton 257
8. Burials from Coatepec 93 28. Comparative Average Measurements of 10
9. Measurements of Adult Skulls .... 103 Dried Cotton Boll Segments .... 257
10. Measurements of Immature Skulls 104 29. Number of Coprolites per Phase and Cave . 265
11. Measurements of Male and Female 30. Contents of El Riego Phase Coprolites . . 267
Mandibles 105 31. Contents of Coxcatlan Phase Coprolites . . 269
12. Estimated Statures 108 32. Contents of Abejas Phase Coprolites 270
....
. . .
13. Length of Immature Diaphyses 108 33. Contents of Selected Santa Maria Phase
14. Comparative Data on New World Skulls . 111 Coprolites 273
15. Small Rodent Remains from Tehuacan Sites . 142 34. Contents of Selected Palo Blanco Phase
16. Food Animals from Coxcatlan Cave and Coprolites 275
Terrace 160 35. Contents of Venta Salada Phase Coprolites . 279
17. Food Animals from the Arroyo Lencho Diego . 164 36. Volume of Meat in Liters Estimated from
18. Food Animals from Teccoral Canyon . 166 Faunal Remains 298
19. Food Animals from between Subareas 37. Volume of Food in Liters Estimated from
1 and 2 following 168 Wild Plants following 304
20. Food Animals from Subarea 1 following 168 38. Volume of Food in Liters Estimated from
21. Maize Remains from All Sites ....
.
182 Cultivated Plants .... following 304
viii
The Prehistory of the Tehuacan Valley
ENVIRONMENT AND SUBSISTENCE

Introduction
by Richard S. MacNeish
find it difficult to begin the Tehuaean reports ex- the Christian era in the southwestern United States.
cept on a personal note, because the seasons spent Contemporary botanical studies concerned with the
I in the field at Tehuaean have been among the most origin of corn advanced three theories: (1) that corn
exciting and satisfying of my life. Since childhood I had evolved from the wild grass, teosinte, in Mesomer-
have been interested in Mayan and Mexican archae- ica; (2) that com had evolved from a grass in South
ology and in the processes by which these civilizations America; and (3) even that com had been first domesti-
came into being. Later, of course, I learned that the cated from a local grass in southeast Asia. We were a
origin of civilization is closely connected with the long way from having any answers to the corn prob-
development of agriculture, and that in the New lem, and our guesses were not even very close.
World this meant corn. I spent many years before I In 1948 and 1949 the situation changed for the
undertook this study of the rise of civilization in better. Tiny primitive corn cobs were uncovered in
Tehuaean in tracking down leads on the origins of the summer of 1948 at Bat Cave, New Mexico. Later,
corn agriculture. charcoal found with them placed their age between
For me the Tehuaean began in 1945—
project really 3000 and 5000 years. Almost as important as their
46, when urging of Dr. Fay Cooper Cole and
at the antiquity was the submission of the cobs by their finder,
under the auspices of the Department of Anthro- Herbert Dick, to Paul C. Mangelsdorf, Fisher Pro-
pology of the University of Chicago, I undertook an fessor of Natural History at Harvard University, for
archaeological reconnaissance of coastal Tamaulipas study and analysis. At almost the same time, digging
in northeastern Mexico to seek prehistoric cultural in La Perra Cave duringthe second Tamaulipas expedi-
connections between Mesoamerica and the southeast- tion in early 1949, uncovered other remains of primi-
I
ern United States. In the course of this project we tive corn. Some of this, dated by radiocarbon labora-
uncovered evidence of early man and two ancient cul- tories to about 2500 b.c., also found its way to
tural complexes characterized by food-gathering and Mangelsdorf’s laboratory. He discovered that the early
the first steps toward agriculture. The season was also corn from La Perra Cave exhibited certain similarities
important to me because I gained valued experience in to Nal-tel, one of the primitive modern races of maize
surveying and finding preceramic sites in Mesoamer- previously defined by Wellhausen, Roberts, Hernandez,
ica. I had learned at first-hand that caves in the drier and Mangelsdorf (1952). Mangelsdorf identified the
regions of Mexico did indeed contain stratified pre- earliest La Perra corn as pre-Nal-tel (Mangelsdorf,
ceramic remains, and that among these were remains of Galinat, and MacNeish 1956).
plants. The sum of this research indicated that corn was
This first endeavor added little to the search for early probablv first domesticated in Middle America rather
corn and the origin of its domestication —but neither than in South America or Asia, and quite possibly be-
had previous archaeological research in Mesoamerica. fore 3000 b.c. Furthermore, these investigations indi-
Archaeological work before 1945 in South America cated clearly that the origin of domesticated maize
and Mexico had led to a belief that corn agriculture could best be discovered by cooperative research be-
had not been practiced there much before 1000 b.c., nor tween botanists and archaeologists.
was the cultivation of corn thought to be older than In the next stage of our investigations —
from 1951
3
to 1955 — Mangelsdorf and I made a concerted effort I looked for dry caves
in Honduras and Guatemala.
to seek the elusive first domesticated corn. As part of Although found remains that were possibly pre-
I
this effort I went to southwestern Tamaulipas. My ceramic, they included no corn. The following year,
findings there proved definitely that corn was domesti- Frederick A. Peterson, who had been with the New
cated, not in, but to the south of that region. Other World Archaeological Foundation, and I started exca-
information began to come in from several sources. vations in Santa Marta Cave in southern Chiapas.
Studies of pollen secured there indicated that the com
pollen was not sufficiently old to justify consideration
of that region as corn’s homeland. It seemed that
Chiapas was too far south, just as Mexico City, Tamauli-
pas, and northwest Mexico were too far north.
As a result of our archaeological efforts up to 1959,
we believed that corn probably was domesticated be-
fore 3000 b.c., very likely in an area south of the Valley
of Mexico but north of Chiapas. Mangelsdorf’s studies
pointed to the probability that the ancestor of corn
was a highland grass not unlike the pod popcorn
found in Bat Cave and southwest Tamaulipas. Pre-
vious finds had revealed that plant remains were to be
found in dry caves in dry regions. We also knew that
caves or rock shelters usually are to be found in deeply
dissected country. A study of rainfall, climate, topog-
raphy, and geological data revealed only three likely
regions that could meet these requirements. One was
in southern Oaxaca, another was the Rio Balsas or
Mescala region of Guerrero, and the third was the
Tehuacan Valley of southeastern Puebla and north-
eastern Oaxaca.
Although I briefly visited Oaxaca and the Tehuacan
Valley at the end of the 1959 season, I did little cave
hunting. Perhaps the most profitable part of this initial
visit was my meeting Sr. Ricardo Gutierrez, at that
time assistant manager of the Hotel Penafiel in Te-

huacan, who enthusiastically helped me pursue my
“com hunt’’ in many ways. A brief search around
Fig. 1. The search for early maize in the southwestern Mitla, Oaxaca, after Christmas in 1959 revealed nu-
United States and Mesoamerica. merous caves, probably with preceramic remains, but
none seemed dry enough nor the refuse deposits deep
enough to contain really old corn, so I returned to
Robert H. Lister, working in Chihuahua and Sonora, Tehuacan. There, in January 1960, through Sr. Gu-
uncovered remains of corn which demonstrated that tierrez, I met Luis Vasquez, Director of the Museo de
corn was originally domesticated to the south of those la who generously gave his help
Revolucion in Puebla,
states. Meanwhile research by P. B. Sears and K. H. in furthering my project. Armed with an official fed-
Clisby (1955) in the Valley of Mexico led us to think eral permit from Arturo Romano of the Instituto
that corn was domesticated still farther south, even Nacional de Antropologia e Historia, a series of letters
though pollen identified as that of maize, and probably and êquipped with borrowed gear,
to local authorities,
wild maize, was found in a drill core in levels assigned I began explorations in the Tehuacan area.
to a much more ancient era than the pollen judged to In the company of a bored army lieutenant from the
be from domesticated corn and found in more recent local base who was under orders to guard me, I made
levels (Barghoorn et aJ. 1954). my first foray along the highway between Chazumba
Because of the findings in northern Mexico we and Huajaupan de Leon, Oaxaca, visiting sixteen caves.
moved operations considerably farther south. In 1958 All were in volcanic deposits too porous for the preser-
4
INTRODUCTION
vation of vegetal materials. Four other caves south away dirt with a paint brush, I uncovered two more
of Ajalpan were equally unproductive. A second trip, tiny cobs. We held in our hands possible ancestors to
this time under the protection of a very perplexed modern domesticated com.
sergeant, netted four caves near Zapotitlan, Puebla, This impression was confirmed a month or two
and one cave just past Tequixtepec, Oaxaca. The latter laterby Mangelsdorf when he examined the cobs at
and one of the Zapotitlan caves gave indication of Harvard University. Still later the cobs were dated to
containing stratified fill which might conceal preserved 3610 b.c. ± 250 years (M-1089). This radiocarbon de-
organic remains, but both shelters were very small. termination on associated charcoal was made at the
Other caves, visited in the company of a sleepy army University of Michigan Laboratory through the kind-
private, included five near the town of El Riego, just ness of James B. Griffin, Director of the university’s
north of Tehuacan, three northeast of a town along Museum of Anthropology. These were the oldest com
the highway to Veracruz, and three only one kilometer cobs that had ever been found!
south of the town of Tehuacan. All had possibilities, After January 27 we did little further digging in
but the prospects of finding remains of early corn in what came to be known as Coxcatlan Cave. I made a
any of them did not appear be great.
to preliminary reconnaissance of the valley and also be-
After looking at three dozen sites without much gan to analyze the material we had uncovered. The
success, I turned to a series of forms that local school last event of this first season at Tehuacan took place
teachers had filled out for Luis Vasquez. Three of these on February 8 when Jose Luis Lorenzo, then Director
forms mentioned caves in the municipio of Tehuacan; of the Escuela Nacional de Antropologia, paid us a
one respondent spoke of a cave near Altepeji, and an- visit. Lorenzo realized the importance of our finds and
other of a cave east of Ajalpan. I visited these, at last came to Tehuacan to offer his assistance, an assistance
by myself, but they revealed little of interest. The third he continued generously for the next four years.
to give
cave, said to be located in the hill called Agujereado By the spring of 1960, we thought we knew approxi-
south of the town of Coxcatlan, was described by mately where corn had been first domesticated, and
Senorita Berta Martinez of that town. Eventually, after we had a few examples of this earliest com as proof
a number found Senorita Martinez. She
of inquiries, I (see map). Furthermore, we had found at least one
kindly arranged for her brother Hector and a guide, stratified site. Better yet, Tehuacan occupied a stra-
Pablo Bolanos, to take me to this cave, some ten tegic position in almost the exact center of Meso-
kilometers south of Coxcatlan. After a long, hot walk america. The we might
possibility that in this region
along the edge of the mountains, through thick stands find evidence of the development of agriculture and
of cactusand mesquite, we arrived at the rock shelter. the concomitant rise of civilization was fairly great.
Even from a distance it looked promising. The artifacts Therefore, with considerable help and advice from
and refuse on the surface, the size of the shelter, and Paul Mangelsdorf, 1 made an initial plan for the Te-
the quantity of vegetal material that lay beneath the huacan project. Feeling fairly confident that we would
goat dung covering the floor showed that this was a eventually be able to carry out the plan, I also wrote
site tobe tested. Fred Peterson, who had been of great help at Santa
From January 2l to January 27 the three of us, Pablo, Marta Cave, C. Earle Smith, and others to see whether
Hector, and I, tested this cave. Behind a large rock they would be available for the proposed research.
roughly in the center of the shelter we dug a two-meter Because of its demonstrated interest in interdis-
square to a depth of about two meters, using trowels. ciplinary projects and the flexibility of its organization,
We took out everything, including the loose dirt, by we felt that the Robert S. Peabody Foundation for
bucket loads and put through a mesh screen to be
it Archaeology was well qualified to administer a col-
sure that we missed nothing. Slowly we peeled off the laborative project such as we envisaged. Accordingly
successive strata. The uppermost layer yielded Post- I discussed our plan with Douglas S. Byers and Fred-
classic remains; the stratum underlying it contained erick Johnson, Director and Curator, respectively, of
Classic and afew Formative sherds; then there was a that institution. Expressing interest, they helped work
sterile layer. Underneath that was a thick layer which out a program of research to be carried on under the
was obviously preceramic. On January 27 after lunch, auspices of the Peabody Foundation. This program
Pablo, working well down in the preceramic stratum, was embodied in a proposal submitted to the Na-
recovered a tiny corn cob no more than an inch long. tional Science Foundation in September 1960. Briefly,
Only half believing, I took his place in the bottom of our main purpose was to investigate the development
the pit. After a short period of troweling and cleaning of agriculture and the concomitant rise of civilization
5
in Mesoamerica. The area of research was to be the data would not only give us a fuller history of the
Tehuacan Valley in the Mexican states of Puebla and region, but that the co-ordination of information from
Oaxaca. Our approach was to be interdisciplinary, a number of fields might be mutually stimulating.
utilizing the skills of all appropriate scientific fields. Ideally, our specialized studies were to have begun
The immediate objectives were as follows: (1) To find after all the relevant data had been excavated (and
and excavate a series of stratified sites with preserved dated at least relatively) and the particular problems
remains of plants and animals used as food by the of each discipline had been defined. The scientist con-
inhabitants; (2) to establish a relative and an absolute cerned with a special study was then to come to Te-
chronology for the Tehuacan Valley; (3) to reconstruct, huacan not only to work with his own particular set
as completely as possible, changes in aboriginal pat- of data, but to see the region and the excavations as
terns of subsistence and to bring this information to background for interpreting his data. Finally, as the
bear on the problem of the origin and spread of agri- various reports and analyses were completed, we
culture in the New World; (4) to reconstruct the cul- hoped there would be an exchange and a co-ordination
tural pattern of each stage of the complete Tehuacan of each set of information which would solve a wide
sequence; and (5) to compare the Tehuacan sequence variety of problems.
with similar sequences found in other parts of the world Chronologically, our first specialist was Miss Monika
in the hope of learning how civilization develops and Bopp, who came to Tehuacan in the spring of 1961
why it develops. to collect pollen from the profiles of our excavations.
This program was to be carried out by a relatively However, before she could analyze her samples she
small field crew of archaeologists who would be work- married and left Mexico. In the spring of 1963, James
ing in close co-operation with botanists, paleontologists, Sehoenwetter came to Tehuacan also to collect pollen.
zoologists, ethnographers, malacologists, and others Eventually the pollen samples were sent to Elso S.
whom we planned to bring to Tehuacan. Among the Barghoorn and Henry Irwin of Harvard University.
botanists were to be persons whose special skills and The purposes of the pollen studies were to determine
knowledge centered on studies of corn, squashes, beans, the agricultural changes, give us clues as to chron-
and the analysis of coprolites. ology of the strata of the various sites, and to deter-
Our proposed schedule allowed three seasons to mine the changes of vegetation through time. Because
complete the task. During the first season, 1960-61, of poor pollen preservation in our earlier levels, these
we intended to make an archaeological survey and worthy objectives were not attained.
a botanical reconnaissance and to carry out a few C. Earle Smith, now with the United States Depart-
limited excavations; during the second season, 1961- ment of Agriculture, Crops Research Division, came
62, we proposed complete the survey and under-
to to Tehuacan in the summer of 1961 to make a botanical
take our main excavations, while the first group of survey of the valley and into the surrounding moun-
specialists would begin their studies in Tehuacan; tains. He continued the survey in August of 1962,
during the final season, 1962-63, the specialists were and in 1963 he identified the vegetal remains from the
to complete their studies, the archaeological analysis archaeological sites. His subsequent analysis of these
was to be completed, and the final reports written. materials gave us pertinent information on the flora
Now that I look back on this original schedule I am of the valley as well as much information about the
surprised to find how closely we followed its spirit, ancient subsistence patterns.
even though we did not carry through each detail Closely allied to the study of the ancient subsistence
exactly. Obviously, field conditionschanged, unfore- were the analysis of the archaeological cucurbits, un-
seen circumstances arose, personnel came and went, dertaken in the spring of 1962 bv Hugh C. Cutler,
but perhaps the factor that most changed our specific Executive Director of the Missouri Botanical Gardens,
plans was that we just simply did better than we had and Thomas W. Whitaker, Senior Research Geneticist
expected to in the beginning. We found more sites, of the United States Department of Agriculture; the
more stratigraphy, more artifacts,more preserved or- study of beans, made in the fall of 1962 by Lawrence
ganic material than any of us (Bvers, Johnson, and I) Kaplan, now with the University of Massachusetts in
ever guessed would be possible. Oh that I can always Boston; and the studies of corn undertaken by Paul
be wrong in my estimates for reasons such as these! C. Mangelsdorf and Walton O. Galinat of Harvard
Inherent in the original Tehuacan project was the University. In order to obtain more precise informa-
intention of incorporating data from as many different tion about diet Eric O. Callen of Macdonald College
sources as possible. We felt that the inclusion of such of McGill University undertook a study of the pre-
6
INTRODUCTION
historic feces preserved in the dry rock shelters. His the project was a co-ordinated set of research and
study, initiated in 1962 but lasting until 1964, was functioned as a unit with the purpose of understand-
carried out in painstaking and minute detail. ing the Tehuacan region’s ancient history. Part of the
Supplementing these studies on subsistence and unifying process came from the participants’ reading
environment were the zoological researches. Kent V. one anothers’ reports and articles, some of it came from
Flannerv, now with the Smithsonian Office of An- correspondence, much of it came from endless argu-
thropology, arrived in the summer of 1962 to make ments and discussions, both formal and informal, by the
a faunal survey of the Tehuacan Valley. In 1963 he various expedition members. Our researches in Tehua-
analyzed the numerous animal remains that had been can taught me how artificial are the barriers between
uncovered in excavation. This information added not the different disciplines and how stimulating it can be
onlv to our knowledge of the ancient subsistence but to break through them.
also gave us hints of changing environments. The cor- The first season of the Tehuacan Archaeological-
relation of all this information, both zoological and Botanical Project began in December 1960 when
botanical, gave us a rather full picture of the diet and Frederick A. Peterson, C. Earle Smith, Melvin L. Fow-
also indications of the actual seasons of the cave oc- ler, and I met in Chicago to plan our initial program.
cupations. We arranged for the purchase of a jeep, which Peter-
Furthermore, data derived from the botanical and son was to drive to Mexico, where I would meet him
zoological studies furnished us with hints about chron- early in January. Smith found that his schedule would
ology in the Tehuacan region. However, the most sig- permit him to come to Tehuacan the following summer
nificant endeavors in this line were in the capable to initiate his botanical survey. Fowler, now at the Uni-
hands of Frederick Johnson of the R. S. Peabody versity of Wisconsin at Milwaukee, would join us in
Foundation. Not only did he teach the archaeologists Tehuacan in March.
in 1962 and 1963 how to collect carbon carefully, but Early in January 1961 I arrived in Mexico City,
he collected the majority of the carbon specimens where Peterson and I spent two weeks bargaining for
himself, organized the information about them, and and buving field equipment. Jose Luis Lorenzo, as
then selected the specimens which were to be sent Director of the Departamento de Prehistoria, Instituto
to Isotopes Incorporated for radiocarbon determina- Nacional de Antropologia e Historia, issued us an offi-
tions. Finally, it became his lot to interpret the mate- This was one of many
cial federal archaeological permit.
rial dated. ways in which the Tehuacan project was indebted to
Beside these studies which are easier to classify to- INAH. This institution, under the able direction of Dr.
gether, other specialized information came from a Eusabio Davalos Hurtado, made available to us facili-
variety of sources. Geological studies of the Tehuacan ties,equipment, and knowledge without which our task
Valley were made by Jean Brunet of the Centre would have been immeasurably more difficult. We wish
Scientifique et Technique Francais au Mexique. Geo- here to express our appreciation to Dr. Davalos for his
graphical studies of the region were made by Douglas many kindnesses.
S. Byers of the R. S. Peabody Foundation. Prehistoric We proceeded to Tehuacan, where we finally estab-
and modern irrigation systems were investigated by lished our headquarters in the downtown area, in a
Richard B. Woodbury, Acting Head of the Smithsonian rented house large enough to accommodate most of the
Office of Anthropology; James E. Neeley; and Aubrey scientific personnel as well as the household staff. At first
W. Williams, Jr., of the University of Maryland. James we used the house only as sleeping quarters and office
E. Anderson, now of the University of Toronto, under- and took our meals elsewhere. Fortunately this period
took a study of the morphologic features of the human did not last long. Mrs. Peterson —Angelita to all of us
skeletal remains taken from the excavations. Most of —came in March
assume responsibility for the
to
these studies took place in 1964 and continued after house. She brought with her Guadelupe and Francisco
the field project closed. At various times information Molina. Lupe became our cook and Francisco became
was collected concerning the ethnohistory and ethnog- indispensible to the operation in many ways.
raphy of the Tehuacan Valley by Carmen Cook, Au- Although I had undertaken some archaeological sur-
brey W. Williams, and Robert Chadwick. vey alone in 1960 and my testing of Coxcatlan Cave
All in all, these specialized studies yielded informa- had yielded significant results, our first task in January
tionfrom a wide variety of sources. I have described of 1961 was further reconnaissance. Initially, Peterson
them under different subdivisions and in terms of a and I sought by survey to delimit the region in which
—
number of different problems and yet this aspect of we were to work. The area had to be dry if we were
7
to find remains of food plants from which to establish Before the season was out we had been visited by
a sequence of developing agriculture. There had to be, Carmen Cook de Leonard, who had offered to do
as well, some measure of cultural coherence if we were ethnohistorical research for us, and by Inngard W.
to trace the rise of a civilization. We therefore com- Johnson, who had volunteered to analyze archae-
menced by working round and round the Tehuacan ological textiles.
Valley, gradually eliminating surrounding areas where Excavations at the severalsites ended in early June.
the cultural materials or artifacts characteristic of the During the summer Smith began the proposed bo-
valley were no longer dominant or areas which lacked tanical survey. During the summer also, with consider-
semidesert features. During January and February of able help from Douglas Byers and Frederick John-
1961, having defined the region in terms of these fac- son, I prepared a report of our first season’s activities
tors, we started an intensive search for sites. At first, which was published by the R. S. Peabody Founda-
Peterson and I carried on the hunt alone. We soon tion.
found that Angel Garcia Cook, a student from the The second season began in December 1961. Except
Escuela Nacional de Antropologia who was sent to for a Landrover station wagon, a wide-carriage type-
us by Jose Luis Lorenzo, had learned the technique writer, materials for visiting consultants, and replace-
of making an archaeological reconnaissance, so I left ments for digging equipment and art supplies, the
the two to carry on without me. expedition acquired few new items. We did, however,
In the meantime, I made preparations for the exca- convert the large front room of our headquarters into
vation of Coxcatlan Cave, for Fowler was to arrive a laboratory with shelves, drawers, large tables, lights,
toward the end of March. Before we could undertake and other necessary equipment. Miss Nelken and
a large-scale excavation there, we had to cut a road Angel Garcia returned to work with us again. Narciso
from the highway to the cave through the spiny scrub Tejeda had mastered the intricacies of our cataloguing
and While Peterson and the crew of workmen
cactus. system, and he carried on this duty leaving Toni Nel-
prepared the road, I began staking out the area to be ken free for other work.
excavated and started two small cross trenches into At the beginning of our second season Peterson
the cave. Fowler joined us as planned and eventually showed me sites that he had found during the sum-
took over the excavation. We started with a modest mer. We also found some promising caves in Tecorral
crew of three men, which we gradually increased to Canyon and the Arroyo Lencho Diego, in which Purron
about twenty by the end of April, as the excavation Cave is located. By this time Peterson’s survey was
became increasingly productive and the workers in- producing an average of two new sites a day. My main
creasingly well trained. Angel Garcia moved to the activity during this period was testing the likely caves
cave as student supervisor and assistant to Fowler, and sites with six of the past year’s “El Riego” crew.
where he learned excavation techniques so well that Testing usually consisted of digging a trench one-meter
he was soon well qualified to run a dig by himself. wide and as deep as the deposit of refuse across the
As specimens came in from this excavation, it be- part of the cave or site that appeared to be best strati-
came clear that we needed a laboratory assistant, and fied. We tested six caves in Tecorral Canyon, three at
again we appealed to our friend Lorenzo, who sent El Riego, one at San Andres, four in the Arroyo Lencho
Miss Antoinette Nelken of the Escuela Nacional de Diego, and one near Coxcatlan Cave. We also sank two
Antropologia to fill the post. tests in theopen site near Ajalpan. By the end of Jan-
I had by this time started excavations in the small uary we knew what we had to excavate and where we
cave called El Riego, just north of Tehuacan. There were going to do it.
I dug with Garcia as my assistant for about a week Toward the end of January, Angel Garcia joined me.
until I realized that he knew my methods as well as In Tecorral Canyon we started excavations in two
or even better than and could run the dig without me.
I caves, and the work was finished by mid-February. We
By the end of May, Garcia was running the second then commenced digging in Purron Cave, where
excavation in El Riego Cave while Fowler was about Garcia completed the initial excavations in late
to finish the first season in Coxcatlan Cave. All this March.

time Peterson, now accompanied by Francisco Molina, The excavation of Coxcatlan Cave resumed in late
continued the search for sites. His archaeological sur- February, when Fowler returned. For a time Miss
vey was running smoothly and gave every indication Nelken acted as his assistant, but he was subsequently
of continuing to do so. assisted by Robert Chadwick, then a student at Mex-
8
INTRODUCTION
ico City College, and Arturo Arvide, a student of excavation of Coxcatlan Cave, a job estimated to take
archaeology at the Escuela Nacional. three or four days. However, important features kept
Visitors to Tehuaean that winter included Frederick turning up, and the job took five weeks. With the
Johnson and Mrs. Johnson. They joined the expedition completion of this work our principal excavations were
for the puqiose of collecting charcoal for radiocarbon brought to a close.
dating. Mrs. Johnson spent much time concentrating These few months of 1962 were the most exciting
the charcoal in the samples. Johnson instructed the of my life. I would moment, to
like to digress for a
staff in techniques of collecting charcoal in order to indicate how the days were spent, and give some idea
keep it free from contaminating impurities. Next came of their pace. We were up at dawn. Breakfast. The
Thomas Whitaker and Hugh Cutler to collect modern scramble of getting equipment for each dig into the
Cucurbit a and begin their studies of our archaeological right truck. Workmen milling about. Suddenly it was
specimens. Their work continued into March, when seven o’clock, and three trucks loaded with workmen
they returned to the United States. The Johnsons left and equipment erupted from headquarters and drove
about the middle of the month. Eric Callen arrived off to the digs.
week to undertake analysis of coprolites found
in Easter On days when I did not go to one of the digs, I
in the excavations and was occupied well into May. breathed a sigh of relief, took a second cup of coffee,
In March we were joined by Douglas Byers and and went to check the cataloguing or visit the labora-
Mrs. Byers. Byers was concerned with the planning tory. All too frequently the cataloguers were in the
of our research program and the preparation of a gen- process of washing a new and exciting find. Although
eral routine for contemplated publication. When Pet- I had not intended to linger, I would stop to look at
erson and I undertook the excavation of the stratified it, and then at others. This often took me to the li-
Formative site near Ajalpan, the Byers were of great brary to see whether I could find published accounts
assistance. Indeed, Mrs. Byers excavated with pains- of similar finds from other regions.
taking care most of an important bell-shaped burial In the laboratory Toni Nelken might have an arti-
pit. fact sequence laid out on the table for my inspection.
At Ajalpan there are several kilns where
tile is manu- Or we might start a new chart, plotting types against
factured. The mining of clay for these tile factories levels. Our work was so absorbing that before we
has exposed a number of archaeological deposits in knew it lunch time was on us.
the flat which lies just west of the town. The location The kitchen, spick and span since the chaos of break-
at which we started working became known as the was usually filled with the fragrance of freshly
fast,
Ajalpan site, or Ts 204, and nearby was the bell-shaped made soup, concocted by Lupe and Angelita. When
burial pit, which we called Ts 204 C. Ts 204 A, B, D, we were lucky enough to have a visiting scientist with
and E were small excavations subsidiary to the larger us, the lunch hour would be given over to discussion
Ajalpan site. of aspects of his work. Or there might be general dis-
Through April our campaign moved ahead rapidly. cussion of happenings of the day. The mail, newly
Peterson and I worked at Ajalpan with a steadily arrived, often broughtnews of another visitor, inquiries
augmented crew which finally reached twenty-five, or congratulations on some new find, or perhaps news
while Fowler, with the help of Chadwick and Arvide, from Johnson about a radiocarbon date that could
continued to work in Coxcatlan Cave and started cause Fowler to direct greater efforts to a particular
excavation of the open site known as Coxcatlan Ter- part of Coxcatlan Cave.
race just outside the mouth of the cave. Late in April, The hours after lunch passed as quickly as the hours
Arvide and I began the excavation of Abejas Cave, before. Suddenly Narciso and “Raton” would pass the
located near Purron Cave in the Arroyo Lencho Diego. laboratory window; it was three o’clock and their, work
We then undertook to prepare Purron Cave for Angel day was done. But not ours. Toni Nelken and I would
Garcia’s return. continue our efforts, perhaps experimenting with a new
In May, Peterson, helped late in the month by method of analyzing artifacts or continuing the end-
Arvide, continued testing stratified Formative sites, less measuring, recording, and comparing of their
among them Ts 204 D and E, Ts
Ts 368e, and Ts
367, many attributes.
368w. Angel Garcia finished excavations in Purron Why had a truck come back early? It’s not early.
Cave by the end of the month. During the last week The afternoon is gone. First comes Peterson in the
of May, Chadwick was assigned the task of finishing the jeep. The hustle and bustle of unloading and shouted
9
news some minor triumph might be followed by

of preliminary reports of our findings. Those reporting
careful unwrapping of a special find. Next would come included Byers, Fowler, Garcia, Nelken, Peterson,
Fowler in the Landrover, and there would be more Smith, and myself. After the Congress a number of
unpacking, more things to see, and more discussion. delegates came Tehuacan and were taken on tours
to
Then Garcia would come in, and the whole business by Smith, Peterson, and Byers.
would be repeated. In September, Lawrence Kaplan came down to make
A “sundowner” at the Hotel Penafiel or on the plaza a careful study of the beans found in our excava-
would give us a chance to continue our discussions tions. At about the same time Flannery came back and
of the day’s finds, or of possible improvements in ways resumed collecting fauna. Some months later he began
of digging, or of whether I should visit one of the sites, to analyze faunal remains from the excavations in light
perhaps to help determine stratigraphy. The move of his newly made collection of modern fauna.
to the kitchen for Lupe’s delicious supper could hardly December 1962 brought a larger group, and we be-
interrupt our excited talk. Nor did jokes or bad puns gan the analysis of our findings. Paul Mangelsdorf,
affect us. accompanied by Mrs. Mangelsdorf, and Walton Gal-
More work in the lab to finish the afternoon’s tasks, inat came from the Botanical Museum of Harvard
or perhaps to shift the artifacts on the table, followed University to study close to 23,000 pieces of archae-
dinner. We might try using chi-square for detailed ological corn. With some assistance from Edward
inspection of artifacts in order to find new attributes. Wellhausen and William H. Hatheway of the Rocke-
Before I knew it Toni Nelken would be gone. I would feller Foundation, they were able to complete pre-
stumble to my room, slip off my clothes in the pitch liminary studies by early January.
black, and bed wondering why I stayed up
fall into In January 1963 we were joined by Miss Mary Hill
so late when was so much to be done tomorrow.
there Gilbert, who was to be our secretary, and by Miss
The next day would start the same way, but I would Nancy Hansen, who at once took over the duties of
be lucky enough to be on one of the trucks going out expedition artist but waited three months before taking
to the digs. Perhaps I intended to coordinate labora- over uxorial responsibility as Mrs. Kent Flannery.
tory and field findings or just see a new feature, but Frederick Johnson returned in February with the
before I was aware would be busily at work
of it I object of collecting radiocarbon samples from the
with a trowel, brush, and notepad. And then it Formative sites known as Ajalpan, Las Canoas, and
would be lunch time, even for me. Lupe with her usual Coatepec, and from a few others for which we had
foresight had put in food for me, and my tacos would only inadequate samples. By this time Flannery had
join those of the others in the hot coals when the cry segregated certain bones found in the lowest levels
of “Lumbre!” would bring our work to a halt. It would of Coxcatlan Cave and had tentatively identified them
be impossible to tear myself away after lunch. There as being from animals now extinct in the valley. When
would be a burial boundary
or other feature to see, the his identifications were Confirmed, Johnson took over
of a zone on a newly cleaned profile, or a
to follow the responsibility for these levels. He found additional
certain point type discovered in an unexpected zone. artifacts associated with bones of extinct animals, and
And suddenly it would be time to go home. Then we he found charcoal as well. His painstaking care in
started through the whole business again unpacking, — excavating these zones produced evidence of disturb-
discussions, comparisons. And so it went for two al- — ance by rodents, and evidence suggesting that the char-
—
most three months. coal found near the bones and artifacts had entered
In June, however, the excitement died almost as the zone through rodent burrows. It thus was not in
quickly as it had begun. Callen, Nelken, Chadwick, clear association with the finds and was of no value
and I departed early in the month, leaving Arvide for establishing the date of occupation.
and Peterson to finish the work at Coatepec and Other visitors included Robert J.
Drake, who came
Quachilco. to study shells collected in the excavations, and James
Kent Flannery appeared in July, and at once set Schoenwetter, who came to collect pollen samples and
about collecting samples of the modern fauna. This discuss the possibility of ecological studies in con-
task occupied him into August. junction with the archaeological program.
August the XXXV International Congress of
In late The Byers paid us a brief visit in April, at which
Americanists convened in Mexico City. Once again time Byers and discussed plans for publication and
I
some of our group assembled this time to make — for preparing the final reports. He and I set out during
10
INTRODUCTION
a few days when there was comparative calm at head- tices, Aubrey Williams has undertaken a thorough
quarters to survey the area above Xaco, near Chilac, study of modern irrigation practices.
to see if we could enlarge our sample of preceramic In late February we were again visited by Mr. and
open sites. We
found what we took to be a pit-house Mrs. Johnson, Johnson having come to take first steps
village, and some months later, we succeeded in uncov- leading toward a final correlation of radiocarbon dates
ering remains of one complete pit house and evidence of and cultural remains. They left about the same time
others that had been badly damaged by erosion. Byers Miss Paulina Heard arrived to take Miss Trull’s place
subsequently undertook excavation of what we took to as secretary.
be a pit house, only to find that it was the remnant of a Concomitant with and often directing our field op-
silted-up irrigation ditch. erations was the analysis of the archaeological and
The season came to a close in June, but the next environmental data. The analvsis was of two general
season opened in August when I returned to Tehuacan sorts; that done in the field with specific reference to
in the company of Callen and Smith to study the re- digging and survey procedure and that done by visiting
mains of plants. In this work we were assisted by Miss scientists on material that had been excavated in order
Gilbert, Miss Nelken, and Narciso Tejada. to help us achieve a more complete archaeological,
After this work was completed, I became involved environmental, and chronological reconstruction.
in the preparation of an application for funds for pre- Preliminary classifications and cataloguing of our
paring the manuscript and in writing and preparing specimens had begun almost as soon as we gathered
for publication a second annual report. In mid- them. One of the Nelken when she
first duties of Toni
November Peterson departed for a new job, leaving joined the expedition in April 1961 was to devise a
the laboratory and household running smoothly. general cataloguing system. However, within a short
The final season, which began in December 1963, period her work with the classification of stone arti-
was devoted primarily to analysis and reporting of our facts came to demand so much of her time that she
finds. In this work I was assisted by Miss Ann Harvey, was forced to turn over the cataloguing of the col-
a Bennington student who had training in both botany lection to Narciso Tejada. During the following sum-
and archaeology, and Miss Elizabeth Trull, a new mer he was assisted by Francisco Molina and R. Perez,
secretary. Unfortunately, Miss Trull’s work on our and during the final stages of 1963 Augustin Tejeda
manuscript was terminated when she was injured in joined this group. Initially, they washed, numbered,
an automobile accident. and classified into general categories the materials un-
In January 1964, James Anderson arrived to study covered, but by 1963 they had become so skilled that
the skeletal remains from our digs. In February, Rich- they made the actual counts of the different types from
ard B. Woodbury and Mr. and Mrs. James E. Neeley most of the surface sites.
joined us. As our field research progressed, we had The classification of stone artifacts was begun in
observed numerous signs that irrigation had been the spring of 1961 by Toni Nelken and myself, and
practiced in the Tehuacan Valley in ancient times. she was able to complete a preliminary classification
We were fortunate in persuading Woodbury, a recog- of all of the excavated stone artifacts by the beginning
nized expert on arid land, to undertake studies of this of our second season. The artifact trends shown by this
aspect of prehistoric life. Results of his studies will study indicated definite gaps in the sequences in our
appear in a later volume. first two excavations, and this information became a
Woodbury
After returned to Washington, James and determining factor in the selection of sites for exca-
Mary Ann Neeley stayed on in Tehuacan to make a vation in the 1962 season. In that year, as stone arti-
survey of irrigation features. They not only found facts were uncovered, we modified our preliminary
many “fossil’ irrigation ditches preserved by salts classifications and aligned our excavated components
carried in the water, but to our great surprise, dis- into chronological order on the basis of these classifica-
covered in the Arroyo Lencho Diego remains of a tions. In 1963, with the help of Mary Hill Gilbert, we
—
huge dam over which we had driven every day going wrote our first draft describing the artifacts. In 1964,
to and from work at Purron and Abejas caves. There the stone artifacts from surface collections were in-
were also remains of several aqueducts and other am- corporated into our report. Except for the report on
bitious engineering projects undertaken to bring water textiles contributed later by Irmgard Johnson, the
to the fields. In order that the study of irrigation fea- drafting of Volume II was almost completed by the
tures might be interpreted in terms of modern prac- time we left the field.
11
Formal pottery classification was started in the sum- ological order, and their cultural phase determined,
mer of 1962. When we were making the preliminary we began an analysis of the survey material to deter-
survey and during our testing in 1962, Peterson and I mine changes in settlement pattern and population.
had roughly sorted the surface collections into general A resurvey of many sites carried out by Neeley at this
periods. Peterson began by sorting some of the Post- time contributed greatly to our understanding of set-
classic pottery, and later classified early Formative tlement patterns. I was able to complete a study of the
(Ajalpan) ceramics. At about the same time he was preceramic sites and Neeley the Formative sites be-
doing Kent Flannery undertook a preliminary
this, fore the field season closed. Neeley’s final analysis of
classification of middle Formative sherds. In 1963 I the Classic sites and mine of the Postclassic sites were
undertook a classification of the sherds of the Forma- completed later.
tive period from the Coatepec site. Later Peterson The preparation of site reports involved processes
studied the Quachilco remains, and all three of us at- analogous in many ways Although
to the foregoing.
tempted to co-ordinate our classifications. The final a description of excavation could have been made im-
classification of sherds was carried out in 1964 under my mediately at the termination of each dig, reconstruction
direction, with the help of Ann Harvey, Mary Ann of the way of life and the environment at the time of
Neeley, and the cataloguing crew. The fine perrographic each occupation of each site had to await data from
study of Arturo Sotomayor was incorporated into several sources. By the 1964 season, however, we did
our reports at this time. During the late spring of 1963 have studies of all the artifacts from every floor, as well
Jose Luis Lorenzo had put us in touch with Ingeniero as a considerable number of studies of special materials
Sotomayor, who was then working with the laboratory from these floors. As all these data became available we
of the Departmento de Prehistoria making thin sec- began to correlate them, and Fowler started to write
tions of pottery' for the purpose of identifying minerals. his site report on Coxcatlan Cave. Using the field notes
We anticipated that studies of this sort would be of as well as data assembled by Ann Harvey from many
great value in connection with the analysis of the sources, I began preliminary drafts of reports on San
Tehuacan pottery. Marcos, Tecorral, El Riego, Purron, and Abejas caves.
In June 1964 we
also assembled a field conference At a dinner held at the Hotel Pehafiel in April 1964
to discuss early Mexican pottery. This was attended all the findings of the Tehuacan expedition were pre-
by James A. Ford, Matthew W. Wallrath, and Alfonso sented to Dr. Ignacio Bernal, Director of the Museo
Medellin Zenil, all of whom had been working on Nacional de Antropologia, for the people of Mexico,
the coast of central Veracruz; Robert Squier, who had by Mr. Byers, acting on behalf of the Tehuacan Project,
worked on the coast of southern Veracruz and Tabasco; the Robert S. Peabody Foundation, and the Trustees
John Paddock, who was familiar with the pottery' of of Phillips Academy. The formal presentation was
Monte Alban; Melvin L. Fowler, who was at that time made at this time because many of the persons involved
working near Puebla; Kent Flannery, who had worked would have found it impossible to be present when the
on the Pacific coast of Guatemala as well as in Te- Tehuacan project came to a close in August. The ac-
huacan; and Bruce Warren and Gareth Lowe, who had tual transfer of specimens occupied several weeks dur-
just completed their study of the pottery of central ing the summer.
Chiapas. This conference produced significant infor- In June Dr. Anderson returned to complete his study
mation regarding the distribution of types of pottery of the skeletal material. At about the same time Jean
found in the excavation of the Tehuacan Valiev. J
Brunet undertook a study of the geology of the Te-
Thus, again by the time we left the field, we had huacan Valley.
at least tentative descriptions of most of the ceramics July 1964 was the beginning of the end. Although
and the bulk of our third volume in rough draft, even we were working feverishly to finish the analysis of
though it was not readied for publication until some some of our material, we nevertheless began to ship
time later. our excavated specimens to the Museo Nacional in
One of the final steps of the field analysis took place Mexico, to send materials gathered bv the site survey
in 1964 when James Neeley and I attempted to analyze to the Museo de la Revolution in Puebla, and to set
the data from the 454 sites found in the archaeological some specimens aside for a museum to be established
reconnaissance. Obviously such a study had to await in Tehuacan.
the classification of all the stone, bone, shell, and perish- August saw us packing, selling the furniture and
able artifacts as well as the ceramics. Once these re- equipment or giving some of it to friends and em-
mains had been classified, the sites aligned in chron- ployees, and making one thousand and one final ar-
12
INTRODUCTION
rangements. Then suddenly, on August 20, the house helped us and to whom, again, I owe much. In conclu-
was empty. The Tehuacan field project was over. sion, let me say that all of us who participated in this ex-
This brief summary relating how so many people pedition worked very hard, struggled to keep to an
contributed to making the Tehuacan project a success impossible schedule —necessitated in some respects by
will, I hope, serve as a way of thanking all of them. my own compulsive neurosis as well as by the usual
When I wrote it, it seemed a better way of expressing —
shrinking exchequer and in many cases pushed our-
my gratitude than the usual dull acknowledgments selves to do a little better than our best. Thus I would
with a thank you to a long list of names. I am
have, I like not only heartily to thank my colleagues, but also
afraid, omitted some of the many Tehuacaneros who to ask their forgiveness.
REFERENCES
Barghoorn et al., 1954.
MacNeish, 1959.
Mangelsdorf et al., 1956.
Sears and Clisbv, 1955.
Wellhausen et al., 1952.
13
CHAPTER 1
An Interdisciplinary Approach
to an Archaeological Problem
Richard S. MacNeish
he immediate objective Tehuacan Archae-

of the the northeast, sequences in central Veracruz (Medellin
T was the establishment

ological-Botanical Project
of an uninterrupted archaeological column span-
ning the time between the first human occupation of
Zenil 1960) and northern Veracruz (Ekholm 1944, Mac-
Neish 1954) exhibit developments from villages to cities.
To the southeast, the early Olmecs developed toward a
the valley and the Spanish Conquest. This hoped-for civilization in southern Veracruz and adjacent parts of
archaeological column would almost certainly be di- Tabasco (Drucker 1943). To the south was the sequence
visible into cultural phases, representing progressively at Monte Alban in central Oaxaca (Caso 1938), while even
more advanced societies, culminating in civilization. further south, in the Mayan region, were a number of
Thus we should, in effect, recover a sequence of cul- sequences from folk to urban culture (Kidder, et. al, 1946).
tures, which we assumed might be representative of Second, we had assumed that the rise of civilization
developing civilization in Mesoamerica, one of the was intimately connected with the development of
major culture areas of the world (Kirchhoff f952). This agriculture. In the New World and Mesoamerica, agri-
sequence we could then compare with similar develop- culture, we believed, depended on the domestication
ments in other areas, and by thoroughly studying and of corn, Zea mays.
classifying the various features of each, determine the The origin of corn was one of the major puzzles of
reason that each unit came into being (Steward 1955). New World botany and archaeology. Various hypoth-
Thus the ultimate objective of the Tehuacan project eses had been advanced to explain the existence of
was the discovery of processes and causes leading to this grain which, unlike the Old World grains —
wheat,
the rise of primary civilization. In order to attain this barley, rye, oats, and millet —
had no identifiable wild
end we would require several sequences of culture in ancestry, but none was truly satisfactory. In the In-
relatively independent areas, each leading to civiliza- troduction we recounted, in detail, the series of steps
tion. by which the regions be searched for the origins
to
Our assumption that the sequence from the Te- of this important cereal were narrowed by eliminating
huacan Valley would be typical of Mesoamerica was first Asia and then South America as possible centers,
based on two bodies of data derived from previous and next how the northern limit of the target zone was
research. First, the region is centrally located in Meso- carried southward from New Mexico, through the
america. Previous investigations on all sides of it had Mexican states of Tamaulipas, Sonora, and Chihua-
revealed sequences of development from early villages hua to the Valley of Mexico. Discovery of pollen identi-
to urban civilization. To the northwest, in the Valley of maize at a depth of some seventy meters
fied as that of
Mexico, a relatively complete sequence beginning with from Mexico City made plant geneticists
in a drill core
the Zaeatenco villagers of about 1000 b.c. led to the city- aware that the ancestor of corn was corn itself, and
dwelling, empire-building Aztecs (Vaillant 1966). To that corn was not developed from teosinte or Trip-
14
INTERDISCIPLINARY APPROACH
sacum, or a cross between them. As this pollen came Now let us turn again to one of our basic problems.
from levels attributed to the Yarmouth interglacial, it In order to establish a chronologic arrangement of the
cannot be connected with human activity and is pre- cultural units we had discovered, we resorted to a
sumed to be that of wild maize. Other pollen, abun- variety of methods, some of which established a rela-
dantly present in upper levels, is judged to represent tive, but not very precise, chronology, while others
cultivated corn (Barghoorn et al. 1954). established a more precise chronology. Regardless of
We have also told in the Introduction how excava- whether the resulting chronology was relative or pre-
tions in Santa Marta Cave gave us reason to believe cise, we relied on a number of methods from a number
that the original homeland of corn lay north of Chiapas of fields of endeavor, using a wide range of techniques.
and how Mangelsdorf and I ultimately confined the We brought all to bear on a single problem chro- —
area in which to concentrate the search to the Te- nology.
huacan Valley. When preliminary testing in Coxcatlan One method for establishing chronology relied on
Cave produced cobs of primitive corn that were dated stratigraphy. We chose for excavation those sites which
to 3610 b.c. ± 250 years (M-1089, Crane and Griffin would give the maximum amount of cultural stra-
1962), it was clear that the Tehuacan Valley, more tigraphy. We dug a total of sixteen pits and uncovered
than any other spot yet tested, gave promise of sup- in them a total of 156 occupational layers, or zones, as
plying an answer to this problem. we shall call them.

In order to obtain, in Tehuacan, a complete sequence Each zone we considered to be a distinct deposit,
of culture representative of that in Mesoamerica and laid down during a single interval of time. Usually
of use in wider comparison, two basic problems had each zone was relatively uniform in color and in con-
to be solved —
the cultural chronology had to be de- tent, whether it was ash, roof fall, waterlaid sands,
termined and cultural contexts had to be established. clay, or other material, and each zone was capped
Perhaps I should say that we had to reconstruct ancient by a well-defined stratum or break in the soil profile.
cultures and their environments throughout our se- This stratigraphic discontinuity might consist of a char-
quence. At times we studied both problems together, coal floor, a layer of vegetal material, a tramped dirt
but we preferred to study chronology before we turned floor, or other matter, but it definitely marked the end
to interpretation, because we felt the latter could not of the period of deposition of refuse in the zone be-
be truly understood unless the former was known. neath.
In attacking these problems, we used an interdis- Coxcatlan Cave (Tc 50) was our best stratified site,
ciplinary approach, bringing to bear upon the solution for we identified there 28 zones. Each of the lower
of our rather specific archaeological problems facts or zones, XXVIII-XXV, consisted principally of an irreg-
conclusions from a number of scientific fields, derived ular stratum of fallen rock and wind-blown sands, and
from an even wider range of techniques and methods. each was capped by a thin charcoal floor in a small
Thus the project, instead of relying on one kind of central portion of the cave. It should be noted that Zone
evidence derived from a single line of research with XXVIII, capping the base level of the cave, was with-
its peculiar set of techniques, made use of many kinds out artifacts but did contain animal bones. Zones XXIII
of evidence from many directions, all leading to the and XXIV were layers of ash and/or fine dirt or silt
solution of our problems. capped by well-defined charcoal floors in portions of
Needless to say, the so-called facts of the various the west end of the cave. Zones XXII to XIX were
disciplines must be derived by a careful and well- similar, but covered small areas in the east end of the
executed and laboratory techniques. This
set of field cave. Zones XIV-XVIII, and Zone XI as well, were also
is particularly true of archaeology and our techniques ashy layers, but they extended over the entire area of
will be described in some detail in the following chap- the cave and were capped by charcoal or burned floors
ter.Mainly because of the careful application of these with ever-increasing amounts of preserved vegetable
techniques we were able to attack successfully the remains. Zones XII and XIII were of similar content in
wider problems with a great mass of data. From many the west end of the cave, as were Zones VIII, IX, and X
standpoints, our techniques were unique for Meso- in the east end. Zone VII represented a break in the
america. This is an area where great designs for re- sequence, as it was an all-encompassing layer of vege-
search often have been defined, but rarely implemented table remains. Zones I to VI were floors of vegetable
successfully. This, is because well-designed
I believe, refuse over ash or disintegrated rock. The top zone was
superstructures often on a poorly laid foundation,
sit a layer of goat dung, refuse not directly made by man.
built with rather inadequate techniques. The zones in the East Niche of El Riego Cave (Tc
15
35e) were very similar in makeup to the upper zones a zone of brown refuse, Zone 4. A
some 15 cm.
layer
of Coxcatlan Cave. Zones C-E were ash layers capped thick including the transition from Zone 4 to Zone 3
by floors of preserved refuse that extended over the was peeled off and designated level 3-4. Above this
whole surface area of this small cave. Zone B was a was a deposit of brown refuse that we divided into
layer of ash and vegetal material and was capped three zones: the top zone, 1, above a charcoal plaster
by a plaster floor. Zone A consisted of a layer of vege- floor; asecond layer, Zone 2, from which graves and
tal material above the plaster floor, and it, in turn, was grave pits could be traced downward; and the lowest,
capped by goat dung. Zone 3, composed of refuse beneath the burial pit.
The stratigraphy in Purron and Abejas caves re- At the open sites less clearly defined floors capped
sembled the lower zones of Coxcatlan Cave, in that the zones. Perishable materials were not found in
there were rather thick layers of fallen rock, sand, and them, and conditions at these sites therefore differed
finer sediments, capped by recognizable floors com- from those encountered in caves. The site known as
posed of charcoal, vegetal remains, or burned material. Coxcatlan Terrace (Ts 51), below Coxcatlan Cave, was
The earliest thirteen of the twenty-five zones in Purron capped by Zones A and B that differed only slightly
Cave (Tc 272), Zones L-U, covered small areas, meas- in color and appeared to have been formed by material
uring approximately 5 by 8 meters, in the southwest dumped down the talus from the cave above. They
part of the cave; the occupational floor capping each covered stream-deposited sediments called Zone C.
zone comprised chiefly charcoal. The upper twelve Beneath this was a burned layer forming a charcoal
zones, A-K, were capped by floors of charcoal and floor labeled Zone D. Zone E consisted of a thin layer
vegetal remains, laid down over the entire surface of of refuse beneath Zone D. These zones, like those in
the shelter. the other open sites, were only tested. Undoubtedly
Cave (Tc 307) there were eleven zones, all
In Abejas they covered much of the terrace in front of Coxcatlan
of limited extent, and all but the uppermost capped bv Cave.
charcoal or ash. The upper Zone A included vegetal The other open sites that we tested were consider-
remains. The top layer consisted of a capping of goat ably larger. Coatepec (Ts 368), was more than 200-
dung. 300 meters long by 100 meters wide. Refuse was very
San Marcos Cave (Tc 254) was small. In it, the five deep and appeared to have been laid down in a short
zones, B-F, were comprised of small flakes of fallen time. In the deepest trench at the east end of this site
rock overlain bv thin layers of vegetal remains. Refuse (Ts 368e), we recognized 18 layers. All but the three
in the rock fill of Zone C caused it to be designated lowest were composed of dark lenses of refuse and
a separate layer; the charcoal floor above it was rec- earth capped by floors. The lowest layers, K K2
1
, ,
and
ognized as a separate zone and called C The upper-
1
. K3 were arbitrary' divisions of a rather deep layer of
,
most layer, goat dung, was identified as Zone A. refuse. Zone D near the top also had from one to three
Tecorral Cave (Tc 255), next to San Marcos Cave, charcoal floors, and although they actually comprised
was even smaller. A small floor of charcoal and vegetal three zones, because of the paucity of artifacts in each
remains at the top of the deposit was designated Zone they were usually combined for purposes of analysis.
A. It apparently represented a short occupation of the A smaller test at the west end of the site (Ts 368w),
cave. Below it was a floor of ash from which a pit ex- disclosed only five zones, A-E, which were refuse-
tended downward. Together they constituted Zone B. capped floors that had abutted against a small pyramid
Although no clearly defined floor was seen below and eventually covered it completely.
Zone B, the rock fill with artifacts that extended down At Quachilco (Tr 218) there were similar layers of
to the base rock of the cave was designated Zone C. refuse capped by floors. There were six such zones in
The stratigraphy in the West Niche of El Riego Test 10, seven in Test 11, and eight in Test 6.
Cave (Tc 35w) was even less clear, and the floors cap- Conditions at Las Canoas (Ts 367) were slightly dif-
ping its six zones were less well defined. The lowest ferent in that the two top zones, A and B, were thick
occupation layer. Zone 6, covered only four or five strata of refuse, Zone A being dark brown and Zone
square meters. It lay directly on the bottom of the cave B lightbrown. Above these strata, laid down during the
and was readily discernible as a thin layer of ash and final Venta Salada period, we recognized no floors.
charcoal. A layer of dark refuse covering about twice Below Zone B was a recognizable charcoal floor over a
the area, and lying above Zone 6, was called Zone 5. little sandy refuse designated Zone C. Below that was
The upper margin of Zone 5 merged gradually into Zone D, a thick layer of charcoal covering a layer of
16
—
refuse designated Zone D 1 which covered sterile soil.

,
to changes in tastes and habits of work that they them-
At the Ajalpan site (Ts 204), with ten layers of refuse, selves —
undergo slight changes and that a proper iden-
conditions resembled those at Las Canoas. Here again, tification of artifact modes will reveal cultural change
Zones A and B were refuse, light brown and dark and trends.
brown respectively, extending down a meter from the Initially, we established artifact types on the basis
surface. Zones C and D were waterlaid sands contain- of mode clusters — series of interrelated features of
ing some refuse. Zone E below them was a zone in artifacts having significance in time and space. We first
which there was charcoal. Next came Zone F 1 a hard- ,

determined artifact types and trends for the rich strati-
burned floor capping rather a small thin layer of fied site in Coxcatlan Cave, Tc 50. By using trends of
burned refuse, Zone F. The latter covered a burned artifact types from Tc 50 as a standard with which we
floor with pits extending down to it, Zone G 1
over refuse could compare artifact types from other stratified sites,
Zone G. At the very bottom of the site, above sterile or for that matter, from sites represented only by sur-
soils, was Zone H, a thin charcoal floor representing face collections, it was possible to correlate any other
a relatively short occupation. site provisionally with the sequence in Coxcatlan Cave,
Another part of the same site was designated Ts 204 assuming that a zone or component was probably
D. Here again there was at the top a layer of undif- closest in time to that zone in Tc 50 with which it
ferentiated refuse, Zone A-B. Beneath this refuse came shared the greatest number of artifact types. As a re-
Zones C and D, also refuse, but with some sand con- sult of our studies of types and sequences of artifacts
tent. A definite floor (Zone Super-E or E 1

)
overlay a from the 156 excavated components and approximately
stratum of refuse called Zone E. This, in turn, rested 350 surface sites, we were able to align all in relative
on refuse, Zone Sub-E, that capped the sterile clay of chronological order (see Table 32, Volume II, and
the vicinity. Volume IV).
The various zones described above represent closed Typology, besides making it possible to describe
intervals of time during which people laid down refuse artifacts and assisting us in aligning the archaeological
and sediments built up. These also were the temporal components in order of relative chronology, also served
units of occupation or the components that became the to define the actual components. Needless to say, some
basis of the stratigraphic sequence of cultural materials archaeological components fitted into the sequence
and cultural phases in the Tehuacan Valley. more easily than others, and some could be more
These 156 components of the stratified sites revealed easily defined than others. A component represents the
cultural chronology by reason of changes in their con- occupation of a place during a single continuous unit
tents, comprising in round numbers 250,000 artifacts of time by a single cultural group. Because some com-
of clay and 13,000 of stone, 25,000 remains of domesti- ponents were less easily defined than others, we divided
cated plants, and 70 human burials. No single site we all components among three categories: pure
the
excavated yielded the full sequence. Furthermore, components, probable components, and possible com-
studies of the actual strata by visiting geologists and ponents. By a pure component we mean a zone which
soil scientists revealed no method for correlating zones is so well defined physically that it could only have
of one site with those of another. Studies of 11,000 been occupied for what is considered a single moment
bones of animals, 25,000 molluscan shells, and 80,000 in time, and which contains enough artifacts so that
remains of wild plants, all taken from excavations, we may feel sure that it represents only one cultural
gave hints of the alignment of the various zones, but group. Probable components are those whose physical
were not sufficiently sensitive in our region to aid in features lack sharp definition or whose inventory of
establishing a relative chronology. We had hoped that artifacts lacks a sufficient number of diagnostic types
pollen studies might add to this aspect of the research, so that we cannot with certainty say that they repre-
but the pollen was too poorlv preserved. sent a single occupation by a single cultural group.
Perhaps the best means of establishing relative chro- Sites or zonestermed possible components have even
nology was by the taxonomic method, using material more limited complexes of artifacts or less distinct evi-
from the stratified sites. Assumptions, techniques, and dence of occupation, so that it is only barely possible
applications of this method, as we used it, will be fully that they were occupied by one group at one time.
explained Volumes II and III. Suffice it to say here
in Another kind of taxonomic study, that of remains of
that this method depends on isolating artifact modes domesticated plants from the sixty excavated compo-
those attributes of an artifact which are so sensitive nents in which preserved food stuffs were found, con-
17
—
firmed at least part of the sequence of components lished by radiocarbon dates served to place in time
based on artifact typology. Races of com determined the zones in the Tehuacan Valley in which they were
by Mangelsdorf and Galinat allowed them to align found. Dates from other regions could thus be applied
in chronological order Zones I-XIII of To 50, Zones to the sequence which we had
in the valley. In cases in
B-F of Tc 254, Zones A-K of Tc 272, Zones A-E of Tc a radiocarbon date for a zone in which a “foreign”
35e, and the top zones of Tc 307, Tc 35w, and Tc 255. object was found, it could be used as a measure of
Kaplan’s study of the beans demonstrated parallel se- correctness of our date. In cases in which we
quences and a study of the cucurbits by Whitaker and had a number of confirming dates in the Tehuacan
Cutler confirmed the order established by the corn series, they could be used as a measure of correctness
and also helped to establish the position of Zone XIV of the “foreign” dates. The same may be said regard-
of Tc 50. Smith’s study of other domesticated plants ing types native to the Tehuacan Valley which were
produced results in harmony with the relative position identified in sequences in other regions. Such cross-
of components revealed by the foregoing plants. dating not only confirmed the relative chronology but
Thus, our relative chronology was established on supported the validity of our series of radiocarbon
the basis of stratigraphy, artifact typology, and the dates, particularly in the case of dates for the later
developmental sequence of domesticated plants. We phases when pottery was plentiful.
attempted to supplement our data on chronology with All in all, our sequence of components, established
studies of the hydration of obsidian. Although this tech- bv making use methods and techniques of sev-
of the
nique seemed to offer promise (Friedman and Smith be based upon a fairly firm
eral disciplines, appears to
1960; Evans and Meggers 1960), painstaking studies foundation. Obviously, the sequence would be im-
by Dr. Donovan Clark failed to substantiate its value proved by larger numbers of artifacts or remains of
as a gauge of time in the Tehuacan Valley. plants or animals from certain segments, by more
Convincing substantiation of the relative chronology radiocarbon dates, and by thorough mineralogical
came from chronometric studies. After we had tenta- studies of trade objects, along with more and better
tively established the relative chronology, we care- radiocarbon dates to place these objects beyond ques-
components to be dated by the radio-
fully selected tion in the time scale of their native region.
carbon method so that we would have dates repre- Comparison of artifact types and related data from
senting the entire sequence. All told, we obtained over each component not only yields chronological informa-
120 radiocarbon dates for sixty-two of the 156 compo- tion, but also information on another level of abstrac-
nents. As a result, no period of over 500 years in the tion. For instance, comparison of our regional artifact
last nine millennia is without at least one radiocarbon tvpes with those of other regions could yield informa-
date. The results completely confirm and substantiate tion not only useful for cross-dating but also for
the ordering of the less precise relative chronology. determining spatial relationships, such as the origin
Further statistical study of the artifact sequence in all and spread of these artifact types.
components in conjunction with the radiocarbon ages In addition, comparison of the various components
of the selected components made it possible to estimate with each other and with clusters of artifact types with-
the ages of the undated components. This analysis, like in the Tehuacan sequence allowed us to discern major
that of the relative chronology, will be fully explained and significant periods of culture change. This latter
by Johnson and MacXeish in Volume IV. process revealed that there were eight spurts or clus-
Dates mentioned in this and the two succeeding vol- ters of new artifact types and other diagnostic traits
umes are used only as rough approximations and may in our long relative chronology, rather than just grad-
have a range of error of as much as 500 years either ual accumulations of new assemblages. We believe
way. We have used approximate dates in order to give these spurts represent periods of significant cultural
the reader some temporal orientation, since the first change. The periods or segments of our sequence which
three volumes were written before analyses for Volume have new complexes of types recurring in a number of
IV could be completed. sequential components are designated “phases.” Our
—
We employed still another method cross-dating actual comparison of artifact types and other traits
to make our dates more precise. Our studies of arti- from each component from Tehuacan revealed nine
fact types at once revealed objects —
mainly pot- phases that we called, from early to late, Ajuereado,
sherds—brought to the Tehuacan Valley from outside El Riego, Coxcatlan, Abejas, Purron, Ajalpan, Santa
sources. Those artifact types whose position in the cul- Maria, Palo Blanco, and Venta Salada.
tural sequence of their native regions was well estab- Throughout these reports such terms as Early Man,
18
Incipient Agriculture or Archaic, Formative, Classic, Schoenwetter, Elso S. Barghoorn, and Henry Irwin
and Postclassic are used. Obviously, this usage infers have worked on samples taken subsequently, none
a further set of units of classification. I use these terms has been able to isolate a sufficient quantity of pollen
not as stages but as periods. The Early Man period from the earlier components to warrant any statements
would be before 7000 b.c. (in Tehuacan, the Ajuereado about stratigraphy. Pollen grains appear to undergo
phase), the Archaic or Incipient Agriculture period a progressive reduction in density with the passage of
would extend from roughly 7000 b.c. to 2300 b.c. (the time, eventually disappearing. No pollen that is not
El Riego, Coxcatlan, and Abejas phases), the early representative of the present vegetation has yet been
Formative period from about 2300 to 900 b.c. (the Pu- found in any sample. We are forced to conclude that
rron and Ajalpan phases), the middle Formative from the amount of pollen is and the number
insufficient
1000 or 900 b.c. to about 600 or 500 b.c. (early Santa of species represented by preserved grains is inad-
Maria phase), the late Formative period from before equate to disclose a record of past climates in the
600 b.c. to 200 b.c. or the beginning of the Christian Tehuacan Valley.
era (late Santa Maria phase), the Classic period from Studies of plant remains indicate no change in cli-
before the time of Christ to a.d. 700 or 800 (the Palo mate any magnitude (see Chapter 12). Bones of
of
Blanco phase), and the Postclassic period from a.d. 700 animals from very early levels include those of several
or 800 to the time of the Spanish Conquest (the Venta species now extinct or not represented in the Tehuacan
Salada phase). Let me state specifically that these Valley (see Chapter 8). It is possible to postulate the
terms are not meant to designate stages in a unilinear existence of a climate in early Ajuereado time that was
cultural development. I have considerable doubts that slightly cooler than that of the present day, but cor-
any sort of “unilinear” scheme of evolutionary stages roborating data is lacking.
led to civilization in Mesoamerica. Detailed studies of micro-environments, including
Although the phases of Tehuacan prehistory were rainfall, drainage, exposure, and evapotranspiration,
defined primarily on the basis of artifact types and of soils and bedrock, of minute variations in tempera-
cultural traits, a full understanding of the culture each tures, and of the plant cover that has developed in
represented was obtained by a conjunctive or interpre- response to variations in these factors may someday
tive study using an interdisciplinary approach. Again, be combined into a whole which will shed much-
as in our interdisciplinary approach to chronology, we needed light on processes of cultural differentiation
attempted to use information derived from a number and specialization in restricted environmental niches
of fields in our effort to reconstruct the ancient way of and on processes of cultural evolution within one val-
life of people who made the artifacts that we now clas- ley. We had hoped to take preliminary steps in this
sified into arbitrary divisions termed phases. direction, but because necessary data were unavailable,
Studies of the geology and modern geography of the progress along this road has not been either great or
region suggest that the physiography and topography sure. Such environmental information as we have
have not undergone drastic change since men first gathered is presented below in Chapters 3, 4, 5, 8, and
entered the valley. Geological studies, however, show 12 .
that there has apparently been a continuous lowering In order to reconstruct the extent to which the pre-
of the level of ground water during that interval. His- historic inhabitants exploited their environment it was
torical data, also, show that a marked reduction in necessary to utilize data derived from researches
effective water supply has occurred in at least one carried out by scholars from several disciplines. Their
town within historic time. Geological studies hint at work, most of which is presented in this volume, covers
possible sources of raw material for aboriginal weap- studies of both wild and domesticated plants, includ-
ons, tools, or works of art. ing corn, beans, cucurbits, and other cultigens; the first
We had hoped to be able to reconstruct past cli- published list of the fauna of the valley and an interpre-
mates, or at least to point toward suggestions of possible tation of the significance of remains of animals found
changes in climate through studies of land snails and in local archaeological sites; measurements and ob-
of pollen secured from our excavation. A study of servations on such human skeletal remains as survived
snails was undertaken, but although preliminary results to us; research on physical and
historical geology and
suggest slight climatic changes, it proved impossible hydrogeology; a consideration of the climate of the
to carry this work to a satisfactory completion. Pollen valley; and an examination of the developing practice
found in samples taken in 1961 led us to hope that of irrigation. An estimate of the actual sustenance of
pollen studies would be rewarding. Although James the ancient inhabitants is based upon interpretation
19
Hd
A
/s
Fig. 2. The interdisciplinary solution of archaeological problems.
20
—
of three kinds of excavated materials: remains of food information derived from the more abundant garbage
plants and animals found in the various components, remains.
feces, and tools connected directly with subsistence. A study of the tools and features connected with
Owing to the dryness of the valley, fruit pits, frag- subsistence activities is obviously the least reliable
ments of stems and leaves, chewed quids, tiny seeds method of estimating subsistence. Here one may as-
— literally any object that the ancient inhabitants sume on the basis of ethnographic data that projectile
dropped on the cave floors or laid aside for storage points indicate hunting; that mortars, pestles, milling
have been preserved, albeit in a dehydrated state. To- stones,and mullers represent plant collecting; or that
gether, these normally perishable remains and the dis- manos, metates, comales, and molcajetes are used in
carded bones of food animals constitute what we have the preparation of agricultural foods. Obviously tools
referred to as “refuse” or garbage. that we assume were for preparing agricultural prod-
Before we could analyze the refuse with a view to ucts could also be used for wild plants, and vice versa.
reconstructing the sustenance of the early people, it Further, it is easier to make and then lose a projectile
was necessary to identify the species of animals point than it is to do the same with a rnetate. In spite
represented and to estimate the minimum amount of of these weaknesses, a study of technology permitted
edible meat derived from these animals. Remains of us to make rather rough estimates of the proportions
both wild and domesticated plants, once they were of food derived from the several sources, and a floor-
identified, could also be expressed in terms of the mini- to-floor comparison of these estimates did show trends
mum amount of food that they would have supplied. in subsistence.
Thus for each zone for which there were preserved Although each of the kinds of subsistence data has
foodstuffs, we had estimates of the total amount of food definite limitations, estimates based on a study of all
which these remains supplied. We could divide this of them, together with some knowledge of ethnography
total among sources of food —meat, wild plants, and and characteristics of the skeletal remains, did allow
cultivated plants — in order to arrive at percentages of us to make a more acute judgment about subsistence
the diet derived from each source. A floor-to-floor activities. Since the different kinds of subsistence ac-
comparison of these percentages revealed a shifting tivitiesand changing trends through time exhibited
reliance on the different sources of food with the pas- by artifacts and other materials associated with them
sage of time. Obviously the range of error in such are discussed at length in the concluding chapter, I
calculations is rather wide, and such factors as uneven will not go into further detail here. Techniques of
preservation of vegetal foods in earlier levels forced preparing food — as
reflected by the artifacts, the
us to make statistical adjustments. Be that as it may, refuse, the feces, and even the skeletal remains also —
our errors are consistent for each floor through time, changed through time. These techniques are discussed
and the trends revealed by the estimates are probably in the final chapter, as well as in the chapters below
valid.These estimates are discussed in detail in Chap- dealing specifically with animal bones, plants, copro-
ter 14. lites, and physical anthropology.
Over two hundred feces from the sixty floors in which We also attempted to reconstruct the ancient tech-
perishable materials survived were available for an- nology and take note of any changes. This is what
alysis (see Chapter 13). Of these, more than one archaeologists often claim they can do best by making
hundred are interpreted as being of human origin. inferences from careful studies of the material culture
Thousands of small particles of food from the feces of living peoples. Quite frankly, neither we nor the
were separated and identified under the microscope, majority of Mesoamerican archaeologists have done
in an effort to reconstruct the diet for each floor and this kind of reconstruction as thoroughly as we should.
sequential phase. However, estimates of proportions We, like most of our colleagues, studied the chipped-
of the sustenance from different sources based on an stone tools. After examining specimens under a hand
analysis of coprolites are subject to peculiar kinds of lens or microscope and taking note of features charac-
errors, for finely ground corn in tortillas seemingly teristically produced by techniques that we tentatively
leaves no trace, while remains of unground plant food identified, we have been able to turn to ethnographic
and roughage may be easily found. Nevertheless the and modern accounts of flint-knapping and make in-
identity of many plant and animal foods and changes ferences regarding techniques used by ancient inhabi-
in their use were revealed by analysis of the human tants of the Tehuacan Valley. Although our conclusions
coprolites and served to supplement or corroborate may in the main be valid, it is unfortunate that we were
21
— —
unable to undertake thorough studies of implements pology concerning age, and pathology is
sex, stature,
under the microscope and attempt by experiment to as yet too limited to be useful demographic study.
in a
reproduce features so identified, as was done in the Future excavations in the Tehuacan Valley could easily
pioneering studies of Semenov and his colleagues. Such supplement these data.
investigations are essential to a more exact reconstruc- A second method of deriving an estimate of popula-
tion of flint-knapping techniques and ways of using tions and population trends is to work backward in time
tools than we now have. from post-Conquest records. Some of the data of Cook,
Criticism of the same order could be leveled against Simpson, and Borah relate to the Tehuacan Valley. As
our studies of pottery. Although Ing. Arturo Sotomayor listed and discussed in Chapter 3 below, counts of post-
was able to make thin sections of sherds and complete Conquest population of many towns exist, though often
petrographic analyses, we did not have sufficient time in terms of families or tribute payers. These estimates
to sample clay beds in the valley in an effort to find justify a guess that the total population of the valley
sources of the clays. Although we could easily deter- about a.d. 1514 was between 90,000 and 120,000. Close
mine the hardness of the pottery and could make judg- study of documents relating to the Senorio of Teotit-
ments about processes of manufacture and firing, we Camino in Conquest and pre-Conquest times
lan del
had no opportunity to undertake experimental and would probably allow even more exact estimates.
ethnographic studies in support of these observations. Nevertheless, the post-Conquest population figures
Although surface finishes, vessel forms, and decorations provide some basis for judging the accuracy of the site
are carefully described, there has been little attempt survey and estimate of the total number of occupied
to understand their psychological meanings. Studies settlements of the Venta Salada phase. By dividing the
of the mineralogy and chemistry of paints and of our total population at the time of the Conquest by the
few copper specimens is even more superficial. number of known sites of the Venta Salada phase, we
Analysis and description of objects made of fiber obtain a ratio which we can then apply to settlements
string, nets, mats, sandals, and textiles —
is somewhat represented by occupation floors in excavated sites and
better, because these —
especially the textiles under- — by survey sites during each of the preceding phases
went careful study and could be viewed in the light Palo Blanco, Santa Maria, Ajalpan, and so on. Thus
of knowledge of similar products made by living peo- the trend of population can be worked out for the
ples or known from other dry archaeological sites. An- whole sequence, using the 454 sites found by a survey
alysis of these objects would have been more valuable and the 156 occupations uncovered by excavations.
if we had been able consistently to identify the fibers Data obtained from the site survey and from exca-
from which they were made. An investigation of com- vations afford a basis for outlining settlement patterns
parable industries among living peoples in the area during each of the sequential phases. Studies of arti-
would also have been profitable. factsprovided knowledge of the forms characteristic of
With the aid of a microscope we studied the ground- the several phases. This, in turn, allowed us to place
stone tools, and objects of bone, shell, antler, and wood. about 350 of the sites recorded by the survey in their
Here again, for comparative data we often relied on proper phases. Next, the survey data for each site was
discussions of similar industries in ethnographic ac- examined for information regarding location with
counts. We made no attempt to reconstruct the tech- respect to topographic features, size and plan, arrange-
nologies of stone quarrying, masonry, architecture, ment of settlement with respect to architectural fea-
sculpture, and other aspects of aboriginal life. Thus tures, and the relationship of the site to other sites with-
our discussion of the total Tehuacan technology to be in the same phase. In this manner we gradually came to
described in seemingly endless pages in Volumes II, some understanding of the pattern of settlements char-
III, V, and VI is far from complete. I hope that future acteristic of each phase. Had we been able to excavate
investigators will be able to make great improvements a number of examples of the different types of settle-
on it. ment for each phase, our study would be considerably
Our studies of the aboriginal population are also enriched. Even so, it is quite complete and relatively
incomplete. Obviously seventy-three burials, often unique for Mexico, particularly in the ceramic phases.
fragmentary, representing nine cultural phases which In the preceramic phases the information from the
span nearly 10,000 years are a very inadequate sample excavated components was even more instructive than
of the prehistoric population of the valley and certainly the site survey data. Populations of camps could be
provide little basis for judging population trends. The estimated on the basis of the size of occupied zones and
valuable information derived from physical anthro- the number of hearths each contained, qualified by data
22
limiting the period of occupation derived from animal changes in them may reflect changes in the kinship
and plant remains and as documented by analysis of system. Different kinds of houses and architectural
feces. Together, these gave us some basis for judging features yield some information about class as well as
the time of year at which people moved to the site about political and religious organization. Analysis of
and the length of time they stayed. Estimates of popula- artifacts suggests the possible existence of full-time
tions could be made by dividing the duration of each specialists or craftsmen. Discovery of objects brought
occupation in terms of approximate number of days by trade from other regions offers grounds for forming
into the estimated total amount of foodstuffs, expressed ideas about commerce. The finding of sites of the Palo
in liters, represented by the refuse in the zone. On the Blanco and Venta Salada phases connected with a salt
assumption that a person ate about a liter of food a industry unfolds information about special products
day, the result should give an indication of the popula- carried by commerce.
tion occupying a certain site at a given time. An analy- The aspects of society concerned with the super-
sis of populationand settlement pattern will be one of natural or with religious organization can be inferred
the main concerns of Volume V. on the basis of information derived from a variety of
All this information — length of occupation in terms sources, included burials, figurines, types of incense
of seasons, density of population, seasonal occupation burners or xantiles, and types of architectural features
of camps of the earlier phases —can be viewed in the and their arrangement. Again the ritual codices, par-
light of reconstructed subsistence patterns and envi- ticularly the BorgiaCodex, which may be closely con-
ronment during the earlier preceramic phases and thus nected with the Venta Salada phase, reveal much about
used as a basis on which to make postulations regard- ceremonies, gods, and ritualistic and religious aspects
ing the organization of society among the earliest of the life of the people.
people. Steward and other social anthropologists have Such inferences about social aspects of the people
established that peoples resident in a harsh desert en- known to us only through archaeological phases will be
vironment in which there is a low population density, dealt with in future volumes of this series. Since the
whose subsistence is based upon the food-collecting phases of the Tehuacan sequence will be mentioned
habits of seasonal nomadism that brings small groups repeatedly throughout this and succeeding volumes,
together in seasons of abundance, are usually organized brief summaries of the characteristics of each are given
into bands, recognize band territories, and have a below:
weak patriarchal leadership. All evidence bearing on Ajuereado. Ended well before 6500 b.c. Traces of cave
population density, seasonal nomadism, subsistence, occupations, probably by small groups of a few families.
settlement pattern, and environment during the El People were hunters, trappers, and plant collectors. A
Riego and Coxcatlan phases points to conditions anal- few chipped stone tools, most of them crudely made
ogous to those described above, and it therefore seems from flakes.
probable that during these phases people were orga- El Riego. 6500-5000 b.c. Many
wet- and dry-season
nized into bands, subscribed to rules of band exogamy camps of people who by hunting, trapping, and
lived
and territoriality, and were under weak leadership. collecting. First hints of plant cultivation. A larger num-
Environmental conditions, density of population, ber of types of chipped-stone tools and the first ground-
settlement pattern, subsistence pattern, and other char- stone implements, nets, coiled baskets, twined mats.
acteristics found in our later phases — Abejas, Ajalpan, Ritualistic multiple burials, with suggestions of human
Santa Maria, Palo Blanco, and Venta Salada —have not sacrifice.
as yet been correlated with specific types of social or- Coxcatlan. .5000-3500 b.c. Fewer sites, occupied over
ganization. However, this represents a worthy avenue longer periods by larger groups. People were basically
of investigation and I shall attempt to pursue it in who also trapped and hunted. Finn evi-
plant collectors
Volume VI. The social organization of the final phase, dence of cultivation of a number of plants, including
Venta Salada, could be better learned by close studies com, beans, squash, and chili peppers, and indications
of such sources as codices and relaciones in which of experimentation with fruits. Enlarged inventory of
there are data bearing on the organization of society at chipped- and ground-stone tools; metates begin to re-
the time of the Conquest. place milling stones; improved techniques of basket-
Other archaeological data can aid an attempt to re- making and netting.
construct the social organization. Burial practices give Abejas. 3500-2300 b.c. Significant change in settle-
some hints as to class and caste, as well as the type of ment pattern; some people living, possibly year-round,
family and kinship organization. Figurine types and in pit-house villages along river terraces on the western
23
flanks of the valley. Agricultural foods now supply over stone tools, woven cotton cloth, and possibly twined
20 percent of the diet and include new species. First sandals.
evidence of the domestic dog. New types of chipped- Palo Blanco. 200 b.c.-a.d. 700. Farmers of the valley
and ground-stone artifacts; split-stitch baskets; possibly regularly used irrigation; new domesticates include pea-
cotton thread. nuts, guavas, and turkeys. Numerous settlements, con-
Purron. 2300-1500 b.c. Known only from two cave sisting of villages and hamlets oriented toward, or
occupations, this phase is characterized by distinctive, adjacent to, large hill-top centers with elaborate stone
crude, crumbly potter)' that is among the earliest found pyramids, plazas, ball courts, and other structures were
in Mesoamerica. Other artifact types continue in forms located onmany of the hills or mesas between Calipan
represented by the Abejas and Ajalpan assemblages. and Teotitlan del Camino; others were built on hills
Ajalpan. 1500-900 b.c. People have become subsist- along the northwestern side of the valley near Zapo-
ence farmers living in wattle-and-daub villages on the titlan. These centers suggest that priests and kings or
flood plains. Crops include com, beans, squash, chili priest-kings ruled, assisted by a hierarchy of bureau-
peppers, amaranth, avocados, sapotes, and cotton. Large cratsand consultants. Effigy incense burners indicate
quantities of chipped- and ground-stone tools; sizable that worship of some of the gods later recorded in
samples of well-made pottery in a limited number of Codex Borgia and other codices may date from this
vessel forms. Figurines develop from early, crudely time. Artifacts include finely made tools of obsidian,
made, solid types, unlike any others in highland Mexico, bark cloth, woven fabrics, fine gray and fine orange
to later, hollow or solid large figurines resembling early pottery.
examples from lowland regions (see Vol. III). Venta Salada. a.d. 700-1540. Valley was probably
Santa Maria. 900-200 b.c. Increased number of com- divided into city-states or kingdoms made up of towns
ponents indicates population growth. First structural and hamlets surrounding urban centers, often fortified,
evidence of true irrigation; people were also farmers of situated in the flanks of the valley in the same locations
the flood plain and barrancas, growing improved races as present-day towns. Camps, shrines, villages, salt-
of corn and many other plants. Villages of wattle-and- making sites, quarry and guard-houses and habi-
sites,
daub houses, oriented to a larger village containing tations connected with a wide range of irrigation features
ceremonial structures, were located in valley bottoms were spread throughout the region.
near Ajalpan, Tehuacan, Tepanco, and at the mouth of Economy was based on irrigation agriculture, com-
the Arroyo Lencho Diego, not far from Tilapa. Well- merce, salt-production, processing of cotton, and making
made white and gray pottery shows resemblances to various stone implements. Material culture as a whole
early pottery of Veracruz and Monte Alban. A varied is a variant of the Mixteca-Puebla style and artifact
sample of from other
figurines, including types intrusive complex. Social organization was intricate and stratified.
parts of Mexico, provides inferences of religious and Development of a hieroglyphic system made possible
secular organization involving a pantheon and powerful the documentation of an elaborate religion, a compli-
priests and chiefs. Other artifacts include new kinds of cated calendrical system, and much ceremonialism.
REFERENCES
Barghoorn et al., 1954. Kidder et al, 1946.

Caso, 1938. Kirchhoff, 1952.
Crane and Griffin, 1962. MacNeish, 1954.
Drueker, 1943. Medellin Zenil, 1960.
Ekholin, 1944. Steward, f.
H., 1955.
Evans and Meggers, 1960. Vaillant, 1966.
Friedman and R. L. Smith, 1960.
24
CHAPTER 2
Field and Laboratory Techniques
Richard S. MacNeish
he more general problems dis-

solution of the or which lacked the characteristic desert features.
T
and
cussed in the preceding chapter depended basi-
cally upon information derived from field work
Although the techniques we used
field laboratory.
First,we worked as far west as the Pan-American
Highway, at Huajuapan de Leon in Oaxaca. Next we
drove into the high sierras to the northeast of the
in the field are well known to most archaeologists, a town of Tehuacan, and then made a foray beyond Te-
brief description of them here may be useful to future otitlan del Camino in the southeastern end of the val-
workers. Obviously a number of rather practical ob- ley. Finally we worked northwestward until we were
jectives and problems aided and at times determined just south of the city of Puebla. By this process, we
the formulation of field and laboratory techniques. gradually developed a general concept of the cultural
Although the discovery of Coxcatlan Cave with its and environmental dimensions of the region with which
early corn had brought us to Tehuacan, the initial field we were concerned. This was finally defined as the
problems were, first, to define the area in which we Tehuacan Valley from Tlacotepec de Diaz, on the north-
would work and then to discover the sites within it west, to Tecomavaca on the southeast, a distance of
whose excavation would give the most complete re- about 80 miles. The northeastern limit of our area lay
gional cultural sequence. Obviously the region had to in the rain shadow of the Sierra Madre de Oaxaca on
be dry if we were to find the remains of perishable the southeastern flank of the Sierra de Zongolica. The
agricultural products grown in early times, but the southwestern edge was formed by the foothills of the
region also had to be a cultural unit. Study of maps Mixteca Alta on the southwestern side of the Tehuacan
published by the Papaloapan Commission provided the Valley. To the valley proper we added the nichelike
environmental data that we needed. At first we defined extension on the west formed by the Zapotitlan Valley
the region in terms of an annual rainfall of less than as far as Acatepec.
800 mm. and the consequent xerophitic vegetation. Although we later modified the limits of the area to
We also felt that the region should be in some measure some extent, we found that generally the region lay
a topographic unit. With these three factors in mind, north of Tecomavaca, south of which point the Post-
we turned to the maps again and set the rough bound- classic pottery was and the character
of a different kind
aries of the region in which we would work. At this of the desert vegetation changed, though the rainfall
time we made note of problem areas within the region and topography did not. Northward, Tlacotepec ap-
which shared only one or two of the three character- peared to mark a boundary beyond which new kinds
istics. Then Fred Peterson and I took to the field to of Classic and Formative pottery began to appear,
test our hypothetical environmentally defined region although the valley’s distinctive Postclassic pottery and
and see if the archaeological material correlated with it. the area of low rainfall extend about as far north as
We commenced by working around the Tehuacan Tecamachalco. Moreover, the vegetation noticeably
Valley, gradually eliminating surrounding areas that changed, and the valley was interrupted by a series
either contained cultural materials of a different type of low hills. The east and west sides of the valley were
25
: .
^4
Tlacotepec
Y .
' vs
H ,
• Cacaloapan <>
V
'
\
n A
r"
BTc
' '
’"i
•
- -
^ - -77
<7
7 r l ..w
ÊL RIEGO CAVE

• Tehuacan
K- TECORRAL CAVE '

Yr f
——
. ..
- » " ...
S AsAN MARCOS CAVE, 'f
\ I8°00 -
\ v x
\
-
'Chilac
• Altepexi
> Ajalpan c
. AJALPAN ..
QUACHILCO ,LAS CANO AS

COATEPEC /
Miahuatlan
'.^Callpan
a;
a • Coxcatlan
i ^ ( r ' -*Jk.O >XCATLAN TERRACE

-Y >
abejas cavea4 c° xc at i_a>;_ cave;
w* "
_ _ -
' '
— i
/
HROSCAVE^ •
y
'
Y
y ft
• -
Y>-' / 'v
7
• :
.Teotltlan- *
- 'V. fa"" del Camino-
« Santa Maria lx cat lan'
Quiotepec*
Cuicatlan
Fig. 3. The Tehuacan Valley. Triangles indicate archaeological sites.
26
FIELD AND LABORATORY TECHNIQUES
easy to define by the increased rainfall, wetter oak and came so fast that it was impossible to keep up with the
pine montane vegetation, and mountain-side topog- analysis of them. Further testing of hypotheses about
raphy. To the west of Acatepec even the Classic pottery settlement patterns was more subconscious than con-
was Within the boundaries we set was a cul-
different. scious. Peterson continued the archaeological survey
tural subarea or region in the “Kroeberean” sense of for eighteen consecutive months and eventually cov-
the word, and this is where it seemed advisable to con- ered the region thoroughly.Our sequence in the mean-
centrate our intensive reconnaissance efforts. time was established, not on the basis of analysis or
Having defined our region in a general way, the next survey, but on the basis of the findings of tests at sev-
task was to find the most productive sites in it. We eral stratified sites. The study of settlement patterns
wanted to excavate sites with deep, well-preserved, was accomplished only in 1964, after the materials from
stratified deposits and sites that contained materials the survey had been totally classified. This final study,
typical of each part of the sequence. We wanted to undertaken by James E. Neeley, was, from many points
find the places in our region where the ancient inhabi- of view, more successful than our ideal plan, for by this
tants preferred to live. Using this knowledge of settle- time the sequence had been firmly established, and we
ment or community pattern, we hoped to dig the sites had rather definite ideas about the settlement pattern
which would give the maximum information with the of each phase. Therefore, Neeley had a good opportu-
minimum of time and effort. nity not only to check our hypotheses regarding settle-
Ideally, this aim could have been achieved by choos- ment patterns in toto, but also to check on features we
ing a small part of the region, surveying it intensively, might have missed in the early part of the survey.
analyzing the materials from the survey to establish a We decided at the outset that we would record the
sequence, and then evolving hvpotheses about the set- survey as simply as possible so that our relatively un-
tlement pattern. These hypotheses could then be tested trained assistants would have little trouble cataloguing
and modified by reconnaissance in the various sub- objects found. Each site was given a number, preceded
regions until the entire region was covered. Analysis by an abbreviation keyed to its character. The letter
of all the surveyed materials would then be the basis “T” for Tehuacan project always preceded a small let-
for choosing the best sites for excavation. ter, “r,” “c,” or “s,” which meant, respectively, a ruin or
Needless to say, we did not adhere strictly to this site with architectural features, a cave, or an open site
ideal pattern. First of from the preliminary survey

all, without architectural features. Thus “Tc 9,” for exam-
I made in had roughly formulated a hypothesis
1960 I ple, was a cave site and the ninth site found by the
regarding the kinds of caves that would have stratified Tehuacan project.
deposits. Second, Peterson and I had sufficient knowl- When a site was found, it was given its appropriate
edge of Mexican archaeological chronology to enable designation according to this system, and this site num-
us to set up hypotheses about the sequence of artifacts ber was written at the appropriate spot on one of the
that should be found in the Tehuacan Valley. We did, detailed contour maps of the Papaloapan Commission.
however, survey rather intensively around El Riego, We then entered on a data sheet essential information
just north of Tehuacan, and we did test my hypothesis about the site, such as location, ownership, municipio,
about characteristic features of stratified sites. We also topography, architectural features, dimensions, and so
undertook some digs to test our theories regarding the on. The sheet also had space for a sketch map of the
sequence of the artifacts. As a result of these efforts, site. Peterson made rough maps of some of the sites,
we postulated that caves containing stratified deposits but later Neeley supplemented these with maps made
would be characterized by a wide mouth, a depth with a Brunton compass and tape.
from front to back of three or more meters, a dry ashy Artifacts found on a site were placed in cloth sacks.
floor, vertical walls indicating a workable deposit of When a surface collection was completed, we identified
refuse, and would exhibit on the talus a wide variety of the bags by writing the number of the site in India ink
sherds and chipped-stone artifacts. Also, we discovered on each of two parts of a perforated manila tag. We
that hill tops were capped by large ruins with pottery then placed one part of the tag inside the bag with the
like that of the Classic period of Monte Alban and that specimens and tied the other part outside at the neck.
on the sides of the valley were large ruins with Post- Thus there was little chance that site numbers could
classic Mixteca-Puebla pottery. be lost before the artifacts were washed and the site
With these hypotheses in mind, we settled into our designation written on each specimen. In 1964 all the
region, planning on a long-sustained reconnaissance. artifacts from each site were classified, and the number
However, the survey was so productive and materials of artifacts of each type was recorded on a sheet for
27
each site which we clipped to the original survey sheet. horizontal stripping deepened it until we reached the
In his resurvey, Neeley made some attempt to associate bottom of the deposit, we could move into one wall of
various types of artifacts with various architectural the square and then the opposite one, stripping identi-
features; in special problem cases, the exact location fied strata for a distance of one meter from the vertical
of particular types were noted. Generally speaking, profile. Later, as conditions demanded, we could ex-
however, we did not use random-sample techniques, tend this three-meter trench by the same method until
and all visible artifacts and rim sherds were collected. we reached the end of the deposit. Once the trench was
Nor did we employ dog-leash survey techniques, al- finished, we studied carefully the artifacts from each
though the results in cactus forest and among rocky stratum, and perishable materials recovered, and the
crags might have been interesting. strata themselves to decide whether the cave or site
The greatest weakness of our survey methods was was worth a major effort. Of the thirty-nine sites tested,
lack of time, not the lack of statistical techniques. The —
only eleven Tc 35, Tc 50, Ts 51, Ts 204, Tr 218, Tc
ideal approach, asI have said, would have been to sur- 254, Tc 255, Tc 272, Ts 367, Ts 368, and Tc 307— proved
vey a small segment of the valley, analyze the results, to have deep, well-preserved deposits or to contain
and then continue surveying by sub-region until the materials representing large segments of the regional
valley was fully covered. Analysis of the total results sequence.
of the survey would have determined a tentative se- In the major digging of these sites, procedure and di-
quence and typology, and on the basis of this informa- rection of vertical slicing or stripping varied slightly
tion we would have selected our sites for excavation. according to local circumstances, but the general meth-
However, at the beginning of the project we were not ods of excavation were roughly the same. First, we
sure that we
could support so long a survey period. studied the vertical profile exposed by our test in order
Furthermore, our initial testing seemed to be giving to distinguish the strata so that we might give each a
more reliable sequential data for choosing stratified numerical or alphabetical designation; next we drew
sites than could a seriation study of surface collections. and photographed the profile; then we emphasized the
Regardless of shortcomings, we did discover 454 limits of each stratum by incising dividing lines and
sites within the limits of our region. On the basis of driving into the outlined stratum a spike carrying a
about 100,000 sherds and 5,000 artifacts, we were able tag with the designation number or letter.
to place 350 of the sites in their proper cultural phase. Our next step was to lay out a one-meter grid over
This gave us considerable information about changes the area to be excavated, numbering stakes according
in settlement pattern. Furthermore, we tested thirty- to a relatively simple system of coordinates. A stake
nine sites with potential stratigraphy and found that designated “0-0” was placed near the center of the area
about a dozen of these were good enough to warrant that we planned to excavate to mark the intersection of
full-scale excavations. the north-south base line and the east-west base line. At
The principle basic to the preliminary tests was also one-meter intervals lines paralleling each base line
basic to our digging technique: that is, to strip off were designated Nl, N2, N3, and so on; SI, S2, S3, and
actual strata exposed and identified on a cleaned verti- so on; El, E2, E3, and so on; Wl, W2, W3, and so on.
cal profile or profiles that clearly showed a cross-section Stakes set at the intersections bore the designation of
of both cultural and natural stratigraphy. By digging the intersecting coordinates with the N-S ordinate
from a vertical profile, where the stratigraphy could always first — thus the stake at the crossing of North 1
be seen clearly, we could move forward carefully and and East 1 was N1E1; that at North 4 and West 7 was
take off each stratum with relative ease and the great- N4W7. Each square carried the number of the stake
est possible exactness. in its southeast corner.
In the test, then, as well as in the major excavation, With the squares adjacent to the exploratory trench
the first task was to expose a profile. Usually we started staked out and numbered, the profile studied, and the
a trench one-meter wide and from three to six meters zones within it given a letter or number, we were ready
long that cut across what appeared to be the deepest to move into the vertical face or faces of the trench. We
stratified deposits. In caves, we often laid out the trench usually dug alternate squares next to the test trench,
to cut across themiddle of the shelter from the edge of using three-man teams in each square. The first man,
the talus to the inner wall. In open sites, we laid out the skilled digger, stood in the test trench and stripped
the test to get greatest return according to local con- off the top stratum in the square in front of him. His
ditions. tools varied with the composition of the stratum and
After we staked out a one-meter square and, by the kinds of materials being uncovered. Ordinarily, he
28
Fig. 4a. The application of the alternate-square technique: the grid is laid out and excavation begun at Ajalpan.
Fig. 4b. A large pit results from excavating the isolated squares in Coxcatlan Cave.
29
Fig. 4c. The zones have been isolated and identified, so that the strata can be followed without difficulty in Abejas Cave.
30
Fig. 4 d. Even in this deep pit at Coatepec the excavation is under perfect control through tracing of zones and location of
corner posts.
used a trowel, but on occasions he might use a small loose-leaf printed forms we called “square descrip-
pick, a spoon, an ice pick, a grapefruit knife, a paint tions.” As each zone of each square was exca-
level or
brush, a whisk-broom, or even a teasing needle. He vated, a sheet was filled out recording site number,
never used a shovel! The second man was the clean-up square number, date, names of workers, level, zone,
and bucket man. He sometimes cleaned up the exca- depths, and comments about artifacts uncovered and
vated dirt with a shovel, but most of the time he used the characteristics of the zone. As we completed the
a dust pan or flour scoop, putting the loose dirt in a excavation of each level of each square, we prepared
bucket so it could be carried to a quarter-inch mesh a two-part tag, giving the site number, designations of
screen, where the screen man and collected
sifted it square, level, and zone, the date, name of worker, and
any artifacts or bits of plant material missed by the other information. One part of the label went into the
other two workers. Using this system, we carefully re- bag with the specimens and the other part was tied
moved each stratum in the alternate squares to a con- outside. We also made an attempt to separate, wrap in
venient depth of about two meters. In like manner we foil if necessary, and individually tag various kinds of
removed the isolated squares, standing out like but- —

specimens projectile points, corn cobs, scraps of tex-
be seen on all three sides.
tresses so that the strata could tiles, sherds, and so forth —
that went into the labeled
In this way we were able to obtain both north-south bags. The person in charge meanwhile made notes in
and east-west profiles and to continue the system in- the daily diary, indicating possible correlation with
definitely. If we were required to make special exca- zones in adjoining squares, listing artifacts found, and
vations, such as those for burial pits, we might be recording contextual data and the drawings and photo-
forced to depart from this procedure, but such occa- graphs that had been made. There was a special sheet
sions were relatively rare. in the loose-leaf notebooks for a photographic record.
Perhaps even more important than this well-con- Floor plans, drawn on loose-leaf sheets of graph paper,
trolled technique of excavation and its resultant finds were also included.
was the recording of the progress
of artifacts in context When we uncovered burials or features such as pits,
of both excavations and findings. First and foremost, fireplaces, architectural structures, or concentrations
our field notes, maps, and so on were the only record of artifacts, we described them on Burial Record or
of the archaeological evidence we were so carefully Archaeological Feature Record forms. These forms
destroying with our digging. Secondly, we believed were supplemented by further descriptions in the daily
that the best “policy decisions” concerning our excava- diary notes, by drawings, and by photographs.
tion techniques should be based upon analysis of the Among the most important records we kept were
materials from the excavation. Thus, ideally, analysis profile drawings. Profiles were drawn at a scale of 1/100
and excavation should go hand in hand, and our field or 1/50 on millimeter graph paper, one sheet for each
records should not only describe precisely what, where, line on the grid. We also attempted to photograph as
when, and how the material had been exhumed, but many vertical profiles as possible.
should do so in such a manner as to assist and speed the I hope that the following pages, and the other vol-
field analysis. Obviously, with a number of scientists umes of this series, will demonstrate that our records
describing the excavation, a uniform system had to be were carefully kept. The excavation supervisors and
established, otherwise the individuality of notetakers their aides were as busy as the diggers. One difficulty
would have made the task of the laboratory analyzers with our field technique was that it demanded more
much more difficult and slower. We therefore adhered student assistants than were available to us to help
to thesystem described below, not because we felt this keep records and thus leave the excavation supervisor
or any systemwas perfect, but because we wanted both free to deal with more important tactical and analytic
the digging and the analysis to be co-operative and problems. Because we did not have enough student
interstimulating. assistants, the daily diary and square descriptions were
We usually made a contour map of the area to be occasionally a little brief.
excavated shortly after the initial staking and number- Sound archaeologieal techniques and as complete
ing of the grid. At this stage we also took photographs recording of excavations as possible are basic to any
of the site before digging began. We kept a daily diary archaeological study, but it is equally important that
on looseleaf sheets, describing the excavation, the the specimens uncovered be recorded carefully in the
numbering of the grid, the strategy at the beginning of laboratory and prepared for study and analysis. Thus,
excavations, the names of workers, and other details. ideally, field laboratory work should keep pace with
As soon as we began excavations, we started filling out field excavations so that the former may supplement
32
FIELD AND LABORATORY TECHNIQUES
and at times guide the latter. One valid criticism of came. These boxes were stored in sections of shelves
our execution is that we should have had a crew of reserved for the various kinds of artifact. Thus all the
twelve to fifteen workers to prepare the artifacts and boxes containing projectile points occupied one cubicle
carry out the cataloguing, so that analyses could have on our shelves, the boxes of bifaces another, and the
paralleled excavations a little more closely. boxes of unifaces or bones, beans, com, and so on, still
One of the prime duties of the field laboratory is to others. Throughout our analysis the boxes were shifted
catalogue the material found. This work was done for around considerably. Initially we studied stone and
us by a three-man team, two of whom did the washing, other nonceramic artifacts for chronology, and visiting
repairing, and numbering while the third entered the scientists studied their respective kinds of materials.
record in the loose-leaf catalogue. We wrote numbers When these studies were completed, the boxes of
directly on specimens or on a small tag that was at- material involved were put in dead storage and other
tached to the specimen or its container or foil wrapping. boxes replaced them in the various cubicles. Still later,
If came from a surface collection, it was

the artifact when we analyzed sites by floors, the cubicles were
labeled only by the site number. A separate loose-leaf filled with all the materials from a specific floor of a
sheet was made out for each site, listing the various specific site. Later, boxes of ceramics replaced these.
types or kinds of artifact uncovered there. Usually this After we had completed our analysis, we transferred
sheet was clipped to the appropriate survey sheet and the collection, in boxes classified by type and kind of
filed together awaiting further study. specimen, to the Museo Nacional de Antropologia in
Cataloguing of artifacts from an excavated site was a Mexico City.
little more complex. In order to record all essential By employing the laboratory and field techniques
facts we employed a fractional designation with the described above, we found it possible to attack the
site number numerator and the
in the position of the problems of chronology and the interpretation of cul-
denominator consisting of three parts, representing tural contexts through time in the Tehuacan Valley. Of
the square, the level or zone, and the kind of material. course, the methods employed in these studies, dis-
This number was written on the specimen or an at- cussed in the previous chapter, allowed us to consider
tached label and also on a loose-leaf catalogue sheet even wider cultural problems. From any standpoint,
for each site. however, whatever success we may have had in solv-
After the specimens were numbered, there came the ing such problems we owed to our basic field and
question of storage. To keep the catalogued specimens laboratory techniques. The field techniques gave us
easily accessible, each category of material was stored data which were scientifically reliable and our labora-
in a box properly labeled with inclusive catalogue tory techniques organized the data in such a manner
numbers, the kind of material, and the site, zone, and that they could be brought to bear relatively easily up-
location within the zone from which the specimens on the larger problems.
33
CHAPTER 3
The Region and Its People
Douglas S. Byers
ehuacan, one of the important cities of the state Saint Elias Range in Alaska — is the principal local land-
T
Oaxaca.
of Puebla, lies in south-central Mexico, close to
the border between that state and the state of
Its name, variously interpreted to mean
mark of the Axis, but many more, although
dences of vulcanism and of igneous extrusives are to
be found in central Puebla and adjacent Veracruz.
lesser, evi-
“Place of the Gods” or “Place of Stones” (Barrera 1946), An easy passage from the Tehuacan Valley through
has been given to a valley that occupies part of the the sierra is to be found at Puerto del Aire, where the
northwestern portion of a great down-faulted trough highway to Veracruz, about 145 kilometers away as the
or graben, usually known as La Canada Oaxaquena- crow flies, crosses to descend by a series of hair-raising
Poblana, the Oaxaca-Puebla Trough. In common with switchbacks into the Barranca de Acultzingo and the
that of other geological structures of this part of Mid- valley of Rio Blanco. Other crossings, many passable
dle America, the trend of this great structural valley is only on foot or by burro, are to be found to the south-
in general north-northwest to south-southeast. east of the city of Tehuacan. The principal modern
The northeastern wall of the valley is formed by the route, passable by jeep or truck, is from Teotitlan del
Sierra Madre de Oaxaca, a series ofconnected ranges Camino by way of Huautla. A network of roads and
for which there is a multitude of names. It is a south- trails connecting the highland towns and hamlets may
ward extension of structures of faulted and folded well preserve the pre-Columbian highway net. 1
marine elastics and metamorphics, sometimes with in- —
The Sierra de Zongolica a part of the Sierra Madre
clusions of crystalline rocks, that are characteristic of —
de Oaxaca forms the northern and eastern rampart of
the Sierra Madre The
Sierra Madre de Oaxaca
Oriental. the Tehuacan Valley. Confusion surrounds the naming
was formerly considered be a part of the Sierra
to of mountain ranges, for Tamayo shows the Sierra de
Madre Oriental, but in recent years it has been dis- Zongolica curving to the eastward, while the Sierra
tinguished from it for several reasons. For one, the de TIacotepec joins with the Sierra de Huautla to form
Sierra Madre Oriental forms the eastern rampart of the the northeastern side of the Tehuacan Valley (Tamayo
Central Plateau, while the Sierra Madre de Oaxaca 1962, I: map, facing p. 300). Jean Brunet (Chapter 5)
forms the eastern rampart of a series of valleys and
deeply dissected terrain (Tamayo 1962, I: 413). Fur-
1
For a study of the geology of the Tehuacan Valley see
Chapter 5 below.
thermore, the distinction between the Sierra Madre Excellent treatments of the geology and natural history of
Oriental and the Sierra Madre de Oaxaca corresponds Middle America by Maldonado-Koerdell, West, Tamayo and
with the crossing of this structural complex by the West, Collier, Vivo Escoto, and Stevens are to be found in the
Haruibook of Middle American Indians, Volume I (University of
Clarion Fracture Zone and Neovolcanic Axis (Mal-
Texas Press, 1964). As it is, complexities of the natural setting
donado-Koerdell 1964). The latter two features mark of the Tehuacan Valley constitute so small a detail that treat-
the southern limit of the Mesa Central (West 1964). The ment in the Handbook could not be adequate for the purposes
Tehuacan That volume should, by all means, be
great cone of Citlaltepetl, or the Peak of Orizaba at — of the project.
consulted in order to place in proper perspective much of that
5750 meters, the highest mountain south of the Mount which follows.
34
—
THE REGION AND ITS PEOPLE
calls these ranges Sierra de Aeultzingo, Sierra de tricts where water is from
scarce, pulque, prepared
Zongoliea, and Sierra Mazateca. We shall refer to the juice of the agave, is be used in place of
said often to
range bounding the valley on the northeast as the Sierra drinking water. Southeastward from Tlacotepec the
de Zongoliea, the name used in the Annual Reports of valley floor drops gradually but steadily. Water from
the Tehuacan Project (MacNeish 1961fl, 1962). Above the Rio Atoyac Poblano, impounded by the Manuel
the southeastern end of the valley the Sierra Mazateca Avila Camacho Dam to form the Valsequillo Reservoir
reaches an elevation of 2600 meters or more. Together, south of Puebla, irrigates the Tecamachalco Valley;
these mountains catch much of the moisture from winds that which enters the trough south of San Gabriel
blowing off the Gulf of Mexico, and their summits are Tetzoyocan is used in the Tehuacan Valley. Under ir-
clothed in cloud forest. rigation these high valleys yield good crops of wheat,
The southern and western sides of the valley are alfalfa, and other produce.
formed by the Sierra de Zapotitlan and other ranges, Just northwest of Tehuacan a spur of the Sierra
which are outliers of the heavily dissected northern Madre de Oaxaca extends westward into the valley,
slope of the Mesa del Sur, collectively known as the and an extensive travertine terrace, known as the Cerro
Mixteca Alta or Sierra de Mixteca (West 1964: 63). de la Mesa, juts eastward, narrowing the valley floor.
These mountains are not so high as the Sierra Madre That part of the valley with which we are chiefly con-
de Oaxaca, and they do not receive as much moisture. cerned lies to the southeast of the narrowing.
It is likely that before they were denuded by cropping, The trend of the Tehuacan Valley is a source of
burning, or other activities, they supported a forest of annoyance to cartographers, for it conforms to the
oak and pine, such as is now found in the Mixteca Alta. orientation of bordering geologic structures — approx-
Passages through this highland zone follow either imately from northwest to southeast — and does not fall
river valleys or mountain ridges. The Pacific Ocean neatly into a page oriented to parallels. The limits of
lies more than 240 kilometers from Tehuacan, across the valley can be set roughly between 19°00' and 17°30'
o
these mountains. N. Lat., and between 97 00' and 98°00' W. Long. It
At its northwestern end, the Oaxaca-Puebla Trough occupies more than half of the northern part of the
seems to hang like a ramp from the southern lip of the Puebla-Oaxaca Trough, the drainage of which is ef-
Mesa Central. It is a broad, high valley, with rather fected by the system of the Rio Salado, flowing south-
low walls. Indeed, West (1964) refers to it as the Plain eastward from southern Puebla and northern Oaxaca.
of Tehuacan. Through it there is easy access from The southern part of the trough is drained by the
Tehuacan to the southeastern side of the Puebla Basin, system of the Rio Grande or Tomellin, flowing north-
Tlaxcala, and the Valley of Mexico. The ancient state, ward from Oaxaca. Joining near Quiotepec, the two
or Sehorio, of Tlaxcala with its populous cities and form the Rio Santo Domingo, which cuts through the
religious centers occupied the largest part of the basin Sierra Madre de Oaxaca in an easterly direction and
and, in fact, all the modern state of Tlaxcala and por- eventually becomes the Rio Papaloapan at Papaloapan.
tions of Puebla and Veracruz as well. Between it Although these streams and some of their principal
and the Tehuacan Yalley lay the ancient province of tributaries carry water throughout the year, their many
Tepeacac. lesser tributaries, ephemeral or at best intermittent,
The modern highway to Tehuacan leaves the Puebla carry surface flow only during the rainy season.
Basin near Tepeaca (el. 2243 m.). This remnant of The ultimate sources of the Rio Salado are shown on
ancient Tepeacac, once occupied by some 35,000 peo-
,
the Puebla Sheet (19°00 -97°30') of the Map of the
ple, is now a town of 5,172 souls. Entrance to the north- Republic of Mexico to be on the western flanks of
western part of the Puebla-Oaxaca Trough is by a Citlaltepetl (the Peak of Orizaba). However, detailed
rolling upland, sloping gently toward the southeast. maps of the Papaloapan Commission (sheets 4, 11, and
Turning southward from Tecamachalco (el. 2045 m.) 19) show no through-flowing perennial stream from this
—
the ancient Tepemaxalco the road crosses a barranca source, and a traverse of the valley from San Lorenzo
and climbs toward Tlacotepec, for the drainage of this to the highway between Tehuacan and Chapulco failed
part of the trough is away from Tehuacan. From near to disclose either a stream or a streambed in 1963. It
Tlacotepec (el. 1977 m.) the drainage is southward. must therefore be assumed that if such a stream ever
"Drainage” is perhaps not the correct word to use, for existed, its waters have long-since been diverted and
this part of the trough is a broad, high, arid valley in its bed graded over. The Rio Salado thus finds its effec-
which water is scarce. Many plantations of agave cover tive head in the Rio Tehuacan. One branch of this
the otherwise barren hills, and here, as in other dis- stream which flows from deeply incised canyons that
35
Fig. 5. San Antonio Las Canadas. The church and most houses are along the barranca; a few modern houses are on terraces
on the hillside.
drain the Sierra de San Felipe de las Vaderas joins a and others, imports such as coffee and a tew citrus trees,
longer branch that rises just below the crest of the flowers such as poinsettias for the valley markets, and
divide little more than five kilometers south of Acult- other crops. A number of settlements line its course,
zingo. The Rio Tehuacan finds its way west of the chief among them being San Antonio las Canadas. Its
feet of Cerro Colorado de Tehuacan to debouch into broad fan, now intensively cultivated, supports a
the valley near San Diego Chalma. This small river, checkerboard of crop land from which there is an im-
which becomes the Rio Salado, is joined by a number portant cutting of sugar cane. Up this barranca runs
of rather short, steep tributaries rising in the Sierra the road to Santa Catarina and San Bartolome Lagunas,
Madre de Oaxaca to the northeast, and by fewer but and ultimately to the high country, to cross it to the low
larger and longer affluents that drain the deeply dis- country of Veracruz.
sected Mixteca Alta to the southwest. To the southeast of the Barranca de San Antonio is a
First of the eastern tributaries is the Barranca de mountainous area cut by deep narrow canyons that are
San Antonio which finds its head not far from the drained by the Rio Comulco. Its headwaters lie in
head of the Rio Tehuacan but flows on the eastern side steep canyons where there is no opportunity for villages
of Cerro Colorado de Tehuacan. It follows a rather or crop land. Several short, very steep tributaries enter
deep, narrow valley bordered by narrow shelves of from the flanks of Cerro Paredones. It joins the Salado
what might, out of courtesy, be called flood plains. west of Calipan.
Their width varies from a few meters to as much as ten Barranca de los Mangos enters the valley near
or twenty. Almost every available bit of this shelf is Calipan. It has its source near the very spine of the
planted in native trees, including the local avocado Sierra, beyond the hamlet of Tecoltepec. Extensive
36
Fig. 6. Cactus and thorn scrub near mouth of Arroyo Lencho Diego. The candelabra-like “cardon” (Lemaireocereus
weberi) isin fruit; the angularly branching form is Escontria chiotilla; the smaller form with rounded pads is prickly pear
( Opuntia sp.)
stands of sugar cane and the fact that a sugar mill was channel, joins the Barranca de Soyolapa about two kil-
built there give evidence of thewater that can be taken ometers below Coxcatlan. South of this is a series of
from it. This cannot with safety be judged for water is arroyos, some shorter and some longer, that drain the
now brought from great distances down the valley, western slopes of Cerro Ellotzihuatl and Cerro Chi-
and some that is used below Calipan is said to have chiltepec. Largest of these is the Arroyo Coxcatlan, on
come from near San Sebastian Zinacatepec. The section whose unnamed southern lie Coxcatlan Cave and
fork
dealing with irrigation, to appear in a later volume, Coxcatlan Terrace, beneath the northwest face of Cerro
will go more fully into this aspect of the present-day Agujereado.
culture of the valley. The next large arroyo is Arroyo Lencho Diego, on
Coxcatlan stands at the mouth of the Barranca de whose western side are Purron and Abejas caves. This
Soyolapa, which is the source of water for domestic arroyo, which rises from the southern and southeast-
use and irrigation of the groves and gardens in the ern flanks of Cerro Chichiltepec, must carry a sizable
town. Above the town, terraces, plazas, and pyramids volume of water during the rains. Remains of an an-
climbing the spur on the south side of the barranca cient dam for the impoundment and diversion of irri-
mark the site of the ceremonial center of an ancient gation water can be found among the thorny scrub near
Coxcatlan. It is said by some in Coxcatlan that “the its lower end.
ancients” brought water to the town from the
first Other tributaries from the east are the rios de Tilapa,
barranca, but whether it was brought bv those who re- Teotitlan, and San Martin, all of which rise on the
modeled the mountain spur would be difficult to say. western slopes of the sierra and bring abundant water
The Barranca de Aterango, a shorter and steeper during a large part of the year.
37
In order to say authoritatively just which of the discussion of irrigation that will appear in a subse-
above-named arroyos and barrancas would flow into quent volume.
the valley as living streams had it not been for the Although the valleys of these two streams do not
interference of man, it would be necessary to examine now support a heavy population, they and the Rio San
each one for evidence of such interference. Certainly Pedro, a tributary of the Rio Grande, give access to the
many carry an underground flow of considerable vol- country of the Mixtecs, with which cultural connections
ume. The people of the barrancas employ an ingenious appear to have been at times quite close.
method of diverting this flow without the use of dams, The principal modem settlement in the valley is Te-
which can be washed away or silted up. They first huacan, a busy, modern city of 31,734 inhabitants (1960
dig a ditch diagonally across the channel at a point census, Tamayo 1962: III, 417). It lies at an elevation
where there is an opportunity to trap the underflow. of 1676 meters and is appreciably cooler and fresher
The upstream end of such ditches may reach a depth than the towns below. Ruins of Old Tehuacan are
of two meters or more. Filled with stones, coarse gravel, scattered along the east bank of the river, over two
and sand from the river bed, the ditch then serves si- kilometers away, for a distance of two or three kil-
multaneously as a diversion work and filter, from the ometers. The old town, abandoned in the late sixteenth
lower end of which a stream of clear water flows into and early seventeenth centuries, had been occupied
the aqueduct that carries it to town, orchard, and gar- since early Postclassic times.
den plot. Such diversion works are said to be immune Tehuacan is both the distribution point for the sur-
from damage by all but the most severely eroding floods. rounding area and the assembly point for agricultural
The western tributaries of the Salado are bigger and produce and manufactured goods. The latter include
very much longer than those on the eastern side. Near- cordage, baskets, and palm-leaf hats, some of which
est to Tehuacan is the Rio Zapotitlan, which enters the are brought from Santa Maria Ixcatlan (Cook 1958).
valley near the hamlet of Xaco, just west of San Gabriel Bottling plants, the principal industry of Tehuacan, use
Chilac. It is immediately joined by the Arroyo de Atex- the flow from several springs to supply much of the
cala, an ephemeral stream that drains an extensive soft drinks and bottled w'ater sold in Mexico.
valley just beyond the range of hills that fronts on Flocks of sheep and goats have overgrazed the sur-
Tehuacan and there forms the western wall of the val- rounding country and contributed greatly to forces
ley. Flow in the Zapotitlan is not great, for its drainage eroding the flanks and floor of the valley. The annual
lies almost entirely within the most arid part of the kill is said to produce approximately 125,000 carcasses.
area with which we are concerned. After it reaches the Many small producers together supply a great quan-
valley, its course almost parallels that of the Salado tity of agricultural produce, from and vege-
flow'ers
until it reaches Axusco, where it turns eastward to join tables to grains, including corn, wheat, and barley. The
the Salado near Tlachica. valley is among the grain-producing centers of Mexico
Next, to the southeastward, is the Rio Hondo, or (Tamayo 1962: IV, 296). Smith (1965a), being based on
Calapilla. Its source is in the Mixteca Alta from which it data gathered during the summer, does not reflect agri-
flows first in a generally northerly direction, and then cultural activities of the winter season w'hen wheat and
makes a great bend, first to the northeast and east, and barley are grown. Much of the small grain is consumed
then to the southeast, ultimately joining the Salado locally, but some is shipped out. Several alfalfa-drying
below Ignacio Mejia, about ten kilometers south-south- plants have been established recently, and while their
west of Teotitlan del Camino. Its course and those of its finished product is in part fed locally, some finds its
feeder streams follow' deep, narrow valleys with sides way out of the valley. Three large feed stores supply
that verge on the precipitous and offer scant opportu- grain and allied products to the valley. The value of
nity for farming. manure and fertilizer is now appreciated, and the re-
Next to the Hondo, and inside the curve of its course, sults of the use of fertilizer are clearly evident where
is the Rio Xiquila. This rises near Navitas and Coixt- the farmer can afford it. A new and modern dairy sup-
lahuaca, in Oaxaca, and also flow's northward, east- plies fresh milk to Tehuacan and the adjacent region.
ward, and southeastward to join the Salado about Several large poultry plants that have been built with-
three kilometers below its junction with the Hondo. in recent years supply eggs and poultry to Tehuacan
Like the Hondo, it flows through narrow, deep valleys. and markets in other cities.
Although it offers little opportunity for farming, its Although the city of Tehuacan is the center of popu-
waters fed an ancient aqueduct that is described in the lation, there are villages throughout the valley and on
38
Fig. 7. Arroyo de Atexcala. This broad dry arroyo joins the Rio Zapotitlan just above Chilac. The hills on its far side are part
of the Cuesta de San Marcos in which Tecorral Canyon lies.
the uplands that surround it. Clinging to the mountain- based culture, brought from Spain, has fitted into the
side in the oak-forest zone, with houses nestled into valley and given it its prevailing flavor, but aboriginal
excavated foundations, like so many marmot dens, are overtones still remain. Valley towns may be occupied
villages like Apala, which look down on Coxcatlan by people of one aboriginal linguistic stock, or by those
and Calipan The highland people farm care-
(Fig. 8). of several stocks. Thus, to list only a few valley towns,
fully contoured and come down to the valley to
plots, according to Tamayo (1962: III, 465) Tlacotepec,
work and to trade upland products and fruits for manu- Tepanco, San Gabriel Chilac, and Zapotitlan Salinas
factured goods and produce of the tropics. This rela- are occupied by Popolocas de Puebla; Mixteca occupy
tionship has persisted for many years, probably from a part of Tehuacan; and Mazatec-speakers occupy the
pre-Columbian time. In at least some instances high- southeastern end of the valley. Their principal towns
land towns are subsidiaries of valley towns, and once are Teotitlan del Camino, Cuicatlan, and Huautla de
supplied products to them for inclusion in tribute (see Jimenez — the last-named in the highlands northeast
Cook 1958). of Teotitlan. According to Cook de Leonard (1953: 425)
The valley is like a gigantic mixing bowl of lan- the people of Coapan are Nahua-speakers. Small en-
guages, primarily of the Macro-Mixteca superstock claves of Chinanteca and Cuicateca are to be found
(Swadesh 1959), into which a stirring paddle has been in Cuicatlan (Tamayo 1962: III, 454-75). A remnant
inserted. It has been a principal artery of trade be- of Ixcateca-speakers lived at Santa Maria Ixcatlan until
tween the high central plateau and the Valley of very recently, but there is no information to show
Oaxaca, used by Indian, Spaniard, and Mexican since whether or not this settlement has vanished within the
time immemorial. The lingua franca of the traders has last few years like others in its vicinity (Cook 1958).
been based on Nahuat or Spanish. Mediterranean- Opportunities for the easy exchange of ideas, manu-
39
Fig. 8. Houses of Apala look down on the Tehuacan Valley.
Fig. 9. Buildings with adobe walls and thatched roofs saddled by maguey boards on northern outskirts of Tehuacan.
The right-hand building has a vestibule of cane; sloping thatch fills the end of the gable.
40
THE REGION' AND ITS PEOPLE
factured goods, produce, and other elements of cul- perhaps numbering twelve to twenty-four people in all,
ture exist in weekly markets held in most of the larger who eked out an existence by collecting wild plant
towns. Such opportunities must have existed in ab- foods and trapping or killing any living thing within
original times, providing agencies through which many reach (MacXeish 1962: 31). This population grew, but
archaeological traits achieved wide diffusion. largely retained its genetic identity ( see below, Chapter
The complexion of larger towns is uniformly Medi- 6), in spite of the addition of cultural traits received
terranean, and in them exterior walls of adobe or con- through contacts with other groups of people. It ulti-
crete blocks and roofs of tile or sheet metal prevail. mately reached a figure estimated by MacXeish to be
Aboriginal patterns of house construction still persist, 5,000 times the original population, or 60,000 to 120,000
however, and examples are to be found on the out- people just prior to the Conquest (MacXeish 1962: 41).
skirts of towns or in the occasional house set off by Many of the towns existed long before the Conquest.
itself. Houses may be walled with adobe, cane, corn- All are not listed in the manv works of S. F. Cook in col-
stalks, thatch, or, in the warmer parts of the valley, laboration with L. B. Simpson and Woodrow Borah, a
with palm leaves. Roofs are usually thatched, often circumstance that can undoubtedly be attributed to the
with a saddle of maguey boards (flattened maguey material on which these scholars have based their re-
leaves). Palm-leaf thatch is the accepted roofing mate- search —tribute rolls of the empire of Montezuma and
rial in warmer parts of the valley. Differences in stvles the Spanish revisions of these rolls.
of house construction may be correlated with the cul- Most Tehuacan Valley fell under the Senorio
of the
tural heritage of the linguistic group of the house- de Teotitlan del Camino, an independent state allied to
holder, or with his -economic status. The relative ease the Aztec empire, but not tributary to the Culhua
with which various kinds of building material can be Mexica. For this reason, no doubt, the extremely sig-
obtained also exerts some control over the choice nificant works of Cook and his colleagues overlook such
(see Cook 1936-39). present-day populous towns as Altepexi, Chilac, Mia-
During the recent past the population of the Tehua- huatlan, and Zinacatepec, almost all of which are most
can Valley has undergone many vicissitudes. The orig- certainly old towns, founded in Postclassic times, if not,
inal population consisted of scattered bands or groups. in some cases, before.
1548 1565 1568 1580 1595

Axochitlan
(Axusco) 4,218
Chapulco 752 565 983
Coxcatlan 2,538 1,870 1,472 1,298
Cuicatlan 1,043 1,020
Ixcatlan 726 564
Quiotepec 990 1,078 891
Tecamachalco 47,825 48,423 17,688 19,750 14,400
Tecomavaca 594 586 413
Tehuacan 6,650 12,627 i ,
188 2,288
Teotitlan
del Camino 4,158 4,009 2,798
Tepanco 10,406
Tlacotepec 10,862
Zapotitlan 4,445 8,400 6,056 4,945
TOTAL 69,721
Add 1565 figures
not represented
in 1548 source 25,480
Possible total in
immediate post-
Conquest times 95,201
Sources: 1548, Borah and Cook I960; 1565, Cook and Simp 1948; other years. Cook and Borah 1960.
41
population was drastic. The population of Mexico at

the time of the Conquest is variously estimated to have
been in the neighborhood of 9,120,000 (C. A. Nieve,

quoted by Tamayo 1962: III, 384) or 11,000,000 (Cook
and Simpson 1948). The estimate of Cook and Simpson
has been generally accepted, but not, at first, without
ridicule from some. As time has passed, more and more
scholars have turned to the opinion that the estimate
of Cook and Simpson is too low, and one has suggested
that their estimate for the Mixteca should be doubled
(Cook and Borah 1960:2n). It is evident that there was
a very dense aboriginal population, in excess of 11,000,-
000. Borah and Cook (1963) have now increased this
Fig. 10. Shelter of cane and thatch in the southeastern end we assume that the population
figure to 25,200,000. If
of the valley. of the Tehuacan Valley had declined by roughly the
same proportions, our estimate of the pre-Conquest
The accompanying list gives towns in the valley ac- population in the valley would greatly exceed Mac-
cording to the various studies, with the populations at Neish’s estimate.
the indicated dates. It is difficult to reconcile some of Dobyns (1966) would increase the estimated aborig-
these figures, but we assume that the area and popula- inal population of Central Mexico from the Borah and
tion covered in one count may not be the same as that Cook figure of 25,200,000 to approximately 30,000,000.
in another count for the same town. For example, note His study emphasizes the extent of depopulation fol-
Tehuacan, with figures varying greatly from list to list, lowing the first smallpox epidemic. If a substantial part
as is the case with Zapotitlan. of the population was wiped out by the epidemic of
In arriving at a total for the valley, we have added 1545, as may well have been the case, it appears that
the towns listed by Cook and Simpson that do not MacNeish’s figures are indeed conservative and per-
appear in other lists, but have not used the higher haps should be doubled.
values for Tehuacan, Zapotitlan, Tecamachalco, and By 1548 the population of Mexico is estimated to
Quiotepec given in their count. have been about 6,300,000 (Borah and Cook 1960, cor-
It would be possible to take the figure arrived at by rected by Cook and Borah 1960, “Correction”). Sixty
this somewhat mixed approach and adjust it by a series years later it is set at 1,375,000 or 1,075,000 (Cook and
of arithmetic manipulations, but the validity of such a Borah 1960). A figure of 4,500,000 in 1799 is given by
procedure would be very questionable. While Cook M. Abad y Queipo (Tamayo 1962: III, 384).
and Simpson were able to reach a figure for the popu- There is a distinct possibility that these differences
lation of Mexico at the time of the Conquest by such may be attributable to differences in the area covered
means, we cannot be sure that the population of the by the estimates, and that some groups are included
Tehuacan Valley underwent identical stresses. It by Abad y Queipo who were excluded by Borah and
seems safe to say that MaeNeish’s figures are very Cook. Space to reconcile these differences is not avail-
likely too conservative. able here. As we have seen, some towns in the Tehua-
Following the Conquest, the population of Mexico can Valley seem to have been omitted from the Borah
suffered a pitifully rapid decline. Impressment of Indi- and Cook, and Cook and Borah figures, but this seem-
ans for forced labor brought death to many, especially ing exclusion may result from their inclusion in the
to those sent to work in the mines. Entire urban com- count for other communities.
munities died of hunger and cold when they fled to Brief descriptions of some of the valley towns are
mountains and forests to escape the conqueror’s chains. given by Antonio de Alcedo (1786). Coxcatlan is de-
Diseases brought by Europeans and by African slaves scribed as inhabited by 180 families of Indians, 60 of
further cut the ranks of the survivors. During the epi- Spaniards, mestizos, and mulattoes. Cuicatlan is de-
demic of 1576-78, for example, 60,000 persons died in scribed as the home of 125 Cuicatec Indians who make
the single city of Tepeaca and its suburbs (Cook 1949: much and cultivate much corn, beans, and
saltpeter
26). Authorities in Spain took steps to prevent fur- cotton, from which they make excellent kerchiefs which
ther deaths of potential producers of wealth, but before are their chief item in trade. Of Santa Maria Ixcatlan,
corrective measures could be applied, the decline in Alcedo writes that including its suburbs, it comprised
42
500 families of Indians who concern themselves with

cultivation and trade in grain and vanilla.
Quiotepec . . . has 42 families of Zapotec Indians who

trade in fruits and grains.
Tecamachalco . . . situated on the skirt of a hill so abound-
ing with water that it runs in the streets and is brought into
the houses and with it they cultivate manv irrigated pieces
and gardens that produce infinite flowers, fruits, vegetables,
garden stuff, and grain that make the countryside very fertile
and pleasant, the population includes 122 families of Span-
iards, 17 of mestizos, 30 of mulattoes and 245 of Indians.
Tehuacan de las Granadas, so called for the abundance of
exquisite pomegranates (granadas) which it produces, has
abundant saline springs which are the center of much com- Fig. 11. A fossil irrigation ditch near El Riego was part of
merce, as are the fruits and seeds which the people of Te- an ancient system originating near San Lorenzo.
huacan produce and the fishery which they carry on in the
streams which make the town fertile The waters that . . .
enrich and irrigate the fields are gentle and of pleasing taste
the Tehuacan Valley. Alcedo attributes to it the advo-
but they are full of particles of nitrates, and so thev coagu- cation “de Santa Cruz”: “Administrative center of part
late and petrify the soil in the ditches and conduits through of the Alcaldia Mayor of Tepeaca, it lies at the foot of
which they pass so that thev form borders as hard as if thev some stony sterile hills, has a hot dry climate, extremely
were of masonrv, and from time to time thev must be moved scant water supply, for they use the scant rain which
so that they mav run; as a result of this the fields appear to they collect in a great cistern from which they parcel it
be full of foundations like ruins of ancient buildings: this out economically to the inhabitants who are made up of
same nitrate that is coagulating on the fields is a solvent of
18 Spanish families, 27 mestizo or mulattoes, and 162
stones and sands in the bladder thev do a big business . . .
Indians who speak Chocha and Mexican and support

in flour and meal because of the great deal of wheat that
themselves by butchering cattle and sheep and by the
they gather in the haciendas of the district (and in the Valiev
large harvests of grain that they collect.”
of San Pablo alone there are twentv-two) and with this they
supplv the Plaza of Vera Cruz and the adjoining provinces Figures used by Tamayo show that by 1871 the popu-
as far as Havana and Campeche: there live in this city a lation of Mexico had reached a total of 9,097,056. Part
great many families of Spaniards, mestizos, and mulattoes of this increase may be attributable to people of Euro-
and more than 2080 families of Indians excluding more than pean and African blood who came to or were born in
300 of the former and manv more of the latter who live in Mexico. The total stayed close to this level until 1880,
haciendas in the country Villages under its jurisdiction
. . .
when it began a slow but steady increase. In 1895 the
are: San Gabriel Chilac; San Miguel Eloxuhitlan; San Mar-
first general census listed the population of Mexico at
tin Mazapetan; Santa Maria Conjomeapa; San Pedro Teo-
12,632,427. By 1950
had reached 25,791,017. The
it
titlan; San Pedro Clapulco; Coxcatlan; Acatepec; Miahua-
(Tamayo
eighth general census listed .34,923,129 in 1960
tlan; San Pablo Zoquitian.
1962: III, 384). These last two censuses showed Te-
Tepeaca, Province and Alcaldia Mayor ... of vast extent
and generally warm climate although it has places that are huacan increasing from 23,209 to 31,724, San Gabriel
temperate and others where it is cold it abounds with — Chilac increasing from 5,790 to 6,131, and Ajalpan from
sheep from whose wool they make textiles that are the prin- 5,073 to 6,391 (ibid.: 417). The growth of cities like
cipal business; it produces also much wheat, barlev, and Tehuacan can be attributed in part to the movement
other grains, whose harvests are abundant, and no less of of people from rural regions in search of employment,
fruit, flowers, and garden stuff: in the celebrated Valley of but the increase in the population of such towns as
Balzaguillo which is extensive, there are 56 farming ha- Chilac is probably to be ascribed to improved condi-
ciendas the population includes 80 families of Spaniards,
tions affecting public health and to the possibility of
. . .
102 of mestizos, and 481 of Indians. ... In the provincial

growing more food as a result of improvements in
capital there are 26 farming haciendas occupied bv 177
irrigation works and farming practices.
Indian families who there collect abundant harvests of
wheat, barley, and other grains, and of the first (wheat)
The present-day population of the valley is grouped
alone, 1000 fanegas a year. into towns, between which stretch well-kept plots of
cultivated irrigated land or uncultivated semi-desert,
Alcedo lists four towns named Tlacotepec, of which where goats graze among shrubby growth, including
only one corresponds in its description to Tlacotepec in mesquite which is cut on occasion so that they may eat
43
ENVffiOXMEXT AND SUBSISTENCE
visible instrument other than the plow and oxen, so

that water spreads evenly over it without gullving.
Within recent years the practice of manuring the fields
and of using chemical fertilizers has been accepted,
with consequent increase in the productivity of the
land. The cost of fertilizer is a factor that militates
strongly against wider use.
its
To the west of Tehuacan, on the road to Zapotitlan,

are irrigated and well-cultivated fields bevond which
lies the Milage of Santa Maria Coapan, a community of
Indians to whom the populace of Tehuacan offers a

ready market for tortillas and produce of manv kinds.
The road down the Tehuacan Valley to Teotitlan
del Camino drops abruptly over the edge of the ter-
Fig. 12. Oxen earning the plow on the voke, near Penafiel. race on which Tehuacan stands as it heads southeast-
ward to San Diego Chalma and San Pablo Tepetzingo.
Tepetzingo boasts a parking facility for donkey trains
the leaves and voung shoots. The concentration of the that come down from villages© in the barrancas and on
populace into modem towns may be a post-Conquest the sierra to the north and east. Below the Tehuacan
phenomenon. It mav reflect efforts of the Spanish con- terrace, irrigated fields line both sides of the highway.
querors to bring Indians into small groups so that thev The main irrigation ditch flows parallel to the highway
might be more easily controlled, or. on the other hand, for a way, either hidden behind a screen of carrizo
it mav possiblv reflect the drawing together of people Arundo donax L.) or in an open ditch that proMdes
who worked the haciendas whose gutted ruins are to combined roadside water supply and bathing and
be seen throughout the area. Archaeological remains, laundry facilities. Bevond Tepetzingo the valley floor
especially of Postclassic or Yenta Salada time imme- — becomes wider as it enters a long and gradual slope to
diately pre-Conquest —
can be found close to many the edge of the Llano de la Taza (el. 1374 m.). Some
existing towns, and it therefore seems more likelv that ten kilometers from Tehuacan the road to San Gabriel
the settlement pattern is very old, especially in view of Chilac and its dependent hamlet Xaco branches off.
the fact that some towns are mentioned in pre-Co- Again comes a rather abrupt descent to the eastern
lumbian documents. outskirts of the town of Altepexi (el. 1230 m.) and its
As in much of Middle America, men who work the dependent aldeas, the first sizable settlement through
land make long daily trips to their fields and return w hich the highway passes. Altepexi lies 446 meters
home at night. Valiev men take burros, mules, or even below Tehuacan, and its climate is consequently
horses with them when there are loads of fodder or warmer. Near Ajalpan are several good-sized tile and
crops to be brought back, and do not, as a rule, bring brick works, clav for which is mined from the adjacent
such things in with a tump line, on their backs. plain. Walls of the clav pits have exposed many archae-
Men work their irrigated lands according to the time ological remains, and it is in these pits that the sites
assigned them for receiving water from the irrigation known as Ajalpan, Las Canoas, and Coatepec were
The farmer turns
ditches. his land with a simple found. Cut walls of these pits reveal, toward the top.
wooden plow shod with a steel or iron point, using thin bands of mineral soil and seemingly vegetal mate-
a voke of oxen. In place of the voke that rests on the rial. one above the other in alternating sequence. These
neck of the oxen and is held in place bv a bow that rhvthmite-like deposits which have built up the surface
circles the neck, the Tehuacan farmer uses a yoke that of the plain mav well be the result of irrigation with
is lashed to the brow and horns of the animal. Plowing water from a river that had not vet cut the deep bar-
is done under a variety- of conditions. Land is owmed ranca in which the S ala do now flows.
in common bv the ejido, or it mav be privately owned. Remains of irrigation ditches, “fossilized ' by precipi-
While groups of men bring their oxen and plows to a tation of minerals in the water, seen along the back
given plot at a given time, the circumstances under road between Tehuacan and San Lorenzo, are those to
which the plowing is done vary with the ownership of which Alcedo refers. Remains of other irrigation
the land. Whatever the circumstances, the plowmen are ditches are thickly scattered across the Llano de la
experts. They can turn a field, without benefit of any Taza. Some lead to San Gabriel Chilac. Others wind
44
Fig. 13. The Cuesta de Altepexi. On its surface is the Llano de la Taza.
down the face of the terrace above Altepexi, preserv- A short way below Coxcatlan the valley grows nar-
ing in stone every twist and turn of the watercourse. rower as low hills encroach from both sides. Arable
Following south along the highway from Ajalpan, land is thus restricted to the immediate vicinity of the
one conies next San Sebastian Zinacatepec (el. 1120
to river and to terrace-like margins of tributary streams.
m.). This very “Indian” town is inhabited by Mazatec- In the eastern hills known as Cox-
are the rock shelters
speakers. To the west is San Jose Miahuatlan, close to catlan Cave, Abejas Cave, and Purron Cave. The large
the Rio Zapotitlan, across which is the hamlet of San Postclassic site of Venta Salada lies on a nearly level
Mateo Tlaeoxcalco. terrace-like remnant between the Rio Zapotitlan and
One sees an occasional patch of sugar cane after the Rio Salado, almost directly across from the village
leaving Tepetzingo, but cane only becomes a principal of Coxcatlan.
crop to the south and east of Zinacatepec, for the eleva- At the crossing of the Arroyo Lencho Diego, in which
tion at Zinacatepec i*s sufficiently low to assure warm liePurron and Abejas caves, the elevation of the high-
nights for the cane. A sugar mill operates at the former way has dropped to 850 meters above sea level. The
hacienda of Calipan (el. 1100 m.), which lies above the confluence of this arroyo and the Rio Salado about 1.7
floor of the valley on its northeastern side. Cane for this kilometers to the westward is 50 meters lower. How-
mill grown
is in every suitable location throughout the ever, Cerro Prieto, between the Calapilla and the
southern end of the valley and in other favorable loca- Salado, forms the precipitous right bank of the Salado
tions, such as the lower ends of the principal barrancas. only three kilometers downstream. Hills begin to close
Southeast of Calipan is Coxcatlan (el. 1200 m.), like in from the left bank of the Salado, also. Ignacio Mejia
Calipan set on the northeastern foothills 200 meters lies about nine kilometers downstream from the con-
above the valley floor. Farther down are the two towns fluence and, at an elevation of 742 meters, considerably
of Tilapa (el. 890 m.) and Teotitlan del Camino (el. below it. Teotitlan is 325 meters above Ignacio Mejia
1067 m.), also set on the foothills. Coxcatlan and on the Rio Teotitlan, which empties into the Salado at
Teotitlan are both ancient towns. Their sites are adja- the latter community. Here the river begins to cut
cent to large archaeological sites with pyramidal downward; the elevation of its bed drops 175 meters
mounds, plazas, and terraced ridges. It therefore seems in something over twenty kilometers to Santa Maria
likely that they are modern communities that are con- Tecomavaca. As it passes San Juan los Cues, it has
tinuities of towns dating from at least Postclassic times. cut a gorge that in places approaches a depth of 200
45
Large haciendas formerly occupied the valley, and

may be seen today among fields and towns.
their ruins
Almost without exception they were burned and sacked
during the revolution of 1910-17, when their lands
were expropriated for the restoration of ejidos to vil-
lages from which they had been taken.
On the west side of the Tehuacan Valley are several
groups of saline springs. Most prominent and most
readily accessible are those near the modem towns of
Zapotitlan Salinas and Salinas de Barranca in the Zapo-
At localities in this vicinity salt is pre-
titlan Valley.
pared by evaporating the brine that flows from the
springs and then partially purifying the salt. The salt
Fig. 14. Modern pans for evaporating brine from saline industry dates from pre-Columbian times, as is demon-
springs near Zapotitlan Salinas. strated by the myriad of mounds in the Tehuacan Val-
ley that were built to support filters through which
brine was passed in order to remove impurities. The
meters. Less than ten kilometers farther downstream
filtrate was concentrated by boiling it in pottery vessels
the Salado joins the Rio Grande, at an elevation of
and then heating the concentrated brine until the water
somewhat less that 500 meters, before turning eastward
evaporated and a salt cake resulted. The cake could
to cut through the Sierra Madre de Oaxaca which
then be carried in trade, or sent in tribute to the over-
towers more than 1600 meters above the river on either
lords. Near Cerro Pelon, and to the west of Petlanco,
hand. Cultivated land is not plentiful in the lower
salt mounds innumerable quantity are con-
in almost
reaches of the river, although there is some at Te-
centrated around another group of saline springs.
comavaca. While the valley was surveyed for archae-
The link with the pre-Columbian past is very strong
ological sites as far downstream as Tecomavaca, for
in the valley. Although the inhabitants of the valley
all practical purposes archaeological work was concen-
appear be thoroughly acculturated, there are many
to
tratedbetwen Teotitlan and Tehuacan, or El Riego,
Indians and Indian customs still persist in one form or
which adjoins it.
another. Through this happy circumstance it has been
From Tehuacan to Teotitlan the distance by road is possible to interpret the uses of many artifacts and to
approximately 64 kilometers; in a direct line it is nearly
determine the uses to which many plants were put,
16 kilometers less. The average gradient of the valley
thus bringing life to archaeological research as few are
floor is neighborhood of 19 meters per kilometer,
in the
fortunate enough to be in a position to do.
but this is first 300 meters consist
misleading, as the
of the descent from the Tehuacan terrace to the Llano
de la Taza, and the next 133 meters come in the descent
—
over its edge two rather steep descents, the cuestas
of San Marcos and of Altepexi, separated by a long,
slowly descending stretch. From Altepexi to Teotitlan
the descent is at a more even rate.
The Rio Zapotitlan follows a generally southeast-
ward course toward the Rio Salado almost from the
moment it breaks into the valley above Chilac. Between
Venta Salada and Axusco the river continues on a closely
parallel course, turning abruptly at Axusco to flow
eastward to join the Salado near Tlachica. A narrow
strip of arable land lies along the Zapotitlan; south-
ward from Axusco this becomes wider
for a short way.
There is little water for irrigation here, saving the flow
of the rivers which must be taken from them many Fig. 15. Remains of prehistoric salt pans west of Petlanco.
kilometers upstream in order to raise the water to the The mounds which supported the filtering apparatus have
level of the fields. been opened by treasure hunters.
46
REFERENCES
Alcedo, D. Antonio de, 1786. Dobyns, 1966.
Barrera, 1946. MacNeish, 1961a, 1962.
Borah and S. F. Cook, 1960, 1963. Maldonado-Koerdell, 1964.
Calderon Garcia, 1956. Paddock, 1966a.
Cook, S. F., 1936-39, 1949, 1958. Smith, C. E., 1965a.
Cook, S. F., and Borah, 1960. Swadesh, 1959.
Cook, S. F., and Simpson, 1948. Tamayo, 1962.
Cook de Leonard, 1953. West,' 1964.
CHAPTER 4
Climate and Hydrology
Douglas S. Byers
he Tehuacan Valley enjoys a hot dry climate. circulation is heat received from the sun. This is re-
T Rains occur there chiefly two periods of six

in
to eight weeks each between mid-May and mid-
July and between late August and late September.
ceived in greatest amount in the equatorial regions,
which serve as a heat source. A pattern of general at-
mospheric circulation distributes that heat from the
Winters are ordinarily dry. Tehuacan itself, at an ele- heat source toward the polar regions, which are heat
vation of 1676 meters, experiences frosts during the sinks. The circulation displays salient features resulting
winter months, but days are warm or even hot, except from the rotation of the earth and the position and
when the nolle brings cold, cloudy weather. On occa- configuration of land masses. These include a deep
sion a nolle will bring rain to the valley and snow far zone of easterlies in low latitudes toward which winds
down the shoulders of higher peaks in the neo-volcanic converge from both hemispheres; zones of divergence,
axis. In short, during the winter, woolen clothes are or high pressure, near 30° N. and S. Lat., from which
often necessary for comfort. flow the easterly trade winds of the tropics and the
With the approach of spring, winds increase, dust deep and broad mid-latitude westerlies; and a zone
begins to fill the air, and smoke from burning milpas of polar easterlies, the latter separated from the mid-
adds to the haze. When rains begin, the air is cleaned, latitude westerlies by a poorly defined zone of low
and the oppressive feeling of the late dry season gives pressure at about 60° N. The subtropical highs are
way to a bracing freshness. During the rainy season well developed over the oceans and generally absent
the temperature falls from its dry-season peak, and over land masses, although their effects are felt far
a sweater again becomes comfortable. from the oceans.

This climate is a product of several factors which Oceanic currents also transport heat away from
modify and are modified by each other. We shall now equatorial regions. The pattern of these currents ex-
turn to an examination of them and to the peculiar- hibits many similarities to currents in the atmosphere.
0
ities of the Tehuacan Valley. They also have some effecton climates. Of these, the
Climate is a product of many variables in the atmo- best known to us is the wanning of northwestern Eu-
sphere. These include temperature; humidity in the rope by the Gulf Stream.
form of water vapor, clouds, and precipitation; and The climate of the Tehuacan Valley is a product of
motion arising from differences in pressure and density the physical characteristics of the atmosphere and of
existing between neighboring air masses. the position of the valley with respect to the general
The ultimate source of energy powering atmospheric atmospheric circulation, particularly with respect to
the trade-wind belt of the northern hemisphere. Next
in importance, is its situation, between the Gulf of
0
I am indebted to Reid A. Bryson and David A. Baerreis of
Mexico and the Pacific Ocean. The controlling factor,
the University of Wisconsin for valuable criticisms and advice.
They are in no way responsible for any statements or postula- however, appears to be the topography of south-cen-
tions advanced in this paper. tral Mexico. We shall discuss these subjects in order,
48
CLIMATE AND HYDROLOGY
and then proceed to a discussion of rainfall in the foothills of a mountain range. Air will also rise when
Tehuacan Valley and environs. a mass of cooler and therefore heavier air flows in at
It must be realized from the beginning that discus- low levels, forcing the warmer resident air upward.
sion of the climate of the Tehuacan Valley is based on Under all these circumstances the rising air expands,
data which are inadequate for a thorough study. Al- its pressure is reduced, it usually cools adiabatically,
though series of observations for many stations in and water vapor condenses into droplets forming
Mexico have been collected by the Servicio Meteoro- clouds or falling as rain.
logieo of Mexico since 1921, only a few series beginning Conversely, descending air is compressed and heated
in 1941 are available for the valley and surrounding as its elevation decreases. This may happen to air
parts of Puebla and Oaxaca. These are insufficient for forced over a mountain range and down the other side.
a study of local climate. Therefore we must rely on As grows progressively denser and warmer, its ca-
it
the records of stations established in the Tehuacan pability for retaining water vapor increases and it be-
Valley in 1955 by the Comision del Papaloapan. At comes a drying wind. This explains why a mountain
these stations, data for rainfall, temperature, direction range that is cloud-covered or rain-soaked on its wind-
and velocity of the wind, and evaporation have been ward side may remain clear on its leeward side. A wind
collected, although records in all categories are not that blows down a slope in this way is known as a
kept at all stations. This information is published each chinook in the western United States and a foehn in the
year in the BoJetin HidroJogico. Unfortunately, the Alps. What the ancient inhabitants of Tehuacan called
Boletin covering the year 1961 is out of print and no it was not discovered by the Project.
copy is presently available. That for 1963 was not pub- Where air ascends or descends there is little lateral
lished at the time when this article was prepared. Be- transfer so that calms may prevail in such zones. The
cause the total body of available data is inadequate, equatorial zone of rising air produces the showery equa-
this analysis of climatic details of the Tehuacan Valley torial belt known over the oceans as the Doldrums. The
can be no more than tentative. While details accord zones of subtropical highs, where air descends, are usu-
with the larger picture of south-central Mexico, minute ally characterized by fair, calm weather known as the
variations in rainfall within the valley — possibly of Horse Latitudes.
the utmost importance in explaining the location of The stability of a column of air depends on its tem-
sites and variations in the plant cover — can only be perature gradient. The less rapid the fall of temperature
suggested. A long record of such data for the valley with increase in altitude, the more stable the column
as a whole could be of considerable significance to of air. The temperature gradient of descending air
ecologist and archaeologist alike. masses is not abrupt and therefore such air masses are
Now let us turn to a discussion of climatic factors usually stable. The poleward borders of the trade winds,
affecting theTehuacan Valley, proceeding from the in general, belong in this category. They are dry and
more general to the more particular. usually rainless.
The amount of water vapor that the air around us When there is a steep temperature gradient from
holds varies directly- with the temperature of the air. heated air at the earth’s surface to cooler overlying air,
Relative humidity expresses water vapor as a percent- the air becomes unstable. As the warmed air rises, it
age of the amount the air is capable of holding. With expands and is cooled. Its capacity for holding moisture
close approach of relative humidity to 100 percent, is progressively reduced, so that clouds may form. If
condensation will occur. If the pressure or temperature the process continues, a shower results. The equator-
of an air mass be lowered, it will reduce the ability of ward borders of the trade winds belong in general to
that air mass to retain moisture in the form of vapor. this category. They are often very humid, bringing rain.
As the elevation of any air mass changes, there is This is especially true on eastward-facing land and
usually an inverse change in its density and tempera- mountains.
ture. The change in temperature of dry air amounts to Each year the thermal equator, with its zone of calms
approximately 1° C. for every 100 meters change in and ascending air and its bordering zones of trade
altitude; that of saturated air changes 1° with every winds and subtropical highs swings north and south in
150 meters of elevation. response to the seasons. In June, July, and August the
Air will rise under several circumstances. When air is thermal equator swings north to about 12°, bringing
heated it expands and rises through overlying cooler with it the zone of ascending, unstable air, with its
air as a “thermal.” Air will be made to rise when its flow towering cumulus clouds and rains. At this time, all
forces it upward over a sloping land mass, such as the of Central America and the southern half of Mexico
49
are under the flow of the moisture-laden and increas- Polar air, at the same moves southward on the
time,
ingly less stable portion of the trade winds. During western side of the storm, and it, in turn, may com-
the northern winter the thermal equator swings south- plete a circuit. Such storms can grow to enormous size
ward from this position carrying with it the zone of so that the system spans much of North America east
descending stable air of the subtropical highs. At this of the Rocky Mountains.
time all of Mexico and much of Central America is In winter the zone of cyclonic storms and anticyclonic
under the flow of the more stable portion of the trade highs characteristic of middle latitudes is displaced
winds. This annual swing produces the winter dry sea- southward with the subtropical highs and thermal equa-
son and summer rainy reason of Central America and tor. When a large and vigorous winter storm crosses
much of Mexico. the United States, masses of cold air may flow south-
The trade winds blow across heated ocean water as ward across the Great Plains to the Gulf of Mexico and
they approach the coast of Middle America, taking up produce the nortes which sweep the coast of Tamauli-
moisture from the warm water and carrying it toward pas and Veracruz, reaching into the Caribbean, Central
the land. But it is only in the summer months that this America, and even northern Colombia. They rarely pro-
air carries sufficient moisture to produce heavy rains. duce rain in the Tehuacan Valley, but may do so in Cen-
Nevertheless, some stations on the windward slopes tral Mexico or along the Gulf Coast or on occasion bring
of the Sierra Madre de Oaxaca or on higher elevations snow far down the shoulders of the higher volcanoes.
experience rain during most of the year. Air flow is in They often result in very cool weather in parts of the
general from the land mass of Mexico to the Pacific Tehuacan Valley.
Ocean. As a result, the effects of the Pacific Ocean The class of atmospheric disturbance with the great-
water are felt in only a narrow coastal strip. In summer est effect on the Tehuacan Valley is the tropical cyclone
the wind may be on shore. Along the Pacific coast, or tropical disturbance. Those originating on the At-
winds carry surface water away from the land, result- lantic side of Central America are called hurricanes.
ing in an up-welling of colder water from the depths. Those having their origin off the Pacific coast are lo-
Atmospheric disturbances further complicate the pat- cally known as “chubascos.” Origins, causes, and the
tern that we have just outlined. Air flows clockwise mechanism of these storms are not fully understood.
from the subtropical highs and some of this air of tropi- They appear to have their origin on the poleward side
cal or subtropical origin moves poleward. Similarly, of the equatorial low as a wave disturbance on which
air of polar origin flows outward, and, of necessity, a center of abnormally low pressure develops. Such
equatorward. The zone in which they mingle is a tur- storms may be born in the Atlantic Ocean, the Carib-
bulent one characterized by the passage of cyclonic bean, or the Gulf of Mexico, or off the Pacific coast,
storms and anticyclonic highs. Heavier polar air flows often between the Gulf of Fonseca and the Gulf of
under the warmer tropical air, which rises and usually Tehuantepec. Some never attain the full force of a hur-
precipitates any moisture as rain or snow. Warmer air ricane, but others develop into very powerful storms
is carried northward on the eastern side of the storm with wind velocities of more than 180 kilometers per
and may at times flow completely around the center. hour. Within the Tropics they are usually compara-
tively small symmetrical stonns, with violent winds
rarely extending more than 240 kilometers from the
center, but as they curvenorthward the Atlantic hur-
ricanes increase in size untilwind and rain may ex-
tend more than 400 kilometers from the center. As this
occurs the storm becomes increasingly asymmetrical.
Tracks followed by tropical storms vary from month
to month and are, at best, unpredictable (Fig. 16). In
their early stages the storms usually move westward

with the trade winds at between 8 and 25 kilometers
per hour, gradually curving northward, and eventually
moving off to the northeastward. As the storms turn
northward they pick up speed. Those that approach
the coast of Central America and Mexico often follow
Fig. 16. Hurricane tracks, 1926—1955, after Atlas Climato- northward just off shore, but some turn westward to-
logico, pis. 28, 29. ward land. Those that intersect the land where there
50
97 ° 00
'
'
96 ” 00
“littlesummer.” In this interval crops and plants that
started their growth under stimulus of the early rains
put on a rapid growth. Rains brought by tropical storms
supply much-needed moisture as crops begin to mature
toward the end of the period of rapid growth. In the
absence of a supply of water for irrigation, a change
of track of tropical storms that results in a drought in
August and September may have near fatal effects on
the crops in inland highland valleys like that of Tehua-
can. Thus the tropical storms bring welcome and much-
needed rain to southern Mexico when the beneficial ef-
fects of the June and July rains have begun to wear
off. At the same time, these storms may bring torren-
tial downpours which can do severe damage, destroying
crops and eroding land. At these times, raging streams

and flooded lowlands drown people and livestock alike.
The occurrence of tropical storms is unpredictable, their
courses are erratic, and they cannot be counted on to
reach any one part of the country every year.
Topography exerts strong and often controlling influ-
ence on weather systems approaching the Tehuacan

Valley. In the first place, the Sierra Madre de Oaxaca
provides a barrier to the passage of storms from the
eastward. Its peaks reach altitudes of 2600 meters or
more, and the intervening ridges reach 2000. It casts
a rain shadow Tehuacan
across the entire more Valley,
intense at the northwestern end, where less rain can
be expected, than at the southeastern end. Most of the
Fig. 17. Mean annual precipitation, 1926—1955, showing moisture brought from the Gulf of Mexico and the Car-
greater amounts of rainfall in eastern slopes of the Sierra ibbean by storms and trade winds drops on the eastern
Madre. slopes of the Sierra, at elevations between 1000 and
2000 meters. In three localities shown in Fig. 17, taken
are no high mountains, as at the Isthmus of Tehuan- from Atlas Climatologico e Hidrologico (pi. 10), the
tepec, may cross to the Pacific and follow northward average annual rainfall, 1926-55, was over 4000 mm. In
along the Pacific track, off the coast of Oaxaca, Gue- the exceptional year of 1952, more than 7000 mm. fell
rrero, Michoacan, and Jalisco. at Cataluna. Rapid decrease in the amount of rainfall
The quadrant of the storm that lies to the right of occurs between the eastern slopes and the crest of the
the direction of travel is the more dangerous. For ex- Sierra, where the average recorded is no more than
ample, as a storm turns northward its radius increases 1000 mm. Some moisture from the Gulf crosses the
in the eastern quadrant until gales and rain may extend crest, but rainfall is progressively less at stations down
a great distance from the center. As it travels northward the slope and in the Tehuacan Valley. During the same
with increasing speed, wind velocities increase on its period, 1926-55, annual precipitation at Coxcatlan aver-
its forward speed. On the
eastern side to compensate for aged 833 mm., that at Teotitlan del Camino averaged
western side of the northward-moving storm, wind ve- 546 mm., and that over a large area in the northwestern
locityis less, precipitation is less, and rain and gales end of the valley was less than 500 mm. The peak of
do not extend westward or southwestward as far as aridity must lie close to the area around Tehuacan and
they do to the northeast. Zapotitlan, where annual rainfall in these years aver-
Seasonality of tropical storms is marked. Although aged 440 mm. and 412 mm. respectively. In the excep-
they may occur in almost any month of the year, their tional year 1943 only 78 mm. of rain was recorded on
frequency is greatest in late summer and early autumn. the gauge at Zapotitlan. Furthermore, rainfall within
The period from late July through the first part of the general area of the valley is often highly localized.
August is known in Mesoamerica as the veranillo, or In the nearby Mixteca Alta there is more rain than
51
Fig. 18. Rain in the Tehuacan Valley may be highly localized.
in the Tehuacan Valley, but in only two rather restricted Altepexi during the years 1955-60 and 1962. We were
areas within the territory covered by our detailed map unable to obtain data for 1961. Altepexi lies approxi-
(Fig. 19) does annual rainfall equal or exceed 1000 mm. mately in the center of the valley and cannot be greatly
Records of rainfall show a marked cyclical variation affected by mountains on either side. In the process of
in precipitation, but parallel curves are not recorded averaging weekly totals, extremes of rainfall have been
at the six stations for which curves are given on plate smoothed out, as for example, with rainfall of 280 mm.
10 of the Atlas Climatologico. It thus appears probable from one stonn, part of which occurred in one week,
that these differences reflect different origins for the with the balance in the next. If rain did not occur in
storms that bring rain to southern Puebla and Oaxaca. the same two weeks in any of the six other years under
We shall return to this matter in a discussion of rain- review, this very substantial rain would be almost ob-
fall below. scured. Thus this chart is not a valid chart of average
The detailed map of rainfall within the valley is based rainfall, but of average rainfall over a very’ short period
on records of stations established there in 1955. Stations within what seems to be a dry cycle. Two peaks of
with shorter records have been used as guides in draw- rainfall are characteristic of the rainy season through-
ing isohyets; the yearly map of rainfall in the Boletin out this part of Mexico. Diagrams for other towns in
has also been consulted. Discrepancies between the de- the valley are not presented here since they vary only
tailed map and the more general map are at once ap- in the total amount of rain, not in the distribution of
parent, but they can be explained on the basis of dif- rain through the year.
ferences in periods covered and in detail registered on During the dry season clouds hang on the Sierra
the more general map. Madre de Oaxaca. Cloud forest thrives on the higher
To illustrate the division of the rainy season into parts of the ranges. However, the trade winds, now
two parts. Fig. 20 shows average weekly rainfall at descending the slopes, are warmed adiabatically so
52
Q
• Tecamachalco
•Xochitlan -
Tlacotepec
Cacaloapan
8°00
Coyotepec
p
Âoultzmgo
EL RIEGOmm
• Tehuacan 'W# Lagunas
TECORRAL CAVE
<4001 ^SAN MARCOS CAV
'âpotitlan Salmas
\ 18*00 'Acatepec 'Vi* San Antonio;
Salinas de Barranca / /
Canadas
Acatitlan Altepexi #
Caltepec • Ajalpan
QUACHILCO \ AJALPAN
^LAS CANOAS
1^ COATEPEC
0
ehuipa^®
j'T ^
iljt* Coxcatlan
I I /
iZoquitlan
"IT, San Miguel Astatla 1
k COXCATLAN'
ÔXCATLANÂ I I
•
-r
'
; ABEJAS CAVE?;
.
•
.* PURRON CAVE J!
Tlacotepec
I.Sai Antonio Abad / / de Diaz
Mahuizapa
#T
,
k
>
< 600
"
\^
V\>
M .
:
\ V* Ignacio
M' j:a /
del
/
Camino
J
'
/
~ d *TepeIm eme |
17^, |
0 Magdalena Jicotlan
ISIS Xiquila*$tof~^ - I r I
/' «Teopoxco
Cataluna
• Chilchotla
>4000
. - /
f
IgSa-*"
w ~- : -
I
iHuautla-5,
Coixtlahuaca
•Jocotipac
Cuicatlan
Fig. 19. Average annual rainfall in nun., Tehuacan Valley and environs, 1955-1962.
53
TEHUACAN
..
3 0
IV
20
J 1i I
.u
JAN
Fig. 20.
FEB MAR APR
II
jdlMAY
1
Average weekly
J UN JCL AUG
m
SEPT
rainfall at Altepexi, in
OCT NOV DEC
mm., 1955-
1962.
that they become desiccating winds that bring the dust

storms of March and April and make the heat in the
valley oppressive and close.
Furthermore, the direction of what was once the
northeast trade wind has been changed by topographic
features to conform to the trends of the valley. The
prevailing wind at El Riego and Teeamaehalco shown
in plate 6 of the Atlas Climatologico is southeasterly,
and blowing at a force of 4.5-9.0 mph. At Coxcatlan
it is easterly, at the same velocity'. At Teotitlan del
Camino the prevailing wind is westerly and at only Fig. 21.Monthly distribution of maximum precipitation,
1926-1955 (Atlas Climatologico, pi. 11).
2.25—1.5 mph., possibly in response to the proximity
of the mountain wall to the west.
Further study of rainfall analyzes what is termed evapotranspiration shows the great deficiency of rain-
“useful rainfall.” Rains of less than 5 mm. are of little fall in the Oaxaca-Puebla Trough, for evapotranspira-
use to crops since they do more than dampen the
little tion everywhere exceeds rainfall by a substantial
surface of the ground and evaporate quickly. Of rains amount except during a brief period near the first of
of more than 5 mm. only 75 percent is considered to be October.
available for crops, while 25 percent of the water runs According Koeppen classification, the south-
to the
away over the surface of the ground. Furthermore, this eastern end of the Tehuacan Valley belongs in the cate-
rainfall must occur at times when it can be used by gory Rswh'-steppe climate: rains during summer,
growing crops. If the rain falls when crops are not in, xerophitic vegetation, hot with mean annual tempera-
a volume sufficient to supply the needs of crops will ture of more than 18° C., and mean of coldest month
be of little use if there is no way of storing it. On the above 18°. To the northwest of the confluence of the
basis of usefulness, rainfall in the Tehuacan Valley is Rio Zapotitlan and the Rio Salado it falls in the cate-
found to be even more deficient, for the farmer can gory Bswh: mean of the coldest month below 18° C.
count on an 80 percent probability of useful rain in A small area within the category' Aw (savannah climate,
only the following amounts: Tehuacan (El Riego), with periodic rains and a dry winter) is to be found near
119 mm.; Coxcatlan, 168 mm.; Teotitlan, 132 mm.; Coxcatlan. According to Thornwaite, the climate of the
Tlacotepec, 175 mm.; and Zapotitlan, 107 mm. At valley floor falls into the category DdB' 3 (arid, rain-
Tehuacan, there is an 80 percent probability that use- fall deficient in all seasons, temperate, evapotranspir-
ful rainfall of 24 mm. will fall during May, June, July, ation high), while tire flanks fall into CidB' 4 (semi-arid,
and September; in August and October approximately rainfall deficient in all seasons, temperate, evapo-
10 mm. may' be expected, while in April somewhat less transpiration very high). Only the extreme southern
than 10 mm. may fall. There is a 20 percent probability end of the valley falls within the category DdA': arid,
that droughts of twenty-eight or more days’ duration rainfall deficient in all seasons, hot. These subdivisions
may be expected in seven out of twelve months at are of significance in measuring the degree of aridity.
Coxcatlan, and in ten at Tehuacan. The index of Around Coxcatlan and including the areas in which
54
—
Coxcatlan, Purron, and Abejas caves are situated, the matologico presents analyses of tropical storms over a
Koeppen classification indicates a somewhat less in- thirty-year period ending in 1955. Unfortunately, few
tense degree of aridity. meteorological stations were in operation in the Tehua-
Mean annual temperatures within the Tehuacan can Valley at that time; therefore data for the valley
Valley vary inversely with the elevation of the stations. itself are lacking. The Papaloapan Commission is quite
At Tehuacan (El Riego), el. 1676 m., mean annual tem- properly concerned with the entire watershed, and not
perature is given as 18.9°C.; monthly means vary be- primarily with the arid Tehuacan Valley. In the Papalo-
tween 15.8° in January and 22.0° in April and May. apan watershed, the greatest flow is from rivers on the
The mean for July is 20.4.° The maximum range is be- eastern side of the Sierra Madre de Oaxaca. A storm
tween 13.0° and 25.0°. At Chilac (el. 1219 m.) mean may affect those rivers, yet cause slight rise in the
annual temperature is 21.5°, with a January mean of Salado and its tributaries. The converse is also true.
17.0°, and a July mean of 22.3°. Means at Coxcatlan Detailed records maintained from 1960 in the Tehuacan
(el. 1200 m.) are 21.5° annual, 19.1° for January, 22.2° Valley offer a hint that Atlantic storms do not deserve
for July. The mean at Teotitlan (el. 1067 m.) is 23.8°, as much credit for bringing rainfall to the valley as the
with January mean of 21.0°, July and a mean of 24.3°, Pacific storms do. It would take far longer records to
range between 19.0° in January and 28.0° in May. Just demonstrate this, but in the records available there is
below the confluence of the Rio Zapotitlan and Rio a suggestion that this may indeed be the case.
Salado, mean annual temperature is more than 25° C. We shall begin by considering the discussion of trop-
Examination of maxima and minima shown in Table ical storms in the Atlas Climatologico, and then pro-
I will demonstrate that there are departures from this ceed to the years 1955-60 and 1962. Although detailed
generalization, and that these are most pronounced in data on storm tracks in the latter series is not presently
the case of minimum temperatures. Freezing tempera- available, we seem to have sufficient data on which to
tures occur commonly in the higher, northwestern end formulate an hypothesis.
of the valley in the months of December and January. Tropical storms originating in the Lesser Antilles,
These set a firm limit on the diffusion of tender peren- the Caribbean, the Gulf of Mexico and off the Pacific
nials and trees. They also affect the planting of corn, Coast have some on the Tehuacan Valley. In the
effect
for it cannot be grown safely at or above Tehuacan Atlas Climatologico are courses of seventeen such
in these months, although it thrives just a short way storms which occurred between 1926 and 1955 and
down the valley, at lower elevations. affected the watershed of the Papaloapan. Some of
Minimum temperatures result from a number of fac- these are full-blown hurricanes, others were storms that
tors. Temperature of the which blankets the region
air had not reached hurricane force, while still others were
is but one. Radiational cooling and air drainage are of large areas of low barometric pressure in the Gulf of
great importance, particularly under comparatively Mexico that resulted in extensive movements of air
cloudless conditions and in mountainous country. masses producing prolonged rainfall over a wide area in
Doubtless these factors combined to produce the the Papaloapan watershed.
record —12.5° C. at Tepelmeme in 1961. Similarly, ex- Tropical storms that reach the Papaloapan basin
tremely low readings may occur in the mountain towns not necessarily the Tehuacan Valley most frequently—
of the Mixteca Alta, apparentlyunder favorable local originate in theCaribbean, the Lesser Antilles, or in
conditions. Records of temperature in much greater the Atlantic beyond them. Almost all are powerful
detail than those available would be required for a storms of great force. Hurricanes Hilda and Janet, of
proper study of this aspect of climate. Suffice it to say September 1955, which affected the area around Pa-
that Tepelmene experienced freezing temperatures in nuco, and an unnamed hurricane of 1941 which came
five of the twelve months in 1962, and only one-half ashore in the state of Veracruz are typical of storms
a degree above frost in two more months. Comparably originating in the Lesser Antilles. Storms originating
low temperatures are not recorded for stations in the in the Caribbean Sea follow a similar pattern. Such a
Sierra Madre de Oaxaca —
doubtless a reflection of the hurricane, in September 1944, brought severe floods to
warmth and moisture carried in from the Gulf of Mex- the Papaloapan watershed. Tropical storms originating
ico by the trade winds. off the coasts of Campeche and Tabasco sometimes
A detailed study of rainfall within the Tehuacan dissipate quickly; others grow into full-blown hurri-
Valley necessitates the examination of records of trop- canes, like hurricane Gladys of early September 1955,
ical storms, which, as we have seen, may be vital to which produced severe flooding in affected watersheds.
the success of unirrigated agriculture. The Atlas Cli- Storms that originate in the Gulf of Mexico have a sub-
55
[* i i — « 1 i
* !
i
t 1 | I< i i* i1 i 1I
i1 \ i j •
1 1 H* ;
j < 11 *i< i i
j= CNJ cn CO CNJ LO UD cn cn CO r- CSJ r—m CO CO o; oo ^r H CO CO t— P-. cn CNJ r^ co LO cn cn

OJ
S CO CO CO
’ —l '
CNJ
CO
’ '
id
CNJ
cd .—
CNI
CNJ
1
id LO
CNI
id
CNI
rd
’ 1
rd
*“
cd
“ 4
cd
r—'
d id
CNJ
—d
CNJ
od pd
*“
pd
r—
id CO
CNI
CO cd
CNJ
rd LO LO
CNJ CSJ
^d
CSJ
id
CNJ
CD CD CD LO CD CD CD CD CD LO LO CD LO CD LO CD CD CD LO CD CD LO CD CD CD LO CD CD CD CD CD CD CD
s i CD "* CNJcd cd LO *7 cd cn cd cd cd id LO cd cd cd •7 CD cd
T cn CD cd cd
X CD CD LO CD CD CD CD CD CD CD LO CD CD o LO CD CD CD CD LO LO CD CD CD LO LO CD — CD LO CD p CD
CNJ CNJ CNJ CO CO LO CNJ CSJ LO CO CO CNIcn cn LO cd cn cd cd
CO CO CO CO CO CO CO CO CO CO CO CN CO CO CO CO CO CNI CO CO CNI CO co CO CO CO
c CD °o CNJ CD CNJ CD LO CO p^ CD CNI CSJ CO oo , CD CNI CO CD CO CNJ CO CSJ , p^ LO

<D rd
•“
CO CO
CNJ
cd LO
CNJ
od
•” 1
cn
*—
cnJ
1
id
*” 1
d
CNJ
id
CNJ
rd rd cd
CNJ
r-*
CNI
pd ^d
*“1
1— cd
CNJ
UD cd
CSJ
pd
CSJ
id
CNI
pd d
CSJ
pd
CO LO CD CD CD LO CD CD CD LO CD CD LO LO LO LO LO LO LO CD CD CD CD LO CD CD CD LO CD CD CD
cn CD cd CO f"' CD id cn LO ^r cn LO id cd id CD
s *7 CNJ
7* *7 cn
'V
CD
X CD CD LO CD LO CD CD CD p CD LO CD LO co O LO LO LO LO CD CD CD CD CD CD CD CD CD p CD
CT3 CO
CO
CNJ
CO
CD LO
CO
CNJ d
CO
CO
CO
cn
CNJ
cd
CO
CO cd LO
CO
CSJ
CO
CD
CO
cn
CSJ
cd cd
co
cri
CNJ CO
CD
CO
cd
CO
CD cd 00
CO CO
*3-
CO
CD cd
co
CO
CO CO
c CO LO cn , OO CO CO CNI LO cn o LO LO CNI CSJ p^ p^ CD CNJ OQ LO CNJ CO CD CD CD CNJ LO CD ,
05 —
i CO CO cd CO rd cn
~ CO i — id LO cd pd pd
•“
pd cd CD od
•“*
cd id cd id cd rd id id cd id
CNJ csj w > CNJ CSJ 1
CNJ CNI 1 CNJ CNJ CNI CNI CNJ
CO LO CD p CD LO CD CD CD CD CD CD CD CD CD LO LO CD LO CD LO CD CD CD LO CD CD O p CD CD
cn r^
is 7 CD csj
7 CO cd
T 7 CO LO CO
T cn LO cd cd CO cd cd cd
7 1
1952-1962
><
-
CD CD CD CD CD CD CD CD CD CD CD CD CD LO LO CD CD LCD LO CD CO CD CD LO CD CD 0 cp CD p
CO CO cd cd CD — cd CO LO CO cd CD cd cd pd cd rd CD CO cd
s CO
CNJ
CO CO CO CO CO
,
CO CO CO CO CO csj CO CO CO CSJ
cd
co
. 1 INJ
CO co
CO
CO
cd
co CO
1
e LO CO LO _ CO p^ CO CO CNJ p^ LO CO CD cn CO CD CNJ CO co CO CD CSJ ^p LO CO cp

CO LO LO LO id cd id
a>
S
1
cd
CSJ CNI
CO
CNI CNJ CNJ
CD
CNJ
cd
7; d cd id 1
7: CNI
LO
7; CNI
id
cn CD cd LO CD CD CD CD CD CD CD CD CD LO CD LO LO lO CD CD CD CD CD LO CD CD CD CD CD CD CD
LO
cn CO — rd ad r— LO CD CD — cd LO id cd cd cn cd «d ^d od cd CD cd cd cn cd cd pd
Stations,
s .
1
.
X LO LO LO CD CD CD CD CD LO CD CD LO CD CD CD c=> LO CD CD LO CD CD CD CD LO CD CD CD
S CO
CO
.
CO
I CD —
.
cd
CO
CSJ
co
—
.
CO
d .— cd
CO
CD
CO
cn
CNI
—
. —
,
CO
rd
CNJ
.— cd
CO
—
.
CO CO
cn
CNJ
_
co
cd
CO CO
d id
co
CNJ
CO
rd
co
d cd
CO
rd
ro
cn
CNJ
c CD cn CD CNJ , CSJ CNJ CNJ cn LO CD CO CO LD ID oo CO , CNJ , CO CO ID p^ LO CD LO

CD cd cd cd od cn CO id LO LO cd CO cn id cn id cd od id CO id id UD
Selected
S CNJ CNI CNJ CNI CNJ CNJ CSJ CNI 7; CNJ
period) 1958
CD CD O CD CD CD CD LO CD CD CD LO CD CD LO CD CD CD o LO LO CD CD CD CD CD CD cp CD CD CD
S CO *7 cd CD LO
7 CD cn cd cd cd cd CO CO id CO cd cd cd LO cd cd id CD Csj CO
at X CD LO CD LO CD CD CD LO CD CD LO CD CD LO CD CD o CD LO CD CD co CD 10 LO CD 0 LO CD CD
s
CT3
rd CO
CO
CO
r
LO
CO
LO
CO
CNJ cn! cd CO CO -d cn cd id cd
CO CO
cd —
. cn CO rd
CO
cn ^d CD CO
^r CO
cn co LO cd CD
CO *' CO CO oo CO CO CO CO CO CNJ CO CNI CO co CO co CO
the
(°C)
c °q p^ csj p^ CO CD ,
cn cn , CNJ co CO CD cn CNJ cn LO LO CD CD p^ , CO ^p CO CD CNJ CD cp
<u cd cd cd cn id csj id cd cd cd cn LO CD LO cn CO UD LO id
for S 7; Csj CNJ CNI CNJ CNJ 7: CNJ CNI 7; cd 7;
LO
0-1
CD O CD
rd
CD LO CD CD CD o CD CD CD LO co LO LO CD LO CD O CD CD CD LO CD
—
CD CD CD CD CD
CD .
CD LO CD CD CD LO cd cd 1 cd , I id id cd cd ^d CO I LCD CD , CD id cd cd
1 —• 1 1
records
><
-
CD
LO
CD LO CD
cn
CD
id
CD CD
CD
CD
CD
LO
cd
CD LO
cn
CD CD
id id
LO
—
CD
CO
cn
cd
CD
CO
LO
pd
CD
•d
LO
pd
LO CD
cd LO
CD
cn
CD
id
CD
CD
CD
cd
CD
cd
<=> CD
CD cd
CD
cd
CD
co
p
CD
Temperatures
s CO co
j CNJ
CO CO CO '=r CO CO co CO co
,
CO CO CO CNJ “ cr CO co CO CNJ co CO CO CO
c CO , CSJ cn , CO co cn __ CNJ CNJ CO , CO LO ^p co cn CNJ CO CD LO CD CO p^ LO p^. CD

are
QJ UD cd LO rd LO Csj id cd id CNI cn cn CO id —
, -d CD rd id CO
— LO CO LO rd
— tt id CD' id
s CNJ CNJ r_l CNJ CNJ CNJ CNI CNJ CSJ CNJ CN CNI r_<
CO CD CD LO CD CD LO CD LO CD Cd CO CD CD CD LO CD CD CD CD CD LO CD LO CD CD CD CD
LO ,a-
in cd CD cd CD LO cn cn cd
s 7 CO CO
1
1
•f *7 7“ cd cd
7 7
Mean figures
X cn CD CD LO LO CD C CD LO LO LO CD LO CD CD LO CD CD CD CD CD LO c CD LO LO CD CD
S oo
CO
,
CO
’
, i cd
CO
CD
CO
1 —
,
CO
rd
CO CO CO
LO cn UD
CO
cd
co
d —
.
CO
<n
CSJ
CO id
CO
pd
CO
^d
CO
cn
CNI
CD
CO
CD cd
CO
CO
CO
cn rd
co
r-4
CO
pd
CO
CD
CO
and
C CD cn , CO CNJ LO CO , CNJ cn CNJ cn p^ CO cn co UD
pd LO od id 1 LO r-d —
. cn -d cd 1
— cd id rd id UD
(Boldface
S CNJ *“• CNJ CNI CNJ CNI CNJ CNJ 1—1 CNJ CNJ
LO
LO LO CD CD CD o CD CD cn CD CD LO LO CD CD LO LO CD CD LO cp
cn
S CD
1 T 7 cd cn CD 0 LO LO CD*
T 7 LCD GO CD cd LO cd
Minimum,
X cn CO CD CD CD LO CD CD CD LO CD CD CD CD LO LO cp cp CD CD
LO CD oo LO cn — id cd cd id CD — cd CD — CO
s CO
csj
CO
csj
CO '=r CO CO
.
co CO CO *3’ CO CO CO
.
CO co
.
CO
C ,
CO CD cn CD __ cp
|
p^.
cu rd pd cn id CO id rd 1
UD
S
\
r- ' CNJ r~' CNJ * —l CSJ -H

' '1
LO LD CO cd CD CD CD CD LO CD CD CD CD
Maximum, cn CD id id cd '=,‘ id cn
S cd' csj
cn CD CD CD CD LO CD CD CD CD co CD
S LO
CO
CO
CO
cd
"=r CO
id
CO
cd
CO
cd cd
co
d
co
cd
co
LO
CO
1.
c CD „_p LO oo LO p^ cn , CO
LO id cd cd
CU
s cn
7 id CO 1
CNJ CNJ
Table.
1953
CD CD CD CD CD CD CD CD LO CD CD C
CO LO LO LO CD
s id
7 1
cn CD cd cd
CD CD ,3- CD CD CD LO LO LO CD 0 p
id CO LO LO ^d cd co 00
s
1
CNI
CO CO CO CO CO CO co co CO
c cn CD CO LO LO , ID cn CD co p^ CD
cn cd cd id cn rd id <d ^d id rd rd
S r~* CNJ csj CNJ CNJ *” H
ini
CD CD CD LO CD CD LO CD CD CD cp LO
cn id cd cd cd
S 7 cd cd CO
CD CD CD LO CD CD CD CD LO O cp CD
—
cn CD CO cd cd CO cd CO cd .
CO CO CO CO CO CO co CO
< < C O O O O ooooooooluhz

1 0 0
-5 0 • Tecamachalco -9.0
Cd. Serdan -7.0

Xochitlan -3.0
y
"v_" ;
I9°00
Tlacotepec
mm .7.
Cacaloapan -1.0
98°00 Coyotepec -4.0 7? _

<Chapulco -4.0
/
Acultzingo -1.5
EL RIEGO
El Riego
Tehuacan
Zapotitlan -7.0
TECORRAL CAVE
MARCOS CAVE
|/3AN
'
iSi
San Antonio
Canadas 1.0-
-
i V
I8°00‘
t Salinas de
Barranca
v v-A ’
i
i-
\* Chilac -0.5
Caltepcc
QUACHILCOA' AJALPAN
LAS CANO AS
> COATEPEC
Tehuipango -1.0
Axusco 2. 0
Coxcatlan 4. 5
;an Miguel Zoquitlan 1 .
Astatla -3
XOXCATLAN CAtfE
AâBEJAS cave 7*
X.t PURRON CAVE 7
Tlacotepec de
Diaz 6 .
Calapilla 3
[ignacio Teotitlan del
Mejias 1
Camino 1."
0.3
Tepelmeme
r-
777
Xiquila 4. 6
Cataluna 4.0
Teopoxco 1. 0
/-g Huautla 3.0
[ffjocotipac -9.5 Quiotepec 4.5
£&/ 1.0
< '
Cuicatlan 7. 0
Fig. 22. Minimum temperatures, in degrees Centigrade, Tehuacan Valley and environs, 1955-1962.
57
stantial effect on the Papaloapan watershed only if they In 1955, Hurricane Gladys took shape off the coast
are formed off the coast of Veracruz. Storms originating of Campeche and Tabasco on September 1, and fol-
in the Pacific may follow the west coast or veer sud- lowed a course that took it northward parallel to the
denly to the eastward and cross southern Mexico. coast. It produced heavy and general rains throughout
These do affect the Papaloapan watershed, although the valley and more than 80 mm. throughout the Mix-
they do not produce as heavy rainfall as the storms teca Alta. A remarkable feature of this storm was a
which originate in the Caribbean or the Lesser An- center of heavy precipitation in the Zapotitlan Valley.
tilles. In the accompanying tabulation the daily precipitation
Routes of seven storms that occurred from 1932 to in millimeters during the life of the hurricane at se-
1955 have been charted and rainfall analvzed.
J
We here lected stations in the valley illustrates the spotty nature
discuss some of these storms as they affected the Te-
Sept. 3 Sept. 4 Sept. 5 Sept. 6 Sept. 7
huacan Valley; rainfall east of the Sierra Madre was
far heavier. In September 1944, a hurricane was born in Altipexi 13.0 60.0 25.5 32.5 2.0
the Caribbean west of Jamaica. On the nineteenth it El Riego 17.5 23.0 44.5 34.5 3.5
was in the Yucatan Channel; on the morning of the Calapilla 11.3 38.2 30.3 32.4 10.3
Chilac 41.2 6.0 74.5 31.5 3.6
twentieth it came ashore on the tip of Yucatan near
Quiotepec 2.1 9.0 25.6 44.5 19.5
Puerto Juarez. By the night of the twenty-first it had
Salinas 26.6 15.5 83.4 67.1 1.4
dissipated over the Isthmus of Tehuantepec. However,
Teotitlan 28.5 34.8 32.5 35.5 13.5
it appears to have regenerated as a less violent storm
Zapotitlan 33.0 20.0 80.0 66.0 2.0
and moved northward, with its center off the coast of Coxcatlan 58.5 58.5 32.5 38.5 52.0
Veracruz. This storm brought rain to the Tehuacan
Valley in the following amounts: on the twenty-first, of the rain produced by such storms. Although juxta-
10-30 mm., with the heavier fall on the immediate position of Chilac and Quiotepec in this tabulation
flanks of the Sierra; on the twenty-second, from less appears to point to a movement of the storm, or at least
than 30 mm. to more than 50 inm., with the heaviest an active from north to south, we do not have data
cell,
fall on the flanks of the Sierra Madre and on the Sierra adequate to determine such movement.
Mixteca, while no rain fell in the center of the valley Other tropical storms which do not reach hurricane
between Tecamachalco and Ajalpan; on the twenty- force may produce more rain in the Tehuacan Valley
third, between 50 and 100 mm. fell on the Sierra Mix- than full-blown hurricanes. However, the pattern of
teca and in the lower valley, but in the upper valley, rainfall produced is as irregular as the rain which re-
from Altipexi northwestward the fall was 10 mm. or sults from hurricanes. While on a given day during
less. such a storm no rain may fall at any given spot, al-
A hurricane took form in October 1950, south of the though heavy rains may fall nearby, the succeeding
area in which the hurricane of September 1944 was day may bring from 20-50 mm. of rain.
born. It crossed Yucatan, passing overTulum and Me- Heaviest rainfall may come during the hurricane
rida. It then swung westward and southwestward to season, between August and October. In the thirty-
late
come ashore south of Veracruz. On October 10 it pro- year period 1926-55, a total of 450 tropical storms of
duced amounts varying from nothing in the
rain in hurricane or near-hurricane intensity were recorded.
vicinity of Tecamachalco to more than 50 mm. at Teo- Of this total, 127 storms originated in the Lesser An-
titlan, except for an area around Calipan, Coxcatlan, tilles or beyond; five reached the Papaloapan water-
and Venta Salada, where no rain fell. On October 11 shed, two passed within 100 km., six passed at a dis-
light rains of 0-10 mm. fell in much of the valley, but tance of 100 to 400 km., and 114 swung northward more
the rainless area extended northwestward along the than 400 km. away. The Caribbean was the birthplace
eastern side of the valley from Coxcatlan. On October of ninety-eight storms, of which five reached the Papa-
12, rains of 0-5 mm. from Tehuacan southeastward
fell loapan watershed, eight passed within 100 km., and
along the flanks of the mountains and spread westward fourteen at a distance of 100 to 400 km. Off the coasts
to include much of the Zapotitlan Valley. Altogether, of Campeche and Tabasco seventeen storms took
on October and 12 the storm produced 10-20
10, 11, shape; three of these reached the Papaloapan water-
mm. between Tecamachalco and Tilapa, with
of rain shed, two passed within 100 km., and twelve passed
up to 50 mm. in the Sierra and between Tilapa and between 100 and 400 km. away. None of the storms
Teotitlan. In the extreme southern end of the Salado originating within the Gulf of Mexico passed within
Valley rainfall exceeded 50 mm. 100 km. of the Papaloapan watershed; as a result, their
58
—
effect is slight. A total of 135 storms originated in the then only 0.4 mm. Bains were general on the twenty-
Pacific Ocean and followed a course parallel to the ninth and thirtieth when Estela took shape inHon-
west coast. Of these only three passed within 100 km. duras. This storm passed close to the Oaxaca shore
of the watershed, two were between 100 and 400 km. before going northwestward into the Pacific on Sep-
away, and 130 were more than 400 km. distant. Storms tember 9. General and heavy rains occurred at valley
originating in the Pacific and coming ashore numbered stations September 5-8, but there was no rain at all
fourteen, of which four reached the Papaloapan water- at Huautla.
shed and five passed within a distance of 100 km. Of On September 19 Coxcatlan received 140 mm. of
the 450 storms, 3.8 percent reached the Papaloapan rain; Altepexi, 88.0; Calapilla, 56.5; Salinas de Barranca,
watershed, 4.4 percent were within 62 km., 10.9 per- and Zapotitlan, 57.0.
55.9; Teotitlan, 90.6; Xiquila, 51.5;
cent were 100 to 400 km. away, while the great major- Huautla received only 13.8 mm. This rainfall may have
ity, 80.9 percent, passed at a distance of more than 400 resulted from the tropical storm known as V2 which
km. and had little or no effect on the Papaloapan drain- took shape in Yucatan Channel September 16 and ran
age. westward to the Sierre Madre de Oaxaca, where it
Unfortunately, the data regarding tropical storms is ended its career on September 20.
not included in numbers of the Boletin Hidrologico The only other year between 1955 and 1962 for which
covering the years 195.5-59 and Boletin 13, for the year we have detailed records of tropical storms and of rain-
1961, is not available. Therefore, we have only Boletin fall for the valley During that year, thirteen
is 1962.
12 and Boletin 14, for 1960 and 1962, on which to base tropical disturbances were reported between 15° and
our analysis. 25° N. Lat. and 90° and 105° W. Long. Boutes of the
During 1960 seven tropical disturbances occurred on majority of these storms took them some distance from
the Atlantic side and seven moved up the Pacific coast. the Tehuacan Valley. Within the valley itself substan-
Among Atlantic storms were five whose tracks took tial rains occurred coincident with only two storms
them far from the Tehuacan Valley as far away as — —

V2 and Berenice both on the Pacific slope. There were
the southeastern United States and the open ocean. also rains at Acultzingo and Lagunas coincident with
During the periods of their life no significantly heavy these two storms. Although only light rains were re-
rain fell in the Salado watershed, nor did any fall dur- ported in the valley, more generous rain occurred on
ing the life of the storm known as Annette, born well off the Atlantic slope coincident with the passing of V3
the isthmus of Tehuantepec on June 9, to travel on a from Tabasco Tampico, close in shore. One may see
to
northwesterly course far from land, passing from the a possible correlation between the two events. Rain
record books June 13. Shortly after Annette’s passing, fell generally in the valley and on its western and east-
Bonnie was born off the coast of Guatemala, June 18, ern rims from August 31 through September 3 and
to move up the Pacific coast and die in the Gulf of continued locally for a day or two more at various
California on June 25. General rains, sometimes heavy, stations. The Boletin Hidrologico records no tropical
occurred in the Salado watershed June 12-29. General storm in this interval. It should be noted that few tropi-
rain occurred July 4-7 when there was no tropical cal storms passed close enough to the valley in the year
storm. Hurricane Abbey, born in the Virgin Islands 1962 to have any effect. Without detailed data it is not
July 11, died against the Sierra Madre de Oaxaca July possible to be certain that rainy spells coincident with
16, as Hurricane Celeste was taking shape in Honduras. such storms are in reality caused by them. Generally
This storm passed up the Pacific coast and lost its iden- stormy conditions associated with areas of low baro-
tity as a hurricane July 21. General and heavy rain oc- metric pressure may in reality bring the rains. From
curred in the Salado watershed July 16-20. General these low-pressure areas tropical storms may develop
rain fell again between August 14 and 26. Although under favorable conditions, attracting more attention
heavy rainfall in the valley during July coincided with than larger, less violent storms.
the passage of Hurricane Diana off the coast of Oaxaca, Thus it may be seen that the hurricane season can-
July 17-19, the heaviest rainfall at Altepexi, Calapilla, not be counted on to bring rains to the valley. Further
Salinas de Barranca, Teotitlan, and Zapotitlan fell detailed study is required to determine whether Pacific
August 24 when no tropical storm was reported. storms are of major importance as sources of rain.
Another period of rain began August 29. The twenty- We have observed above that a marked cyclical as-
eighth was the day of heaviest rainfall of the year at pect is characteristic of the rainfall of the region. The
Huautla de Jimenez, although only one of the valley interval of the cycle is not clearly evident in any readily
stations —
Calapilla —
reported measurable rain and available data. The Atlas Climatologico makes refer-
59
1940 1950 1960 1940 1950 1960 portance of amounts of rainfall is eliminated. Instead,
we have graphic representation of the rate of
shall
variation from year to year and in this way be enabled
to compare stations with minimal rainfall with those
where rainfall is abundant. These graphs appear to
demonstrate that the rainfall over the area fluctuates
according to two and possibly three rather
at least
characteristic patterns —
that exhibited by stations in
the valley and especially the southeastern valley and
that exhibited by stations on the eastern slope. A third
possible pattern may be exhibited by Oaxaca and sta-
tions on the western slopes. A far longer record would
be necessary to establish the reality of this division.
Rain, when it does come, may strike with the force of
a cloudburst, cutting gullies and eroding unprotected
slopes. Its effect on sloping fields on the flanks of hills
is maximum. Although we lack information regarding
intensity of rainfall for many of the valley stations,

Boletin Hidrometrico 14 lists, for four stations in the
valley, the figures for maximum intensity in terms of
1000 millimeters per hour according to minutes of duration.
These we present
1
in the accompanying tabulation.
Salinas de
Minutes Calapilla Quiotepec Barranca Xiquila
duration 7/5/59 5/29/49 6/10/59 9/3/62
5 240.0
10 156.0
15 132.0
20 112.5
30 62.0 86.4 100.0 55.0
45 62.2
60 61.5 47.7 62.0 38.0
80 36.5
90 4.5 48.6 26.1
100 30.0
Fig. 23. Annual rainfall at selected stations, 1941-1962, 120 33.8 25.8 37.6 20.0
in mm. on logarithmic scale. Left: east of the crest of the 240 17.6 20.7 10.3
Sierra Madre de Oaxaca. Right: west of the crest.
Thus Quiotepec. with 240 mm. per hour

,
for five
ence to a period of abundant rainfall from 1926 to minutes received 20 mm. of rain in that interval. Over
1937, followed by a period of deficiency from 1937 to a ten-minute interval the intensity was 156 mm. of rain.
1951, followed by abundant rainfall from 1952 to 1955. Rainfall over two hours at the rate of 25.8 mm. per
Annual precipitation for the period 1941-62, as pub- hour brought 51.6 mm. At Salinas de Barranca, not far
lished in Boletin Hidrometrico 14, makes it clear that from Zapotitlan Salinas, the maximum intensity of 100
rainfall was low at many stations over the span of a mm. per hour for 30 minutes produced 50 mm. of a
few years, but that abundant rains come at intervals, total rainfall of 86.3'mm. on June 10, 1959. At Xiquila,
except at Huatusco which is outside our map area, on on September 3, 1962, 27.5 mm. of a total of 43 mm. for
the Gulf slope, where a period of low rainfall persisted the day fell in one-half hour. The greater part of that
1945-54. In many stations Oaxaca-Puebla
in the total fell in four hours, leaving only 1.8 mm. to fall
Trough this interval was interrupted by a year of good during the rest of the day.
rains in 1947. Such quantities of water, falling on dry or nearly dry
If rainfall is plotted on a logarithmic scale, the im- ground, unprotected by vegetation, are capable of pro-
60
during intense washing in a very short interval. Soon 1940 1950 1960
1940 1950 1960
gullies are filled with tumbling stones and a flood of 1500
mud. Larger and the

gullies carry fair-sized boulders, 1000
800
barrancas, boulders of the size of a hogshead. If this
600
water can spread out over a broad flood plain at a 500
400
slackened pace some of it will sink in. Such is the
1500
character of the terrain in the Tehuacan Valley that this
1000
rarely occurs. According to the tables of discharge, the 800
response of the streams in the valley is almost instan- 600
500
taneous, and the flow as quickly dies. Thus, although 400
the Rio Zapotitlan carried 13.1 cubic meters per second 1000
800
on June 10, 1959, it carried only 0.485 cubic meters per
600
second on June 11. 500
400
Extensive disturbance of the natural plant cover 300
accompanied the extension of agriculture to new parts
of the valley, and Miranda (1948: 344-46) has ob-
served that nearly pure stands of cacti may result from
human intervention. Disturbance in the oak-pine zone
would have been severe, especially if ancient agricul- 1000
tural practices were based on the milpa system of slash- 800
600
and-bum, as seems likely in view of its widespread use 500
400
by Indian farmers today. With the increase in popula-
300
tion and developing agriculture, additional inroads on
200
plant cover were made by cutting forest trees to clear
new cropland, for firewood, or for fuel for burning lime
100
for structures faced with plaster or stucco. When steep 80
mountain sides were eventuallv brought under cultiva- 60
tion,such disturbance rendered the soil easily suscepti-

Fig. 24. Annual 1941-1962, west of the crest of
rainfall,
ble to washing. Where the regolith on which soils de-
the Sierra Madre de Oaxaca.
Left: except for Puebla and
veloped was not derived from acidic rock, caliche-
Tepeaca, in the Puebla-Oaxaca Trough. Right: except for
like tepetate has formed, making the soil almost im-
Oaxaca, west of the trough.
pervious to moisture at relatively shallow depths. Such
circumstances in combination must have caused the
water table to begin to drop, a phenomenon that prob- Heavy rains remove great quantities of yellow loam and this
ably had its inception in Classic times at the latest. is the dominant color of the rivers and brooks. Red sand-
Loss of marginal fields to erosion and a lowered water stones near Cuicatlan give a red color to the waters of the
table can only have served as a stimulus to the develop- brooks, but not to the big rivers. With heavv rains toward
the higher parts, and with no rain in the lower, the volume
ment of irrigated fields on the valley floor.
of rivers increases immediately and their turbulent waters
The introduction of sheep and goats by Europeans
become tinged with vellow. It seems, therefore, as if immod-
must have accelerated processes of erosion. During
erate clearing and farming activities on steeplv sloping land
the course of a botanical survey of the upper Papaloa-
in the upper parts of the basin were responsible for the in-
pan watershed, Miranda made apposite observations creased erosive action in places where the aridity of the cli-
concerning erosion. In discussing the higher parts of mate and the irregularity of the rains are already producing
the basin, he remarks (1948: 346—48): bv themselves active erosion.
In other days, groves of oak and pine must have evenlv
. about 1400 meters in the Cuicatlan district and above
. .
covered all high country over the limits noted above. Thev
2200 meters in the country around Tehuacan the xerophitic are now greatly reduced bv clearing; nevertheless one may
vegetation stops abruptly and is replaced bv oak and pine still find great expanses of very beautiful forests in places
forest.Farms in this zone almost exclusively produce rainy - farthest from the towns [my translation].
season crops, especially corn and beans, which thev supply
to markets in the low country. Forests in this zone are being On page 362 Miranda observes that erosion is very
actively felled in order to establish new tillage; as the land intense in allthe upper parts of the watershed. These
is in general very sloping, erosion is progressing rapidly. seem to be the source of the greater part of the detrital
61
EN VIRONMENT AND SUBSISTENCE
Coixtlahuaca. The base is an arbitrary line from Coix-

tlahuaca to the headwaters of the Rio Apoala, thence
to the Rio Grande at Tomellin.”
Cook
gives the elevation of the plaza of Santa Maria
as 1950 meters, but the elevation in Bolei'm Hidro-
metrico 14 is 1850 meters. It is clear that this does not
vary greatly from the higher parts of the Tehuacan
Valley. However, annual precipitation is greater at
Santa Maria Ixcatlan than at any station in the Tehua-
can Valley itself. Fluctuations in rainfall occur at Santa
Maria just as they do in the Tehuacan Valley, the ex-
tremes 195.5-62 being 1063 mm. in 19.55 and 378 mm.
in 1957. Distribution of rainfall through the year cor-
responds to that at stations in the Tehuacan Valley,
there being a peak in June and July, and a second peak
Fig. 25. Farming is sometimes undertaken on steeply slop-
ing fields.
in September and October. In response to this cycle,
—
two plantings of com were made one in milpas set
high on the mountains to make use of moister condi-
material carried bv
J
the rivers whose waters never be- tions, was made in late winter, February to March. It
come clear in any season of the year. He notes that germinated on moisture in the soil and rain brought
from natural causes erosion is most intense in this zone by the rare winter showers, but ripened and matured
because of the relatively immature rolling relief, the under the effects of the rains of the early season. The
aridity of the climate which does not allow a dense second planting was made just prior to the early rains,
vegetal cover to develop, and the irregularity of the so that it might receive their moisture for germination
rains which come all at once at certain times of the year and early growth. This crop grew to maturity on the
and send the rivers over their banks. However, it is his later rains and was harvested in November and De-
view that responsibility for the enormous increase in cember. A similar planting schedule is typical of much
erosion must he assumed by man, especially the resi- of Central America under a climatic regimen in which
dents of the high country with their ill-directed farm- there are two peaks of rain. (La Farge and Byers 1931:
ing practices, their immoderate clearing of the forests 77.) Failure of the first of the rainy seasons to material-
to establish new tillage, their farming activities under- ize can spell disaster for both crops.
taken with no precautions at all, even on steeply slop- Cook was concerned with long-term changes in rain-
ing fields, and also at times their fires and the over- fall. He cites the occurrence of epiphytes and para-
grazing of their sheep and goats. siteson forest trees and the lack of seedlings as evidence
That the dropping water table and consequent dry- of a deteriorating forest,and adduces information from
ing up of springs and wells dates hack at least two natives that forests were once more extensive as possi-
centuries is clear from records of abandoned towns in ble evidence in support of a proposition that there has
northern Oaxaca, just west of the Canada. The present been a reduction in total rainfall. He goes on to suggest
town of Santa Maria Ixcatlan lies only 24 kilometers, as an alternative the loss of effective rainfall to run-off
as the buzzard flies, to the west and north of Cuicatlan. as a result of processes of erosion set in train by the
It is situated in a small valley enclosed by hills on three inhabitants themselves through many years of cutting
sides and drained by the Rio Ixcatlan, a tributary of firewood, through the practice of slash-and-burn agri-
the Rio San Pedro, which in turn drains into the Rio culture, and finally through the devastating effects of
Grande not far south of Quiotepec. The town has been running sheep and goats.
the subject of a detailed study by Cook (1958), who He then cites seven small towns — the pueblos des-
describes its situation as “the center of an extensive jwblados— which formerly existed northeast of Ixcat-
mountain mass of triangular shape, the apex of which is lan. According one of Cook’s informants, four towns
to
formed by the confluence of the Rio Hondo and the Rio lay on the north side of the Rio Seco, “extending to near
Salado about 10 kilometers south of Teotitlan del Ca- —
the Rio Salado” San Miguel Nopalapa, San Jeronimo,
mino.The eastern side of the triangle follows the Rio —
Santa Cruz, arid San Cristobal and three were on the
Grande southward to the vicinity of Tomellin, and the —
south side of the stream San Juan Viejo, Santiago, and
western side follows the Xiquila to its headwaters near San Antonio Nopala. Cook visited the site of San Juan
62
Viejo and observed “across the deep canyon of the present village are at the up-stream end of this terrace.
Rio Seeo to the north four open or bare places
. . . Cook further discusses the possibility that the village
high on the side of the mountain. These can be none was in a mature valley above which were forested
other than the worn and eroded
sites of the four towns slopes or heavy shrub and grassland resting on a mature
of the outer line,” on the northern side of the river.
i.e., soil. this primeval cover by human
Disturbance of
Cook further cites “irrefutable documentary evi- activity would then open Pandora’s Box and let flood
dence” dating from 1578 involving persons from Santia- waters loose, with consequent erosion, deposition, and
go, Nopala, and San Juan Coyula, the last-named said eventual gullying.
to be the equivalent of San Juan Viejo. Records of a In the presence or absence of domesticated plants
surveying party sent out in 1743 to delimit the bound- Cook and rainfall. He re-
sees a guide to both climate
aries of Ixcatlan show that they visited a place “that two loads of cotton paid in 1553 by
fers to a tribute of
they call Santa Cruz, an old and uninhabited town the town of San Miguel Huautla (not Huautla de Ji-
where there are two orchards of fruit trees.” (This and menez, a highland Mazatec town east of Teotitlan). Re-
the translations below from the Spanish are my own.) gardless of the correctness of translating Ixcatlan as
A document dated 1778, from Coixtlahuaca, states “the place of cotton,” Cook avers that if cotton could
“. and although this town of Ixcatlan had other sub-
. . be grown in Huautla it could be grown in Ixcatlan. He
ordinate towns in their time, no one lives in the districts goes on to say (p. 17) that “cotton of the ancient type
where they were, [they have] already gone wild and requires more warmth and moisture than there is in
completely sterile, and only the ruins of their churches Ixcatlan (or Huautla) today. Here a definite change in
are to be seen.” climate may be predicated.”
Thus it is clear that these towns were abandoned at In respect to plants and animals, the Relaciones
some time between the end of the sixteenth century and Geograficas continues:
the end of the eighteenth. Cook cites two possible
Fruit trees include chicozapotes which fruit in the months
causes, including a change in climate resulting in less
of July and August, and also sugarcane, and
there is
rain and a diminution of the effective water supply as a
town and in addi-
black(ish) sapotes in the lower parts of
result of deforestation and erosion of the soil, but he
tion many huaje trees and some papaya trees; also there
does not feel that the evidence justifies the acceptance
are pulque magueys, and anonas [custard apples] and all
of one and the total exclusion of the other. He does are not very abundant: such can only be said of the six-
state “certainly there is no visible sign of water at pres- month corn, sown and harvested in this interval.
ent, and no possibility of profitable agriculture in the
arid, sterile basin of the Rio Seco. All this might argue Cook states that “today there are a few (20-30) sapotes,
for a diminution ofwater induced by a change in cli- very few huajes and anonas and no papayas or sugar-
mate. The gradual desiccation seems clear, since before cane.” The “virtual extinction” of these semitropical
1600 the area was able to support seven towns which domesticated plants he attributes to a change in cli-
originallymay have been of fair size.” mate, after averring that there is no loss of interest in
There is scant historical evidence bearing on the fruit trees because people have kept the sapote, intro-
water supply; however, an excerpt from the Relaciones duced the fig, maintained the mulberry, and introduced
Geograficas del Siglo XV III reprinted by Cook as Ap- some drought-resistant trees in recent years. He then
pendix I states in part: asserts that erosion and soil desiccation cannot be in-
voked and that a change of climate must have occurred.
The aforesaid town (Santa Maria Ixcatlan) is not very According to Cook, the “rather profound alteration in
abundantly endowed with sources of water, for it has only
the physical environment during the past few hundred
one river (which comes to an end after a descent of a half
years” has driven the people of Ixcatlan from diveisi-
a league) and one poor well and all this water is of middling
fied agriculture to corn alone, and from the raising of
quality, and two very small arroyos also run somewhat
brackish water (only in the rainy season).
corn to the manufacture of hats made from leaves of
the overabundant palmetto Brahea dulcis), a plant of
(
The village is situated on a terrace-like formation on arid situations.

the Rio Ixcatlan. This appears to have been a zone of If, now, we turn from Cook’s work to Brunet’s hy-
alluviation such as is described by Brunet in Chapter pothesis (Chapter 5) of a steadily lowered water table,
5 below. Subsequently the river has cut through this we see that it is possible to interpret all phenomena
terrace-like feature and in doing so lowered the water listed by Cook as resultingfrom continued draining
table. Modern wells used for the water supply of the away of water resources and a lowered water table
63
brought on by headward erosion of streams. Effects of ture on the opposite sides, and thus increasing stormi-
this cycle are further accentuated by the rapid run-off ness. Next would come a small but intense anticyclonic
of precipitation, in turn causing a decline in effective belt with a narrow belt of powerful trade winds. The
precipitation which has the same effects on plants as a equatorial trough of low pressure would be deepened.
general decline in precipitation. As a result of these changes
If this be the case in the Ixcatlan valley, it must also
outside the great ice sheets there is evidence of a much
have applied in the Tehuacan Valley. By analogy we
greater rainfall (snowfall in the mountains) in two belts, one
may infer that with a higher watertable, and with a along the new storm tracks a short distance equatorward of
larger effective rainfall, it is likely that the aspect of the ice edges, and the other along the equator. The former
the Tehuacan Valley in 1500 was less harsh and arid gave rise to the great lakes of the Great Basin of America
than that of the present and that sapotes, chicosapotes, and of the interior of Asia, and to a large number of moun-
avocados, papayas, and doubtless other fruit trees and tain glaciers, the latter to the greatly increased lakes of
plants which are alleged not to grow in the Tehuacan Central Africa and to the glaciers of Kenya, Kilimanjaro,
Valley without irrigation were in reality to be found Ruwenzori, and parts of the Andes. Between these two belts
until about two centuries ago in suitable spots there

the evidence of greater precipitation is more indefinite and
irregular.
or in the beds of barrancas leading down from the
Sierra. In such barrancas they must in all likelihood One may observe that at present winter is the dry
have once been found farther down slope than they season in Tehuacan Valley and in central and
the
are at present, especially in the vicinity of Coxcatlan, northern Mexico. Cold winters would not increase pre-
Purron, and Abejas caves where there is now somewhat cipitation, for snow falls on Citlaltepetl, Popocateptl,
more rainfall than in the central, western, and north- and Ixtaccihuatl during the summer rains, and only
western parts of the valley. rarely under winter conditions, when the snow usually
Such a supposition is given further credence by melts.
Wagner’s comments (1964: 253) concerning what may There seems to be no grounds for extending to the
be relict species and associations in suitable locations Tehuacan Valley the “Pluvial period” of the climatic
in the upper Papaloapan drainage, almost surely in- cycle developed in the southwestern United States by
cluding the Salado: “The original cover may have been Antevs (1955). Recent work by Aschmann (1958),
very different from that of the present, and Miranda Martin (1963), Martin et al. (1963), and Mehringer and
(1948) suggests that the local climax may have been a Haynes (1965) has cast doubt on the reality of the
mesquite forest of Prosopis jttliflora." “Neothermal Climatic Cycle.” A paper by Bryan and
No one save Flannery (Chapter 8) has yet adduced Gruhn (1964) points to fallacies involved in extending
any good reason to believe that the climate of the Te- such a cycle, if indeed it exists, beyond the ecological
huacan Valley has undergone much change in post- zone in which it was developed as a means of dating
glacial time. There is no evidence that the Valderan aboriginal occupations or for the purpose of inferring
substage of Wisconsin glaciation of the Great Lakes climatic conditions for dated aboriginal occupations in
Basin brought major climatic changes to the south- North America.
western United States (Martin 1964). In the aboriginal To extend the Pluvial period of Antevs to the Tehua-
condition of mesquite forest postulated by Miranda, can Valley seems particularly unjustifiable, because the
there would have been areas of grassland such as are southwestern United States lies in the flow of winds
postulated by Flannery for the environment recpiired from the poleward side of the zone of subtropical di-
by the antelope and jack rabbits of early Ajuereado vergence now lying near 30° N. These ultimately be-
time. come and contribute to the cyclonic storms
westerlies
In the absence of any pollen record from early times of mid-latitudes. The Tehuacan Valley, on the other
in the valley one is forced either to postulate a period hand, lies within the zone of trade winds originating
of cooler and somewhat moister climate— a postulation from the same source, but flowing toward the zone of
for —
which we have no support or to accept Miranda’s equatorial convergence. There is, therefore, a consider-
view. able barrier to any extension, however hypothetical, of
In regard to a difference in climate and rainfall in the postulated climatic phases of Antevs to the Tehua-
Pleistocene times, Brooks (1949: 59) has pointed out can Valley.
that the onset of glacial conditions would intensify Several climatic models for the time of Pleistocene
present winter conditions. The polar front would be glaciation, including that of Brooks, postulate a narrow-
pushed southward, increasing the contrast in tempera- ing of the zone of equatorial displacement of the sub-
64
tropical zones of divergence —

the subtropical highs now even less well watered than it is today. It may have been
near 30° of latitude. There must of necessity have been less well watered, but possiblv cooler, with a climate
an accompanying narrowing of the belt of trade winds approaching that of San Juan sin Agua in northern San
and an increase in their force. It is also postulated that Luis Potosi.
the seasonal swing of the entire tropical and subtropical Because of these possibilities it would require a series
system of winds was reduced in amplitude. A reduction of radiocarbon dates for glacial features on peaks of the
in the width of the zone of trade winds could only serve neo-volcanic axis, or a series of radiocarbon dates for a
to move and
the drier portion of this belt farther south, number of pollen profiles reflecting climatic changes,
a reduction in amplitude of the seasonal swing would in order to establish an acceptable correlation of as yet
serve to prevent the moister portion from swinging so undiscovered climatic phases in the Tehuacan Valley
far north in the summer. There is thus a good possibility with hypothetical ones in the United States.
that in late Pleistocene time the Tehuacan Valley was
REFERENCES
Antevs, 1935, 1962. Hubbs and Roden, 1964.
Asehmann, 1958. La Farge and D. S. Bvers, 1931.
Atlas Climatologico e Hidrologico, 1956. Martin, 1963, 1964.
Boletin Hidrologico, 1955-60, 1962. Martin et al, 1963.
Brooks, C. E. P., 1949. Mehringer et al., 1965.
Brvan and Gruhn, 1964. Miranda, 1948.
Bvers, H. R., 1959. Riehl, 1965.
Collier, 1964. Tamavo and West, 1964.
Cook, S. F., 1936-39, 1938. Vivo Eseoto, 1964.
Hess, 1959. Wagner, 1964.
65
CHAPTER 5
Geologic Studies
Jean Brunet
n this chapter the term “Tehuacan Valley” applies of Puebla. It developed that they could furnish some
0
only to that part of the entire Tehuacan basin which facts directly concerning archaeology.
I is bounded on the northwest by Tepanco de Lopez
The map shows that the region covered by this sur-
and San Andres Cacaloapan, both in the state of
vey comprises approximately 1600 square kilometers,
Puebla; on the eastern side by ranges of the Sierra
but the total area includes some 2000 square kilometers,
Madre de Oaxaca, including from northeast to south-
thanks to the relief, which unfortunately could not be
east the sierras of Acultzingo, Zongolica, and Maza-
covered in sufficient detail because of the lack of time
teea; on the southwest by the Sierra de Zapotitlan; and
and especially of adequate means of access, particu-
at the southeastern end by the environs of Teotitlan del
larly in the mountainous regions. After covering nearly
Camino, in the state of Oaxaca. The area so defined
12,000 kilometers in this region by jeep, we took stock
comprises the central part of a Cenozoic continental
of the gaps in our observations, and as a consequence,
basin formed from an arm of the Cretaceous sea. The
of the resulting defects in this study.
basin was rejuvenated during Quaternary time as the
The very long history that runs from the crystalliza-
result of its capture by the Rio Santo Domingo and thus
tion of the earth’s crust, some four and a half billion
became part of the watershed of the Rio Papaloapan,
years ago, until the beginning of Jurassic time is hid-
into which the Santo Domingo discharges. The Rio
den from view in the Tehuacan Valley. At best, one
—
Salado and its tributaries surface, subterranean, or
may group certain geologic structures which are clearly
artificial — drain this northwestern sector of the trough.
very old under the term “basal complex.”
is concerned not only with the valley
This study
We have studied this basal complex on the flanks of
but also with a more or less extended strip along
itself,
° Professor Jose Luis Lorenzo, Director of the Departamento
the shoulders of the bordering mountains and with the
de Prehistoria, Instituto Nacional de Antropologia e Historia,
adjoining Zapotitlan Valley, brieflv as the latter may Mexico, made it possible for me to join the Tehuacan Archae-
be treated. ological-Botanical Project under the auspices of the R. S. Pea-
In the field we sought to carry out essentially parallel body Foundation and led by Dr. Richard S. MacNeish. To these
two, as well as to the Peabody Foundation, I wish to express
traverses, perpendicular to the axis of the valley, that
my deepest gratitude. It is also a pleasant task to thank again
is, oriented approximately northeast-southwest. These those who helped me carry out the geologic studies of the valley,
traverses allowed us to gather rather quickly the sam- especially Ing. Arturo. Sotomayor, who kindly undertook a
number of petrographic studies, the results of which are pre-
ples and observations necessary to prepare the geo- sented at the end of this chapter. 1 am indebted also to certain
logical study. Certain details, especially those concern- earlier studies of the Tehuacan Valley, especially the memoir
ing the hydrogeology, were studied near the town of by Barrera (1946) and the guidebook for the excursion by the
Twentieth International Geological Congress (1956).
—
Tehuacan at the former hacienda known as “El This chapter has been translated from the French by M. T.
—
Riego” and near Ajalpan and Tilapa, all in the state Einaudi and Marland P. Billings.
66
GEOLOGIC STUDIES
Fig. 26. Panorama of theTehuacan Valley, looking northwest from km. 11 on the
Teotitlan-Huautla road, Other points are: b, Cerro Tepetroje; c, Nudo Mixteco Moun-
a.
tains; d, approximate location of Purron and Abejas caves; e, Cuesta de Altepexi; f, Sierra
de Zapotitlan; g, Cuesta de San Marcos; h, Cerro de la Mesa de Tuhuacan; i, hill and
hamlet called Chichiltepec.
the Sierra Mazateca and along the road from the town and this may be the same basal complex as that which
of Coxcatlan to San Pablo Zoquitlan 0 and in the Sierra occurs in the southwestern part of the Tehuacan Valley.
de Acultzingo, where it possibly represents the Paleo- The term “basal complex” is preferred here because
zoic or the beginning of the Mesozoic (Triassic). It may it includes numerous rock types rather than schists
go back a half billion years. It is principally composed alone, and because it better indicates our ignorance on
of —
metamorphic rocks schists, slates, sandstones with the subject. We should recall that the basal complex
calcareous cement showing traces of dynamic meta- is visible where erosion has exposed and even strongly
morphism (see note 1 under the Analysis of Hand attacked it — in the southern part of the Sierra de Acult-
Samples below), sandstones contaminated by volcanic zingo, in the Sierra Mazateca, and also some twenty
material (2), quartzite (3), gneiss, mica schists (4), and kilometers west of Zapotitlan Salinas, in the region of
marbles (5), as well. as rocks often extensively altered Santo Tomas Otlaltepec.
and intruded by volcanic dikes. The latter are par- This part of the Mesozoic is more easily studied in
ticularly evident along the road to Huautla de Jimenez, the region under investigation, for it is well exposed in
Oaxaca, in the Sierra Mazateca. the Sierra de Zapotitlan and in the northern part of the
This basal complex has been submitted to numerous Sierra de Acultzingo, to the east of Cerro Colorado de
tectonic and volcanic disturbances, of which folds, Tehuacan or Cerro de Los Cuarteles.
traces of dynamometamorphism, dikes, and sandstone de Zapotitlan, the northeast and east
In the Sierra
with lava debris are evidence. Although the basal com- boundary Mesozoic forms a convex
of that part of the
plex is made up of rocks which one would suspect were arc facing the town of Tehuacan. It passes through the
of marine origin, it is deformed and transformed to outskirts of the former hacienda of Cipiapa (near San
such a degree by metamorphism that one cannot un- Bartolo Teontepec) and crosses the highway between
ravel its Following Salas (1949) geologists
history. Huajuapan de Leon, Oaxaca, and Tehuacan eleven kil-
speak of the basal complex in the region of Santo To- ometers west of Tehuacan. It next includes the Sierra
mas Otlaltepec and Acatlan as the “Acatlan Schist,” de Miahuatepec and the Sierra de Atzingo and finally
° Unless otherwise indicated, towns mentioned in the text
joins the valley of the Rio Hondo, which here forms the
are in the state of Puebla. boundary between the states of Puebla and Oaxaca, at
67
Fig. 27. Cerro de la Mesa and surroundings seen from the Calvario de Coapan, Other
points are: a , Sierra de San, Bartolo; b, Cuesta del Yeso; c, highway between Tehuacan
and Huajuapan de Leon, Oaxaca; d, e, quarries in Cuesta del Yeso; /, Sierra de Teca-
machalco; g, gray soils and other Quarternary deposits; h, Cerro de la Mesa plateau; i,
small gully in Quaternary deposits; k, El Riego Cave; l, former hacienda El Riego; m,
canal from Valsequillo Reservoir; n, Tehuacan; o, Santa Maria Coapan.
about 18° 10' N. This topographic boundary corre- the Sierra de Santa Rosa. Both consist of sandstones
sponds to the end Middle Cretaceous and the be-
of the and conglomerates, but the Red Beds in general ex-
ginning of the Cenozoic. Although it is thought to be hibita more argillaceous facies than the Matzitzi
present in the Sierra de Atzingo, 0 the Upper Creta- formation.
ceous appears to be absent in this region. The Cretaceous is always clearly marine. In the
The invertebrate deposits of San Juan Raya, in the Lower Cretaceous, or Puebla Group, one finds, in nor-
interior of the Zapotitlan Valley, have for a long time mal stratigraphic order, the Zapotitlan formation, the
attracted the attention of geologists and stimulated the San Juan Raya formation, and then the Miahuatepec
study of the Mesozoic strata of the region. formation.
Let us brieflv recall the principal facies represented The Zapotitlan formation, marly on the whole, also
in the Middle and Tapper Cenozoic of the region, and includes limestones with rudistids and neritids. It is
their characteristics. assigned to the Barremian. Two members are distin-

The Jurassic is continental except in the northern guished —
the Agua del Cordero and the Agua del
part of the region under review. It includes two dis- Burro. The Zapotitlan formation is found in the entire
tinct formations between which the stratigraphic re- region surrounding the road from Tehuacan to San
lationships are not clear — the Matzitzi formation, Juan Raya, between the outskirts of San Antonio Tex-
described bv Aguilera (1906), to the south of Zapotit- cala and the place called Mata Calabaza on the far side
lan; and the “Red Beds’’ in the southern approaches of of Zapotitlan Salinas.’
°
The San Juan Raya formation, found to the west of
At the foot of the Sierra de Atzingo we collected a sample
of calcareous breccia (Lower Cenozoic). Micropaleontological the Zapotitlan formation as far as the outskirts of San
analysis, the results of which were transmitted to us by Dr. Nicolas Tepostitlan, is not as extensive as Aguilera
Jacques Butterlin, show that one of the elements in this breccia has been assigned to the Aptien.
(1906) believed. It Its
belongs to the Upper Cretaceous. This element surely came from
Sierra de Atzingo. An analysis of his sample appears under
lithology is similar to that of the preceding formation.
note 6 below. In the vicinity of the hamlet of San Juan Raya an abun-
68
Fig. 28. Alternating and folded limestones, marls, and gypsum of the Cuesta del Yeso, km. 7, Tehuacan-Huajuapan de
Leon highway.
Fig. 29. Extraction of marly limestone and gypsum from quarry at e in Fig. 27.
A
Fig. 30. Lagoonal sediments exposed in Cerro Agujereado. Coxcatlan Cave lies below the exposed face; the talus of exca-
vation shows the lateral extent of the shelter.
dance of about ten fossil species can be found (Cal- much more restricted than that of Aguilera. The for-
deron Garcia 1956; 70); Pterotrigonia plicatocostata mation is composed of limestones with chert nodules
(Nyst and Galeotti), Tylostoma aguilerai Alencaster, and dolomitized zones. They are rich in fossils rudis- —
Ncrinea sp., Neithea Pyrazus scalariformis Nagao,
sp., tids, foraminifera, algae, and so on —
and are reportedly
Corbis (
Sphaera ) corrugata Sowerby, Cerithium bus- lower Trinity- Albian (Applin 1955, Calderon Garcia
tamantii (Nyst and Isognomon lamberti
Galeotti), 1956) or upper Albian-Cenomanian (Viniegra and Lopez
(Mullerried), “Cyprina” sanzi Landerer, and Exogyra Rubio Calderon Garcia 1956).
in
acuticosta (Nyst and Galeotti). One also commonly sees In chronological order, beginning with the Upper
ostreids and corals. During one of our trips, a student Cretaceous and continuing into the Cenozoic, the evolu-
found a portion of what appears to be a crustacean tion of the Tehuacan basin was marked by the following
claw. This specimen is now being studied. essential episodes:
The Miahuatepec formation makes up the Sierra de 1. Principal orogenic stage and the beginning of the
Miahuatepec and the Sierra de Atzingo. It comprises isolation of the future Tehuacan Valley from the Gulf
limestones with black chert and marly beds and dis- of Mexico.
plays many folds. Geologists who have studied the re- 2. Deposition of lagoonal sediments rich in saline
gion hesitate to assign a precise age to this formation; material, in particular fibrous gypsum associated oc-
it is certain that it is intercalated between the preceding casionally with “lanCe-head” gypsum. These lagoonal
formations and the Cenozoic. sediments are well represented in the following re-
San Bartolo and surrounding heights
In the Sierra de gions: Cuesta del Yeso, between the slopes of the
—
the Middle Cretaceous Albian and Cenomanian is — Sierra de San Bartolo and the neighborhood of the
represented by the Cipiapa formation. The name is de- Sierra de Atzingo; Calvario of Sta. Maria Coapan; Cal-
rived from that of the former hacienda of Cipiapa, near vario of San Sebastian Zinacatepec; the vicinity of
San Bartolo Teontepec. It is presently used in a sense Coxcatlan; Pueblo Nuevo; San Jose Tilapa; and Teo-
70
GEOLOGIC STUDIES
Fig. 31. Cerro Colorado de Tehuacan.
titlan del Camino. They may also be seen along the cordant on the preceding ones and show perfect hori-
highway from Tehuacan to Veracruz, between the zontal stratification.
outskirts of Tehuacan and kilometer 260, close to Cha- Some fine cuts permit the description given in Fig.
pulco. 33,which lists the sediments from top to bottom. How
An approximately lateral facies consists of the “Red can one classify and date these assemblages? The
Conglomerates” derived from the extensive erosion of search for macrofossils gave no results, although it had
new mountain chains. From them are formed in par- been made earlier, especially during preparations for
ticular Cerro Colorado of Tehuacan and Cerro Colo- the Twentieth International Geological Congress in
rado of Ajalpan. 1956, and by us in the summer of 1964. We therefore
3. New orogenic movements marking the final iso- collected some samples with the object of searching for
lation of the valley from the sea and the close of la- microfossils. Analyses of some thin sections, results of
goonal deposition. which were furnished us by Dr. Jacques Butterlin, gave
4. Formation of a large lake whose sediments, most not the slightest chronological hint (see corresponding
often consisting of travertines, are well preserved today notes).
in the northwest part of the valley — in the vicinity of The red conglomerates correspond exactly with those
Tepanco de Lopez, Cerro de la Mesa, the cuestas of in the northern part of the state of Oaxaca. They be-
San Marcos and Altepexi and also perhaps in the vicin- long to the Huajuapan formation and are, at least in
ity of the village of Axusco. These sediments are dis- part, a lateral facies of the lagoonal-lacustrine sedi-
71
Fig. 32. Cerro Colorado de Ajalpan and the Calvario of San Sebastian Zinacatepec.
still further the extent of the old “Tehuacan formation”

Travertine (notes 7-11)
of Aguilera.
The remaining lacustrine travertine that is discord-
Alternating marls (12), ant on the saline sediments could then be called the
Cerro de Mesa formation, after Cerro de la Mesa,
la
limestones (13, 14), gypsum,
"Red Conglomerates” which behind
rises the former hacienda of El Riego.
conglomerate (15), some The Cerro de la Mesa formation would therefore rep-
resent the end of the continental Cenozoic in the region
chalcedony nodules
and perhaps the beginning of the Quaternary. The
stratigraphic sequence then becomes as shown in Fig.
Fig. 33. Correlation of Cenozoic lateral facies.
35. It is Tehuacan formation
possible that the is in part
later than the Huajuapan formation. That is, its upper

part doubtless corresponds to the Yanhuitlan beds of
ments mentioned above. These sediments are tradi- Salas (1949).,
tionally termed the Tehuacan formation (Aguilera The presence of vertebrate fossils has made it pos-
1906). However, one could reserve the latter name for sible to assign an Eocene-Oligocene age to the red
the part rich in saline material and in this way reduce conglomerates of the state of Guanajuato (Fries 1955).
72
GEOLOGIC STUDIES
Fig. 34. Cerro de la Mesa, viewed from the irrigation canal north of El Riego.
known for some time. These include, for example, the

andesites of San Bernardino lagoon near Santa Ana
Cerro de la Mesa Formation
Otzolotepec. A map published by L. Blasquez in 1957
shows a volcano in the region of the San Marcos hills,
to the south of the former hydroelectric station No. 2 of

Discordance
Tehuacan, but we have not found anything to corre-
Cenozoic /
spond with this. On the other hand, between Chilac
Tehuacan / Huajuapan
Formation / Formation and Atzingo we did find a volcanic complex (20, 21),
the extent of which could not be defined exactly. On the
Laramide Discordance other side of the valley, to the east of Ajalpan, we

briefly studied units — cataclastic rocks
lavas, (22), and
volcanic tuffs (23) —of a volcano that perhaps is fairly
important.
Mesozoic Lower and Middle Cretaceous
There are also some Cenozoic basaltic dikes (24),
visible inexposed cuts along the road between Tehua-
Fig. 35. Correlation of Mesozoic and Cenozoic formations. can and Huajuapan de Leon. One of these is shown in
Fig. 36. The cuts were visited by the Twentieth Inter-
One may assign the same age to the Huajuapan and national Geological Congress in 1956.
Tehuacan formations (Calderon Garcia 1956). The Formerly volcanic veins were found in the subsoil
Cerro de la Mesa formation would be approximately of the city ofTehuacan at depths of about 80 meters,
Mio-Pliocene in age. according to L. Blasquez. It is not possible to determine
In order to complete the stratigraphic data on the very much in regard to the age of the majority of these
Cenozoic, it should be pointed out that in the region volcanic units, but it is likely that they date from the
of Axusco and near Chilac a certain number of strata, second half of the Cenozoic.
whose correlations are not clear, ought definitely to be The end of the Cenozoic is not clearly marked in the
assigned to the Tehuacan formation, using the term in Tehuacan Valley, and it is likely that the lacustrine de-
its restricted sense. These include limestone (notes 16 posits of the Cerro de la Mesa formation extend some-
and 17), sandstones (18), breccias (19), and so on. what into early Quaternary times.
Vulcanism was of only limited importance in the
Tehuacan region, although it is undoubtedly respon- The Formation of the Present Valley
sible for a part of the mineralization of the underground The present characteristics of the Tehuacan Valley,
waters. Some volcanic units in the valley have been and particularly the fact that it is today a valley and not
73
—
Fig. 36. Basaltic dike exposed at 10.1km. on the Tehuacan-Huajuapan highway. This
dike is possiblymore recent than the one cited in note 24. a, Rocks of Zapotitlan forma-
tion; b, d, highway and embankment; c, basaltic dike with calcitic veins and marks of
alteration; e, limestones and marls of Tehuacan formation.
an interior basin, can be attributed to an event that Rio Salado, Rio Tehuacan, Rio Hondo, Rio de Parian,
probably took place during the early Quaternary. Rio Grande de Cuicatlan, arid their numerous tribu-
Headward erosion by the Rio Santo Domingo, an afflu- taries. Before the capture of the basin, these streams
ent of the Rio Papaloapan, resulted in the capture by were shorter, independent of each other, and dis-
the Santo Domingo of the lake in which the sediments charged into the lake.
of the Cerro de la Mesa formation were deposited. This Such was the hydrography at the time of the capture
probably occurred near the present site of Quiotepec, of the ancient lake. Some interesting comments about
Oaxaca. this capture can be found in Barrera, who refers to it
The old sedimentary basin was approximately as “a case of piracy” by the Rio Papaloapan (1946: 24
bounded by the following points, starting from the and fig. 10). He goes on to explain how the headward
north and turning clockwise: Ciudad Serdan, Atzin- erosion of the Rio Papaloapan — actually its tributary,
zintla, Cumbres de Acultzingo, Santa Maria del
the the Santo Domingo — connected the old, closed basin
Monte, San Pablo Zoquitlan, and Coyomeapan, all in with the lower valley of the Papaloapan, thus bringing
the state of Puebla; then Teopoxco, Concepcion Pa- about the drainage of the waters of the lake.
palo, Aloapan, San Jeronimo Sosola, Coixtlahuaca, and One must emphasize the extreme importance of this
[icotlan, in the state of Oaxaca; and finally, once again capture to the geomorphological evolution of the
in the state of Puebla, Acatepec, Atenayuca, Xochitlan, region. —
The old lacustrine system stable in equilib-
the environs of Tecamachalco, and Quecholac. The lake rium and functioning, one might say, as a closed system
thus bounded extended on both sides of a line drawn — was abruptly transformed into an area of intense ero-
approximately through the following points: San sion. It was an area with two heads, comprising a north-
Gabriel Tetzoyocan, San Miguel Zozutla, Tepanco de western sector, drained toward the southeast by the
Lopez, Tehuacan, and the course of the Rio Salado into Rio Salado, and a south-southeastern part, drained to-
Oaxaca, thence through Quiotepec and the course of ward the north-northwest by the Rio Grande de Cui-
the Rio Grande de Cuicatlan to the region of the can- catlan. Such an arrangement, or “T-shaped drainage,”
yon of the Tomellin. The slopes of the surrounding is fairly characteristic of captured drainage patterns
mountains were drained by rivers in existence todav (see Fig 37). The lacustrine basin, situated some 1700
74
GEOLOGIC STUDIES
a, Sierra de Zapotitlan; b, Tehuacan; c, Orizaba; d, Cordoba; e, Zongoliea; /, Sierra de

Zongolica; g, Teotitlan del Camino; h, extent of former closed lake in Tehuacan-Cuicatlan
basin; i, Quiotepec; /, Cuicatlan; k, Huautla de Jimenez; l, site of capture of former lake
by the Rio Santo Domingo; m, Presa Presidente Miguel Aleman; n, Gulf of Mexico;
o,Alvarado, Vera Cruz.
meters above the sea and less than 200 kilometers from net result is that the valley becomes drier and even
emptied itself of its watery resources in two comple-
it, desert-like. Concerning this last point we may make
mentary processes. First, the lake disappeared, and one observation: there is no reason to believe that the
with it the base-level of the streams that fed it. Then climate as a general phenomenon, could have changed
these streams, finding slopes and indices of erosion that much during the last stages of the Quaternary. How-
are characteristic of torrents, increased the rapidity ever, the microclimates of the valley have progressed
with which water was drained from the neighboring toward ever more marked aridity because of the disap-
mountains. pearance of the water.
The consequences of all this are marked. Reserves of Quaternary sediments in the former basin can be
water in the surrounding mountains have decreased classified under three headings:
with some delay, but steadily, causing the hydrostatic 1. Encrusted material deposited in different parts of
base-level to sink lower as time passes. We shall dis- the valley by springs rich in minerals — silica, carbon-
cuss this point in detail in the section below dealing ates, nitrates, chlorides, sulfates, and so on.These are
with the district of the former hacienda of El Riego. The found along the edge of the Cerro de la Mesa between
75
Fig. 38. Section approximately WSW-ENE between Sierra de San Bartolo, a, and San
Antonio Las Canadas, h. Other points are: b, Cuesta del Yeso; c, Cerro de la Mesa; d,
former hacienda El Riego; e, Tehuacan; f, Rio Salado; g, Cerro Colorado de Tehuacan;
i, encrusted deposits along face of Cerro de la Mesa; j, gray soils and other recent deposits;
k, alluvium; l, “basal complex”; m, Jurassic levels; n, Zapotitlan, San Juan Raya, and similar
formations; o, Cipiapa formation; p, basaltic dike; q, Tehuacan formation; r, Cerro de la

Mesa formations; s, “Red Conglomerates” of Huajuapan formation; t, andesitic extrusion.
Fig. 39. Encrusted travertine near El Riego.
76
GEOLOGIC STUDIES
Fig. 40. Detail of encrusted travertine showing impressions of plants.
San Lorenzo Teotipilco and El Riego. The alignment 3. Slope deposits occur particularly at the foot of
of the springs or manantiales in this region cannot be Cerro Colorado de Tehuacan, Cerro Colorado de Ajal-
attributed to a fault, as has sometimes been thought, pan, and the ancient massif formed by the basal com-
but to the general direction of the erosion front of the plex in the southeastern part of the valley. It appears
travertine of the Cerro de la Mesa formation. Further that these deposits are accumulated at the foot of the
deposits of encrusted material are found along the sides Sierra de Zongolica, a circumstance that is explained
of the cuestas of San .Marcos and Altepexi, near Cerro by the fact that this side of the valley has steeper and
Tepetroje (25, 26), and Cerro Pelon (27, 28). Also of drier slopes and erosion is therefore unable to remove
interest is a geyser-like spring on the summit of the hill the debris that is constantly forming.
called Petlanco which is depositing chlorites and sul- Phenomena of “present-day geology” are an inter-
fates, and to whose waters medicinal properties are esting study in the Tehuacan Valley, especially in con-
attributed. nection with relations between erosion and sedimenta-
2. Alluvium —
clays, sands, gravels —
scattered be- tion on the one hand and hydrologic equilibria on the
tween the environs of San Diego Chalma and the low other. We will return to this matter in the following sec-
part of the valley. In the following section we shall see One could give great importance to karst phe-
tion.
that this alluvium is undergoing a somewhat unusual nomena in this valley, but we do not believe that they
—
geomorphologieal evolution an “internal filling.” This merit being considered the key to geomorphologieal
alluvium is often covered with a calcareous crust or problems, because, in general, near Tehuacan calcare-
“caliche” which greatly reduces its fertility. We prob- ous zones are altered and then carried away by erosion
ably should not draw any climatic inferences from the without undergoing the “evolution in place” that is
presence of this caliche: it simply results from the characteristic of karst phenomena. In particular, clays
supersaturation with carbonates of the capillary and formed by decalcification are totally lacking in this
circulating waters of a large part of the valley. region.
IERRA
;-SAN BARTOLO
V^LA MESA
X, *T 184 - JS LORENZO
S. JUAN^RAYA
C CHACATECA^ TEHUACAN
: T 04
i*KcUESTA ^COAPAN :
C COLMENA :
/
Ao
;
S ATEXCALA/f
'
E l4 / W /CERRO* COLORADO
" AN
DE/TEHUAC
DE y.A/piSi'V'-ii;,
.YESO.Vm *T 173 S. ANTONIO CANADA
ZAPOTITLAN
ZAPOTITLAN
i',
v
A/\Aa/\A/\A
,
^ '' y* /va "<
l!AAAA/W\A/J
:-\/\/v\/\AA/'i
y
;7—
iT 062*«T.-I667^'^ J'
XA
T 064* / VT25,
.
/
A/NA/\ AA
.•VX
i,/\/\/\/\AA /
/X^X /\ NVN
A/''/>J
T A,';/\/\A/\AAA/V/
“• i\/\A/\/\AAA/
T 076:9*^
= /
^^<^^©CHILAC ALTEPEJI X 'AA/VNAA/V/V/

••'Vw V\/sA/\/\A/V (
AJALPAN'/'X/'X^x/x/vw/ I
ATZINGO p *T07I-2X>C?/
.
/ ,V\/\/N/S/S/V/
/
£/ >. ./\/N/\/N/N.' /
,
KILOMETERS
S MATEO
MIAHUATLAN
‘
o *0*
C O LO R A* D*0\ V*
CERR *
r
/DE ‘AJALPAN .*** £/
ZINACATEPEC
THE VALLEY OF TEHUACAN

TENTATIVE GEOLOGICAL MAP
~~~n CONTOUR INTERVAL 200 M

.CALIPAM,
HYDROGRAPHY. F^=q ROADS.
*T 000 ORIGIN AND N£ OF A HAND SAMPLE.
:
3G ARCHAEOLOGICAL SITES.
==TI37# " AJUSCO
/
!© EL RIEGO. CHILAC AND MEQUITONGO SITES
. CERRO
=
lCOXCA-T-LAN
3D FOSSILS. CZZI FAULTS. TEPETROJE
C DE LA MESA F -CENOZOIC OR QUATERNARY

OTHER QUATERNARY DEPOSITS CHICHILTEPEC
TEHUACAN F.
HUAJUAPAN F.
rm BASAL COMPLEX MEQUITONGO

CIPIAPA F.
MIAHUATEPEC F. q E7T71 EXTRUSIVE TILAPA

S. JUAN RAYA F. U INTRUSIVE
r~~~l ZAPOTITLAN
MAZATECA
INDET. CRETAC. -
F. = FORMATION *TEOTITLAN DEL^C AMINO.
BRUNET
Fig. 41. Tentative geologic map of the Tehuacan Valley.

78
1 CERRO DE LA MESA 13 SIERRA DE ATZINGO
2 TEHUACAN 14 CERRO TEPETROJE
3 CERRO COLORADO DE TEHUACAN 15 AXUSCO
4 SAN ANTONIO LAS CANADAS 16 RIO HONDO
5 SANTA MARIA COAPAN 17 LOS CUES
6 CUESTA DEL YESO & 18 SIERRA MAZATECA
CALVARIO DE COAPAN 19 TEOTITLAN DEL CAMINO
7 TEXCALA 20 TILAPA
8 CUESTA DE ALTEPEXI 21 COXCATLAN
9 AJALPAN 22 CALIPAN
10 ZAPOTITLAN 23 SAN SEBASTIAN ZINACATEPEC
11 ALTEPEXI 24 SAN PABLO ZOQUITLAN
12 CHILAC 25 RIO ZAPOTITLAN
26 RIO SALADO
A : CENOZOIC & QUATERNARY

B : BASAL COMPLEX
C CERRO DE LA MESA
: F.
H HUAJUAPAN F.
;
M: MESOZOIC
Q QUATERNARY
;
T TEHUACAN
;
F.
BLOCK-DIAGRAM OF THE VALLEY OF TEHUACAN
Fig. 42. Block diagram to show relationship of topographic features.
79
Fig. 43. Section at El Riego Cave, showing geologic and irrigation features: a, Cerro de la
Mesa plateau; b, stratified travertine; c, encrusted travertine; d, front of erosion at the end
of the rejuvenation of this part of the valley; e, El Riego Cave; f,
modern underground
water sources; g, h, i, wells of “El Huizache” network; a prehispanic canal; k, mass of
fallen rock and earth; l, drain running from San Lorenzo sources to a lower part of the
valley; m, former factory of mineral water; n, collapsed subterranean gallery; o, modern
gallery; p, rubbish; q, gray soils and other recent deposits.
The Quaternary fauna, we note, includes Parelephas of the sugar-cane plantations at Calipan in the lower
—
columbi Falconer near San Jeronimo (Aguilera 1906) valley. Soils are, in general, gray soils mixed with
— Parelephas columbi var. felicis Freudenberg be- — gravel from the alluvium.
tween El Riego and the city of Tehuacan (Miillerried Volcanic and tectonic activity do not seem to have
1931) —
and some unidentifiable remains of a probosci- affected the region during the Quaternary'. Only small
dian found during the summer of 1964 at the bottom of tectonic readjustments are to be noted in the saline
a well near Altepexi. beds of the Cuesta del Yeso; this phenomenon is well ex-
Initiation of the program of the Papaloapan Commis- posed along the road from Tehuacan to Huajuapan de
sion of the Secretaria de Recursos Hidraulicos has Leon between approximately kilometer 5.5 and kilo-
partly offset the harmful effects of erosion and drought. meter 10.5.
This program benefits Tecamachalco and Tehuacan Fig. 41 is a generalized map showing the principal
with waters of the Rio Atoyac Poblano impounded by geologic structures in the Tehuacan region. In order to
the Valsequillo Dam. The valley of the Atoyac has a make it more readily understandable we insert a block
history that parallels that of the Tehuacan Valley, hav- diagram (Fig. 42) prepared from the map and making
ing been an ancient lake that was captured. The lower use of all available topographic and altimetric data.
portions of theTehuacan Valley, in turn, benefit from We see there, inserted among the ancient structures be-
waters gathered in the upper part. Rut in any event longing to the basal complex and the Mesozoic, lagoonal
soils are poorly developed, except perhaps in the region and lacustrine beds of the continental Cenozoic, fol-
80
GEOLOGIC STUDIES
Fig. 44. El Riego Cave. The lush growth along the foot of La Mesa is characteristic, as are the epiphytes growing in the
trees and the xerophvtic plants on the talus.
lowed by beds that represent the end of the Cenozoic sible owing to a series of observations in the immediate
—
and the Quaternary travertine, alluvium, slope de- vicinity of El Riego rock shelter and near the former
posits, and so on. hacienda of the same name. These observations will be
given below. The curve traced from these observations
Hydrological Evolution indicates, in addition, the approximate time during
Environs of El Riego
which pre-Hispanic irrigation canals, preserved by a
The preceding paragraphs have emphasized the im- calcareous travertine coating, were in use. We were also
portant role of the rejuvenation of the valley in acceler- able to determine the sequence in which a certain
ating the erosion that revives and dissects the lagoonal number of canals were used.
and lacustrine structures of the ancient basin of Tehua- The cross section (Fig. 43) represents from west to
can. The lowering of the water table in the valley and east and from top to bottom the following features: the
in the surrounding mountains is a result of this phe- El Riego rock shelter 22.5 meters above the present
nomenon, and it also follows in a precise manner a water table; a certain number of old springs, traces of
mathematical law of acceleration. which are preserved by encrusted travertine; pre-His-
Good fortune permitted a study of the lowering of panic canals “fossilized” by the same travertine; some
the water table as time has gone by, not only in the shallow, collapsed subterranean galleries; and finally,
valley, —
but on its flanks as well in the valley thanks modern galleries about 10 meters deep, used to conduct
to an old well-digger born in San Diego Chalma, who water obtained from underground sources to the sur-
in the course of his seventy years saw the water of face. One of these galleries today supplies mineral wa-
flowing springs slowly dry up, to be found thereafter ter to the El Riego bottling plant. Although the method
only in wells of greater and greater depth, ranging to of constructing the older galleries is similar to that used
10 or 12 meters. On the valley slopes analysis was pos- for modem “filtering” galleries, in the older galleries
81
level). The curve shown in Fig. 46 confirms that the

successively higher levels at which water was ob-
tained represent levels of increasingly greater age. This
makes it possible to date pre-Hispanic canals with a
wholly relative precision. Those radiating from El Rie-

go were in use between a.d. 1000 and the Spanish con-
quest and had maximum use about 1200. This dating
applies only to canals and galleries, both “fossil” and
modem, that radiate directly from El Riego. Other
irrigation systems, originating at San Lorenzo Teoti-
pilco, cross the districtaround El Riego and run about
ten meters above the galleries we have discussed.
It follows from these remarks that when one is
—
studying a fascicle of canals a fan-shaped assemblage
—
originating from a single spring the oldest are the
highest, and inversely, those at the lowest level are the
most recent. The same “geochronological” method
could be applied with fruitful results to the regions of
the cuestas of San Marcos and Altepexi.
The Vicinity of Ajalpan
One may say that erosion progresses by moving up

the drainage basin. In the Tehuacan Valley this means
that shortly after the capture of the basin by the Rio
Papaloapan system, erosion was very active in the
Fig. 45. The subterranean gallery noted at n in Fig. 43.
lower parts of the valley, around San Juan de los Cues
and in the vicinity of Teotitlan del Camino. Later
erosion affected more strongly the regions around the
we are dealing with structures intended to carry water towns of Venta Salada and Coxcatlan, then Ajalpan and
from the foot of the mountain to lower parts of the Chilac. Today the strongest evidence of erosion is seen
valley and not gather filtered water along the way. It in the vicinity of Tehuacan, especially in the zone of
is for this reason that we speak of underground gal- —
the “Red Conglomerates” Cerro Colorado de Tehua-
leries; they are veritable underground aqueducts. can. One notes there the presence of multiple gullies,
In the same order, here are the corresponding chron- or canadas —
la Canada de San Antonio, for instance.
ological data. About 7000 years ago a veritable water- To the north, in the vicinity of Tepanco de Lopez, ero-
fall left travertine deposits in El Riego Cave, a shelter sion is weaker, but one may predict its extension in the
that must have remained very damp up to some 4000 future.
years ago. Various old texts speak of “trenches” built The predominantly sedimentary material tom from
Conquest or somewhat
for irrigation at the time of the regions strongly attacked by erosive activity is not
later, a word that suggests an open ditch with a depth necessarily carried as far as the lower basin of the Rio
of about a meter. At the beginning of the twentieth Papaloapan, nor for that matter into the waters of the
century, the water galleries were at a depth of “a few Gulf of Mexico. After being carried a relatively short
meters,” according to the master well-digger to whom distance, it is dropped in the intermediate sections of
we have referred. These are the abandoned subter- the valley. For several thousand years such deposition
ranean galleries with depths of three to five meters has been most actively building up in the region around
previously mentioned. Ajalpan. Thus the intermediate valley is filled in pro-
In order to define precisely the age of the “modern” portion to the rate of erosion in the upper reaches. It
galleries, we note that a network called “El Huizache” is clear that in turn this filling of the intermediate valley
was dug in 1964. In this porous limestone region the is attacked by erosion and is extensively gullied —
presence of water (hydric level) is closely related to the phenomenon clearly seen in the immediate outskirts of
upper limit of the “zone of cementation” (hydrostatic Ajalpan.
82
GEOLOGIC STUDIES
Fig. 46. Diagram Riego Cave, a. Water

to date prehispanic canals, c, in the vicinity of El
levels in meters on arithmetic between 8000 b.c. and a.d. 1964 expressed in
scale; time
logarithmic scale, b, function expressing correlation between water table and time; d,
trenches built for irrigation about 1519; e, old collapsed galleries; f, modern galleries;
g, hypothetical function for the abandoned canal from San Lorenzo, passing east of El
Riego; h, actual situation of this drain.
When it is difficult, then, to divorce erosion and sedi- One might propose the expression “intermediate al-
mentation, it becomes possible to understand the luviation,” or “intermediate filling,” or “successive allu-
phenomenon in its entirety. The disequilibrium created viation.” If this intermediate filling began in the
by the capture of the ancient basin by a tributary of the Pleistocene —the finding in a well near Altepexi of
Rio Papaloapan set in motion a complex phenomenon fragments of bone from an extinct proboscidian prob-
affecting the evolution of relief. This is expressed by a ably dating from the upper Pleistocene suggests that it
shift of erosive activity toward the northwestern part did indeed begin then — it has developed fully in the
of the valley. After erosion has cleared part of the val- region about Ajalpan during relatively recent times. If
ley, it is succeeded by sedimentation that fills that part we take into account the stratified deposit of numerous
of the valley already cleared. The very intensity of this archaeological remains in the sediments, the filling of
sedimentation brings about a new disequilibrium of the area must have involved a period of some thou-
the system, and a new phase of erosion is initiated. sands of years. The sediments consist primarily of clays
Neither the term “alluvial cone” nor “terrace” is well and gravels, whichalso are more or less finely stratified.
suited to describe the land built up around Ajalpan. It is obvious that the archaeological evidence falls into
83
Fig. 47. Section across the Tehuacan Valley through Zapotitlan Salinas, a, and Ajalpan,
d. Other points are: h. Sierra de Miahuatepec; c, Cuesta de Altepexi; e, Cerro Colorado
de Ajalpan; f, Rio Comulco; g, “intermediate filling” of Ajalpan area; h, Rio Salado; i,
Calvario de Zinacatepec; j, Jurassic levels; k, Zapotitlan formation; l, Miahuatepec forma-
tion; m, Tehuacan formation; n, Cerro de la Mesa formation; o, Huajuapan formation;
p, igneous rocks.
a chronological succession in the order in which the from Tehuacan to Teotitlan del Camino and the rock
specimens appear in the stratigraphic cut under review, shelters called Purron and Abejas. The dams were
the deepest objects being the oldest. One must except built in a geological environment characterized mostly
burials made in graves dug during periods of tempo- by Lower Cenozoic continental gypsiferous beds of
rary dryness in one zone or another, a situation in quite regularly alternating sandstone, marl, marly lime-
which dating can only be done by determining the level stone, and fibrous gypsum —
all rocks that are quite
of the mouth of a grave shaft. easily worked and that were used for building the
The same remarks apply to the Rio Zapotitlan. The dams and adjoining structures. The bottom of the
archaeologicallv interesting zone of intermediate al- ancient impounded lake, now almost invisible because
luviation extends southwestward of the village of of alluvium, was without doubt formed by the ancient
Chilac. Some remains occurred in an exposure along the terraces of the stream system that cut the Arroyo
bank of the Rio Zapotitlan exactly one kilometer south Lencho Diego. The collection basin took form in the
of the village square where a human mandible asso- extreme southern part of the Sierra de Atzingo (Cerro
ciated with two bone points and some sherds of pottery —
Chiehiltepec; Lower and Middle? Cretaceous and —
came to light. These remains date from approximately basal complex).
the beginning of the present era. As for the dams themselves, and without wishing to
It is now possible to pick out some terraces in the infringe on the domain of our colleagues Neeley and
immediate outskirts of Chilac. It is very easy to identify
7
Woodbury, we would like to propose here some ideas
them in the arroyo of San Marcos, where three terraces that set forth a method or “geological procedure” for
can be separated on morphological grounds and dated the study of ancient artificial dams. We here consider
by means of their archaeological content. as artificial dams all those more-or-less massive, man-
made structures intended to retain, at least temporarily,
Meqaitongo: The Pre-Hi.spanic Dams an appreciable bodv of water, and. by the same token,
The dams of Mequitongo, near San Jose Tilapa, are the modification of the natural relief to the extent that
in the Arroyo Lencho Diego between the highway it may be necessary for the fulfillment of this end.
84
GEOLOGIC STUDIES
Fig. 48.Chilac and surroundings: a, Miahuatepec formation; b, Tehuacan formation;

c, Quaternary deposits; d, igneous rocks; e, San Juan Atzingo; f, San Gabriel Chilac; g,
human mandible found April 1964; h, San Mateo Tlacoxcalco.
Some basic principles must be admitted, in particular history of any given region. This premise makes it
the following: the pool above a dam begins to fill with much easier to interpret past events, for it allows us to
sediments practically from the moment of its construc- consider the mass of a dam as a lateral facies of its own
tion. Consequently, in the case of superposed dams, alluvium and vice versa. Another principle is this: the
alluvial deposits that correspond horizontally with each total capacity of the pool impounded by a stage in the
stage of the structure are each contemporaneous with construction of a dam is represented by the correspond-
the stage to which they correspond. In effect, if the ing body of alluvium.
construction of a dam requires one or several years, Finally, the localization of zones of heaviest deposi-
the length of time required for the sedimentation of tion of gravel hasno connection with such concepts as
the basin will be measured in the majority of cases in upstream and downstream, at least in the operation and
some tens of years. This is especially true where erosion scope of a dam, but depends primarily on water turbu-
is very active, asmost of the valleys of the Mexican
in lence. Indeed, gravels accumulate where the flow is
plateau and connected regions, such as the Sierra forced into a confining passage, as in sluice-gates,
Madre de Oaxaca and the Sierra Madre del Sur. arches of bridges, narrow canals, and so forth.
Geologically and even archaeologically speaking, such We hope that these concepts, already discussed in
intervals usually elude our analysis and may be con- the field with the project archaeologists, may have
sidered to be units too small to measure in the past helped in their reconstruction of the history of Mequi-
85
86
GEOLOGIC STUDIES
tongo. Here, briefly, is my own interpretation of that (23 km. from the former, 5 km. from the latter).
history: (1) natural primitive situation, (2) first dam Macroscopic description: clastic rock with light gray
and contemporary structures, (3) second and third dams color. Texture: clastic psammitic. Detrital compo-
(separated by a mud flow), (4) fourth dam, (5) aban- nents: quartz, feldspars, calcite, micas. Cement: cal-
cite. Classification: Comments:
calcareous sandstone.
donment of the region, (6) return of people during
the rock shows evidence of dynamic metamorphism.
Postclassic time, (7) present situation. This inteqireta-
2. Specimen T087, section no. 452. By A. Sotomayor. Field
tion is illustrated in Fig. 49.
notes: sandstone; road from Coxcatlan to Zoquitlan
(18 km. from the former, 10 km. from the latter).
In the long geological history of the Tehuacan Val- Macroscopic description: clastic rock, friable, rose
ley, the most significant event is the capture of the old, color. Texture: clastic psammitic. Detrital compo-
Canada Poblana-
closed Tehuacan-Cuicatlan basin (La nents: quartz, silicified material, feldspars, some fer-
Oaxaquena) by the Rio Santo Domingo, a capture that romagnesian minerals. Cement: kaolinitic clays,
aqueous reserves of the region
set in train the loss of the sandstone con-
chlorite, oxides. Classification: quartz
and completely changed its appearance. The valley, taminated by volcanic material. Comments: the rock
is not fully developed.
which became drier and drier, nevertheless allowed
3. Specimen T092, section no. 453. By A. Sotomavor. Field
man to establish himself there, to adapt himself pro-
1. notes: sandstone; roadfrom Coxcatlan to Zoquitlan
gressively until he was able to take advantage of its
(11 km. from the former, 17 km. from the latter).
scant water resources, thanks to his ingenuity. It is a
Macroscopic description: white, coffee, and black
paradox that this same aridity has preserved for archae- color. Texture: granoblastic throughout. Minerals:
vanquished civilizations of
ologists the details of the quartz, hematite, chlorite, limonite. Classification:
the Tehuacan Valley. Without it, we would never have quartzite. Origin: metamorphic rock.
been aware of them. 4. Specimen T094, section no. 454. By A. Sotomayor. Field
notes: chlorite schist; road from Coxcatlan to Zo-
quitlan (8.5 km. from the former, 19.5 km. from the
Notes on Hand Samples latter). Macroscopic description: gray and reddish
brown color, fohated. Texture: granoblastic in zones.

Specimen T083, section no. 451 By A. Sotomayor. Field Minerals: quartz, magnetite, hematite, chlorite, seri-
notes: sandstone; road from Coxcatlan to Zoquitlan cite, hmonite, graphite. Classification: zoned quartz-
5.
6.
Fig. 49. Dams of Mequitongo.
7.
1. Natural situation: a, Cerro Lencho Diego; b, to Purron and Abejas caves; c, Arroyo
Lencho
8. Diego; d, talus slope.
2. First dam: a, Arroyo Lencho Diego; b, dam; c, probable sluice gate; d, impounded
lake; e, hypothetical monuments preceding accessory dam of next period; No. 450,
monument.
3. Second and third dams: No. 435, principal dam; No. 15, accessorv dam and associ-
ated monuments; No. 67, monument probably near important sluice gate; Nos. 449-453,
contemporary monuments; a, gravels deposited in irrigation canals; b, probable sluice
gates; c, wells or heaps of stones exposed bv erosion; d, flood gates, probablv spanned by
wooden bridges, connecting the two basins separated by dam No. 15; e, gravels deposited
by turbulent water of flood gate or drain.
4. Fourth dam: No. 435, principal dam; a, Arroyo Lencho Diego; b, sluice gate; c, wells
exposed by erosion; d, structures of preceding period become islands or peninsulas in

I water behind dam.
Abandonment in Classic period: a, present streams established; b, natural drains of
the dam; c, remaining bog.
Rebuilding in Postclassic period: No. 435, pyramids and parvis built on former dam;
a, encampment; b, bog, or possibly a cultivated zone.
No. 435, four superimposed dams with later pyramids and
Present situation:
parvis; a, cart track; b, Arroyo Lencho Diego; c, superimposed alluviums dissected by
erosion; Nos. 15, 67, 449-453, well-preserved structures.
( Schematic section through the dams: a, Cerro Lencho Diego; b, d, eroded gullies;
c, Postclassic buildings; e, talus;
f, g,
h, i, superimposed dams.
crystalline limestone, fine-grained, with recrystallized

micro-veins, considerable iron oxides, and rare
foraminifera fragments; (c) medium-grained, micro-
crystalline limestone with granoblastic texture;
without fauna; ( d ) fine-grained, microcrystalline
limestone with numerous thin, recrystallized vein-
lets, feldspars, iron oxides; numerous chambers of
globigerina and other foraminifera, in particular a
well-conserved specimen of Rotalipora sp. and also
P. trejoi and Calcisphaerula innominata. Comments:
the presence of these fossils allows one to assign an
Upper Cretaceous age only to the fragments that
contain them, while the breccia itself may be of a
later age. The appearance of the rock suggests the
continental “red conglomerates” of the Lower Ceno-
zoic that are very common in many areas of Mexico.
7. Specimen T001, section no. 64-2324. Bv Dr. Federico
Bonet and Enrique Riva Palacio. Field notes: hori-
zontal beds of travertine; Cerro de la Mesa, 200 m.
west of El Riego spring, near Tehuacan. Description:
cream-coffee-colored marly limestone, partially re-
crystallized and extremely silicified. In spite of dia-
genesis there are biogenic remains that probably cor-
respond to mollusks, bryozoans, and algae; lacking
characteristic fauna.
8. Specimen T003, section no. 64-2325. Bv Dr. Federico
Bonet and Enrique Riva Palacio. Field notes: hori-
zontal beds of travertine; Cerro de la Mesa, 200 m.
west of El Riego spring. Description: fine-grained,
argillaceous, pseudo-colitic limestone of light-coffee
color; matrix contains small amounts of finely micro-
crystalline limestone, with clastic constituents (small
Fig. 50. Dams Mequitongo: canal or sluice gate exposed

at
and medium-sized) of quartz, potassic, and sodic
in bank of arrovo at left-hand structure No. 15 in Fig. 49, 7.
feldspars, subangular and irregular; parts filled with
Gravel fill at bottom is clearly discernible.

granoblastic spathic calcite and rare mica flakes; the
major part of the pseudo-colites are partially silicified.
Some mollusk remains, a poorlv preserved specimen
ite.Origin: metamorphic rock, formed from a sedi- of Globochaete alpina (?), and other biogenic remains
mentary rock. too poorly preserved to be identified.
5. Specimen T099, section no. 455. Bv A. Sotomavor. Field 9. Specimen T184, section Bv A. Sotomavor. Field
no. 460.
notes: marble; road from Coxcatlan to Zoquitlan (5.5 de la Mesa, 3 km. west of El
notes: travertine, Cerro
km. from the former, 22.5 km. from the latter). Mac- Riego spring. Macroscopic description: coffee-colored
roscopic description: massive rock, compact, light and dark gray rock, with zonation. Texture: crypto-
green color. Texture: granoblastic. Minerals: calcite, crystalline. Constituents: calcite, clavs, phosphates;
quartz, amphiboles. Classification: marble. Origin: silicified zones (with chalcedony, quartz, and opal).
metamorphic rock. Classification: limestone with silicified zones. Origin:
6. Specimen T064, section no. 64-2332. Bv Dr. Federico sedimentary rock.
Bonet and Enrique Riva Palacio. Field notes: cal- 10. Specimen T152, section no. 458. Bv A. Sotomavor. Field
careous breccia; barranca on first stream met after notes: lacustrine limestone; road from Puebla to Te-
crossing the Rio Zapotitlan (or Rio Chilac) on the huacan, near Tepanco de Lopez. Macroscopic de-
road from Chilac to Atzingo. Description: red calcar- scription: red-colored rock with veins of hematite and
eous breccia; consisting of a matrix of silicified lime- calcite. Texture: crystalline granular. Minerals: cal-
stone with abundant iron oxides and foraminifera cite, secondary calcite, hematite, quartz, feldspars,
remains. This matrix encloses the following constitu- limonite, clavs. Classification: limestone. Origin: sedi-
ents (up to 4 cm. in diameter): (a) limestone with mentary rock.
foraminifera, small amounts of microcrystalline ma- 11. Specimen T173, section no. 64-2340. Bv Dr. Federico
terial and many globigerina chambers; (
b ) micro- Bonet and Enrique Riva Palacio. Field notes: lacus-
88
GEOLOGIC STUDIES
Fig. 51. Dams at Mequitongo: top of dam can be seen at right but is bidden in brush at left. The truck is at the end of dam
near No. 450 in Fig. 49, 7.
trine limestone; San Marcos hill, about 5 km. south 14. Specimen T153, section no. 64-2337. Bv Dr. Federico
of Tehuacan (6 km. on the old direct Tehuacan- Bonet and Enrique Riva Palacio. Field notes: lime-
Chilac road). Description: coffee-colored, medium- stone, 2 km. south of Tehuacan, foot of Cerro Colo-
grained, microcrystalline limestone, slightly argillace- rado (Tehuacan viejo) at the level of the terraces of
ous with irregular grains of finer, mierocrvstalline Rio Salado. Perhaps of more recent age than T033
limestone; lacking fauna. and T166. Description: coffee-colored, fine-grained
12. Specimen T033, section no. 64-2326. Bv Dr. Federico mierocrvstalline limestone, with veins full of spathic
Bonet and Enrique Riva Palacio. Field notes: marl, crystalline calcite, slightlv argillaceous. Numerous
continental Lower Cenozoic; cut located at km. 8 on calcified radiolaria: Glohigerina sp. and Nodosaria
the Tehuacan-Huajuapan de Leon road. Description: sp. Lacks characteristic microfauna.
calcareous marl, coffee-ocher color, with very small, 15. Specimen T039, section no. 64-2327. Bv Dr. Federico
irregular,and subangular clastic quartz fragments, Bonet and Enrique Riva Palacio. Field notes: recrys-
and with numerous calcite grains, some of which are tallized conglomeritic limestone; cut located at km. 8
rhombohedral; iron oxides. Lacking in microfauna. on Tehuacan-Huajuapan de Leon road. Description:
13. Specimen T116, section no. 64-2339. Bv Dr. Federico medium- and fine-grained gray mierocrvstalline lime-
Bonet and Enrique Riva Palacio. Field notes: limestone, slightlv argillaceous, with iron oxides; certain
stone; road from San Gabriel Chilac to San Juan parts are strongly recrvstallized. Remains of mollusks,
Atzingo, 100 m. after crossing Rio Zapotitlan (Chi- echinoderms, few ostracods, recrvstallized algae.
lac). Description: medium- or fine-grained, gravish- Lacks characteristic microfauna.
green microcrystalline limestone, slightly argillace- 16. Specimen T140, section no. 64-2335. Bv Dr. Federico
ous, with finely granular iron oxides. Calcified mollusk Bonet and Enrique Riva Palacio. Field notes: lime-
and ostracod remains. Lacks characteristic micro- stone, 1 km. southwest of Axusco, at summit of first
fauna. hills in front of Cerro Tepetroje. Description: coffee-
89
colored, fine-grained microcrystalline limestone, with 23. Specimen T072, section no. 447. By A. Sotomayor. Field
geodes lined with calcite, slightly granular, slightly notes: volcanic tuff; road in mountains about 3 km.
argillaceous. Lacks characteristic microfauna. east of Ajalpan. Macroscopic description: fine-
17. Specimen T077, section no. 449. Bv A. Sotomayor. grained, green-colored rock. Constituents: fine-
Field notes: limestone with altered constituents; grained volcanic ash, plagioclases, oxides, calcite,
barranca on first stream beyond Rio Zapotitlan chlorite; devitrification of original material. Classi-
(Chilac) on road from Chilac to Atzingo. Texture: fication: fine-grained volcanic tuff (ash-tuff).
clastic. Constituents: elastics, calcareous fragments 24. Specimen T044, section no. 444. Bv A. Sotomayor. Field
of organisms, remains of microfossils, calcite, oxides, notes: dike some 200 m. before the Moctezuma
microcrystalline and cryptocrystalline secondary bridge, Tehuacan-Huajuapan de Leon road (enter-
quartz. Classification: silicified “calci-rudite.” ing San Antonio Texcala). Macroscopic description:
18. Specimen T062, section no. 445. By A. Sotomayor. Field light-grav massive rock. Texture: aphanitic, aphyric.
notes: sandstone, 200 m. beyond the Rio Zapotitlan Constituents: labradorite, augite, magnetite, hema-
(Chilac) on road from Chilac to Atzingo. Macroscopic tite, chlorite, limonite. Classification: augite basalt.
and texture description: clastic psammitic rock. Con- 25. Specimen T138, section no. 457. By A. Sotomayor. Field
stituents: detrital components (basalt, quartz sand- notes: old manantial; slopes of Cerro Tepetroje south-
stone, calcite, feldspars, quartz) and cement (calcite, west of Axusco. Macroscopic description: coffee-red
chlorite). Secondary minerals: limonite, chlorite. breccia. Texture: clastic. Detrital constituents: lime-
Classification: “Lithic” sandstone. stone, sandstone, schist, quartz, feldspar, granite.
19. Specimen T137, section no. 64-2334. Bv Dr. Federico Cement: calcite, hematite. Classification: calcareous
Bonet and Enrique Riva Palacio. Field notes: cal- breccia. Origin: hydroclastic sedimentarv rock.
careous breccia; slopes of Cerro Tepetroje about 3 26. Specimen T136, section no. 456. By A. Sotomayor. Field
km. southwest of Axusco. Description: contempo- notes: calcareous breccia; slopes of Cerro Tepetroje
raneous breccia, with subangular fragments (diam- southwest of Axusco. Macroscopic description: cof-
eters up to 4 cm.) of coffee-colored, fine-grained fee-red breccia. Texture: clastic. Constituents: lime-
microcrystalline limestone, with abundant recrvstal- stone fragments, schist, altered rocks, quartz, feld-
lized veins, iron oxides, and a little secondary silica, spars, hematite, sandstone, chlorite. Cement: calcite,
infrequent mollusk remains, calcified radiolaria, fo- hematite. Classification: calcareous breccia. Origin:
raminifera so badly broken that they are impossible hvdroclastic sedimentary rock.
to identify — all in a matrix of granular silicified lime- 27. Specimen T157, section no. 459. Bv A. Sotomayor. Field
stone with abundant iron oxides and a well-crystal- notes: calcareous breccia; between Cerro Pelon and
lized calcite. Lacks fauna. Rio Salado, near Pueblo Nuevo, Puebla. Macroscopic
20. Specimen T079, section no. 450. Bv A. Sotomayor. Field description: white, cream, and coffee-colored breccia.
notes: lava; barranca on the first stream beyond the Texture: clastic. Detrital components: limestone, si-
Rio Zapotitlan (Chilac) on the road from Chilac to licified rocks, quartz, feldspars, chalcedony. Classi-
Atzingo. Macroscopic description: rose to reddish- fication: calcareous breccia. Origin: hvdroclastic
coffee color. Texture: aphanitic. Constituents: labra- sedimentary rock.
dorite, hematite, magnetite, pyroxenes. Classifica- 28. Specimen T157b, section no. 64-2338. Bv Dr. Federico
tion: basalt. Bonet and Enrique Riva Palacio. Field notes: cal-
21. Specimen T076, section no. 448. Bv A. Sotomayor. Field careous breccia; between Cerro Pelon and Rio Salado
notes: reworked lava; barranca on first stream beyond near Pueblo Nuevo, Puebla. Description: greenish-
the Rio Zapotitlan (Chilac) on the road from Chilac gray, medium- or fine-grained, microcrystalline lime-
to Atzingo. Macroscopic description: massive porphy- stone, irregularly agrillaceous, with iron oxides and
ritic rock with light gray color. Texture: porphyritic recrystallized veins. Lacks fauna.
with xenoliths. Primary minerals: andesine, magne-
tite, orthoclase. Secondary minerals: quartz, calcite,
chlorite, oxides. Comments: the rock shows evidence
of xenolithic granite and lavas. Classification: ande-
REFERENCES
sitic porphyry. Aguilera, 1906.
22. Specimen T071, section no. 446. By A. Sotomayor. Field Applin, 1955.
notes: road in mountains about 3 km. east of Ajalpan. Barrera, 1946.
Macroscopic description: massive dark-gray and cof- Blasquez, 1957.
fee-colored rock with quartz veins. Texture: cata- Calderon Garcia, 1956.
clastic. Constituents: quartz, microcline, plagio- Fries et al., 1955.
clases, chlorite, sericite, limonite, secondary quartz. Miillerried, 1931.
Classification: cataclasite. Salas, 1949.
90
CHAPTER 6
The Human Skeletons
James E. Anderson
D
uring the excavations in the Tehuacan Valley, ture comparisons as data become available from other
human skeletal material representing, if only sites; the search for temporal trends showing micro-
fragmentary, more than seventy individuals evolution; the correlation of biological findings with
was recovered from seven sites and from eight of the culture change; and the study of the incidence of
nine phases of the prehistoric cultural sequence. The New World populations.
disease in early
importance of these bones lies in their antiquity (the I measured crania and long bones according to the
six earliest burials being from the El Riego phase of techniques described by Ashley Montagu ( 1960) and
about 6500-5000 b.c.) and in their scientific excavation
in a cultural context with known dating. The bones are
representative of a sequence spanning the transition Table 2. Burials from El Riego Cave
from a nomadic food-gathering and hunting economy West Niche (Tc 35w), by Phase and Level
to that of full-time agriculturalists living in sizable
Lab or
communities.
field no. Age Sex Comments
This report grows out of the two intervals I spent in
Venta Salada
Tehuacan restoring and studying the skeletal material
Zone 1 A adult F Intracranial fragments only
in January and July of 1964.° My emphasis in this
B adult F Fragments of skeleton
report has been descriptive rather than metrical, with C adult M Almost complete except skull
particular attention to morphological variations and D adult F Fragments of skeleton
E adult M Partial including mandible

pathology of bone and tooth. The objectives which
Ll'l under 6 mos. ? Fragmentary
guided the work included: the provision of basic popu- F
Ll:2 adult Fragmentary, no skull
lation data for use in interpreting the archaeological Zone 2 L2:l adult F No skull, legs burned
record; the description of skeletal morphology for fu- L2:2 newborn ? Fragmentary
L2:3 2-4 yrs. ? Almost complete and intact
L2:4 ? Most of skeleton, part of skull

5 yrs.
° I am pleased to record my gratitude to the R. S. Peabody
?
L2:5 6 yrs. Mandible and other fragments
Foundation for Archaeology and to Dr. Richard S. MacNeish
Zone L3:l Fragmentary skull and teeth
for making it possible for me important collection.
to study this
3 13 yrs. ?
I owe a great debt to Dr. MacNeish for his stimulating guidance

L3:2 adult F Fragmentary skull and teeth
in the interpretation of the material and to Mrs. MacNeish for Palo Blanco
her kindnesses while I was at Tehuacan. Zone 4 L4:l adult M Fragmentary, partly burned
During his visit to the site, Dr. Santiago Genoves was most L4:2 11 yrs. ? Fragmentary, partly burned
helpful in discussing criteria for determining age and sex and
El Riego
in providing literature on Mexican osteology. Through the kind-
ness of Dr. Arturo Romano I was able to examine skeletal ma- Zone 5 L5:l adult F Cremated fragments
terial in the National Museum of Mexico, where Senorita Teresa Ajuereado

Jaen was most helpful. I am particularly indebted to Dr. Charles Zone 6 L6:l adult ? Fragment of mandible
Goff for his guidance in interpreting some of the pathological
Unclassified adult F Almost complete; looted grave
specimens.
91
Table 3. Burials from El Riego Cave Tables 2-8 summarize the burial data for each of
East Niche (Tc 35e), by Phase and Level the sites which yielded human skeletal remains. Some
of the skeletonswere given burial numbers in the field,
Lab or
field no. Age
other fragmentary remains were given numbers in the
Sex Comments
laboratory, and a few remains are identified by the
Venta Salada
level of excavation in which the burials were found.
Zone B B:l adult F Fragmentary
Bll adult M Fragmentary In the following descriptions of the burials, the order
Bill 15 mos. ? Fragmentary skull and teeth follows the chronological sequence established for the
burned black
F
Tehuacan Valley, beginning with the most ancient
Zone C C:l adult Partial, some bones burned
black material and proceeding toward the present. The ma-
under 2 yrs. ? Fragmentary
C:ll
terial is grouped according to cultural phase.
Palo Blanco
Zone D 0:1 adult M Fragmentary

Ajuereado Phase
D:ll adult F Fragmentary
Dill newborn ? Fragmentary This, the earliest phase, ended before 6500 b.c. The
Zone E E:l adult ? Fragmentary, burned population consisted of nomadic microbands whose
subsistence depended on the collection of wild plants
and the hunting and trapping of game.
The only human skeletal material recovered from
Table 4. Burials from Coxcatlan Cave this phase, a fragment of the left side of an adult
(Tc 50), by Phase and Level mandible, was found in the West Niche of El Riego
Lab or
field no Age Sex Comments
Venta Salada 7 adult F Jaws, teeth, other fragments
Palo Blanco 1 15 mos. ? Complete, desiccated tissue

Table 5. Burials from Ajalpan Site
El Riego (Ts 204), by Level (all Ajalpan Phase)
Zone XIV 2 5 yrs. ? Almost complete, broken skull
Lab or
3 under 6 mos. ? Almost complete
4 adult M Burned black, almost complete
"Burial 1” 1 adult M Long bones and mandible
5 adult F Almost complete
Zone F 2 adult M Fragments
6 under 6 mos. ? Almost complete, broken skull
3 adult F? Fragments
Zone G 4 subadult M Fragments
Zone G‘ 5 adult M Fragments
6 adult F Fragments
Zone H 7 adult ? Fragments
estimated stature on the basis of the length of limb

bones, using the formulae of Trotter and Gleser (1958).
The age immature individuals was determined by
of
Table 6- BurialIs from Quae hilco (Tr 218),
the stagedental eruption and the progress of
of
by Phase
epiphyseal fusion. Since the symphysis pubis was rarely
available for study in the adult skeletons, only the rela-
Lab or
tiveage could be judged on the basis of degenerative field no. Age Sex Comments
changes in the bones, the degree of wear on the teeth

Palo Blanco 1 adult M No skull or teeth
as related to the cultural level, and the state of suture

Santa Maria 2 adult M Almost complete
closure. Examination of the pelvis, the skull, and the 3 8 yrs. ? Almost complete
*'
articular areas of the appendicular skeleton helped to 4 adult M Almost complete
determine the sex of the skeletons. When other areas 5 6 yrs. ? Teeth, skull and other fragments
6 adult F Teeth, fragmentary skeleton

were equivocal, pelvic criteria as described by Genoves
7 adult M Teeth, part of skull, other bones
(1962) were given priority. Also the size of joint sur- 8 adult M? No skull or teeth
faces, particularly thediameter of the femoral head, 9 adult M F ragmentary
was used in determining sex, rather than the length and

Note: Fragments found in Zone A are probably parts of Burial 1. Fragments recovered from
robustness of the long bones (Anderson 1963). Zones B, B\ C, and C
1
are probably parts from Burials 2-9, particularly Burial 5.
92
THE HUMAN SKELETONS
Table 7. Burials from Purron Cave (Tc 272), in the dry season and exploited the lush vegetation of
by Phase and Level the wet season.
Human material comes from three sites: Burial 4 at
Lab or
Purron Cave, Burials 2-6 from Coxcatlan Cave, and a
fragmentary burial from Zone 5 of the West Niche of
Palo Blanco El Riego Cave.
Zones C-E 1 adult F Fragmentary
Tc 272, Burial 4. This fragmentary skeleton from
Abejas
Purron Cave is b.c. The bone is soft
dated about 6000
Zone M - adult ? Tibial fragment
- adult ? Pelvic fragments

and friableand deeply stained by red ochre. The skull
Zones N-0 3 adult M Fragmentary: parts added from Zone N‘. vault is badly warped and the frontal region is
-
Zones 0-P adult ? Pelvic fragments and R. calcaneus crushed. The temporal bones are complete, but the
Coxcatlan facial skeleton is represented by only the left zygoma.
Zones P-Q - adult ? Cuboid, calcaneus, fibula
There are two mandibular fragments and nine loose
Zone Q‘ - adult ? Rib fragments, lumbar vertebra
teeth. The limb bones are all represented, but are
Zones Q‘-R 2 adult M Partial skeleton
badly damaged, particularly at their joint surfaces.

El Riego
Zone R 4 adult F Partial skeleton All segments of the vertebral column are present, but
most are badly preserved.
The skeleton is probably that of a female, judging
from the delicacy of its supraorbital ridges and mastoid
Table 8. Burials from Coatepec (Ts 368e), processes, marked parietal bossing, indistinct muscu-
by Phase and Level lar markings in the nuchal region, feminine conforma-
tion of the right ischial notch, and a femoral head
Lab or
Venta Salada
Zone A - young ? Fragments

- subadult ? Fragments
Palo Blanco
1 17 years ? Fragments
Santa Maria
3 2 years ? Fragmentary
Zone C - adult ? Femur fragments
Zone C‘ - adult ? Vault fragments, teeth
Zone D - adult ? Two small frontal fragments
Ajalpan
Zone K - adult ? Burned mandible, vault, talus

- subadult ? Mandibular fragment
Zone K‘ - adult ? Burned left talus

2
Zone K 2 adult M Intracranial fragments
Zone K
3
- adult ? R. frontal, L. maxilla, teeth
Cave. The chin is square or bilateral in form with
prominent mental eminences. The mental foramen is

double. Only two teeth are in place. The first premolar
isworn down almost to the neck, exposing its pulp
chamber and resulting in an apical abscess. The third
molar is tilted mesially toward the space formerly
occupied by the long-lost first two molars, whose alveoli
have been resorbed.
Fig. 52. The skull of Tc 50, Burial 4; El Riego phase; adult
male; reconstructed from 58 pieces of burned bone. Note
El Riego Phase
continuous supraorbital ridge, large infraorbital foramina,
This period spans the years 6500 to about 5000 b.c. and zygomatic region which appears vertical in side view
The people subsisted on a low-level hunting economy and flat when seen from below.
93
and leaving short crescents of enamel projecting on

the occlusal surface. One premolar has a carious lesion
in the cervical region of its distal surface which has
opened into the pulp chamber.
There is flattening of the femur, but not of the tibia:
the platymeric index is 71.9; and the platycnemic index
is 74.2. There is no vastus notch or third trochanter, but
a large septal aperture is present on the left humerus.

Anterior and middle facets on the calcaneum are dis-
crete. The articular surfaces of both patellae have
central pitted areas diagnostic of earlv osteoarthritis.
Vertebral body fragments show moderate peripheral
osteophytosis. There is a healed fracture of the distal
end of the left radius with the characteristic “dinner
fork” deformity, but no signs of fracture on the ulna.
Tc 50, Burial 2. This burial from Coxcatlan Cave is
the almost complete skeleton of a five-year-old child,

found in the same grave with the skeleton of a still
younger child (Burial 3). The heads of both children
had been removed from their bodies and were ex-
changed in burial. .Although all parts of the skull
properly belonging to Burial 2 were present, the skull
could not be reconstructed. The occipital region may
Fig. 53. The skull of Tc 50, Burial 5; El Riego phase; adult
female. Note short, broad mandibular ramus; gonial ever- have been broken in before burial, and there was some
sion, vertical malar region; prominent zvgomaxillarv tuber- evidence of charring.
osity; and narrow space for the nasal bones. The age of the child was estimated as five years on
the basis of the following evidence: All parts of the
temporal bone have fused and lateral growth of the
diameter of only 40 mm. Although the vault sutures of tympanic plate is almost complete. Its lateral edge is
this adult remain distinct, there are signs of the aging lacelike, and dehiscences are present bilaterally. The
process in the osteophvtosis of the vertebral body mar- mastoid process is well developed for the age of the
gins. The stature was estimated from the length of the skull. The fusion of the condylar to the squamous part
ulna as 167.8 ± 4.66 cm., or approximately 5 feet, 6 of the occipital bone is incomplete. All the deciduous
inches. This estimate may be distorted since it falls teeth have erupted, and the crowns of the first perma-
outside the range of the other estimated statures for nent molars are just at the alveolar margin. The limb
females given in Table 12 below. epiphyses have not yet fused. The atlas vertebra con-
The neurocranium appears short, with its greatest sists of the unfused right and left neural arches and
height above the marked parietal bosses. The cranial a separate anterior arch segment. Itsforamen transver-
index is estimated at 78.3, which is in the mesocranial sarium has not quite formed. The axis vertebra is com-
range. The only Wormian bones are tiny inclusions plete except for the fusion of the os apicis and the
in the lambdoid suture and there is no sagittal crest or closure of the foramen transversarium. The other
metopic suture. On both sides there are supraorbital cervical vertebrae have fused neural arches, but two
notches, but the vessels and nerves leave no grooves arches have not fused to the centrum. The upper eight
on the frontal bone. The supraorbital ridges are con- thoracic vertebrae have arches completely united to
tinuous across the midline. The occipital region con- their centra, but the lower four show recent fusion. In
sists of a low mound and a Both tympanic plates
ridge. the lumbar region 'arch and centrum are well fused,
are slightly thickened but have no dehiscences. except for L4 and L5, which show spina bifida (see Fig.
A fragment of right mandible shows complete re- 65 b). The neural arch of the sacral elements has fused
sorption of its alveolar bone, following the loss of all to the ala, but the not fused with the body.
latter has
posterior teeth. The anterior teeth are severely worn The sacral canal iswidely open posteriorly for its whole
in an asymmetrical fashion, with oblique slopes on length. There is a pseudo-epiphysis on the first meta-
one or more surfaces exposing the dentin of the crown tarsal.
94
THE HUMAN SKELETONS
On each side of the skull there is a large Wormian

bone deep parietal notch. Supraorbital notches are
in a
bilateral, but no grooves may be seen. There is a promi-
nent tuberosity mainly on the zygomatic process of the

maxilla. The only wear shown by the teeth is a flatten-
ing on the cusps of the deciduous molars. The upper
left central incisor had been lost, and its socket is the
site of an alveolar abscess.

Tc 50, Burial 3. The skeleton is of an infant less than
six months of age. The skull had been exchanged in
burial with the skull of Burial 2. Unfused are: the

symphysis menti, the parts of the occipital bone, the
metopic suture, and the limb epiphyses. The squamous
and petrous parts of the temporal bone have fused,
but the tympanic ring shows no signs of lateral growth.
The crowns of the deciduous teeth are forming and
their alveoli are open. In the vertebral column fusion
of the two halves of the neural arch has not begun.
Tc 50, Burial 4. This is the skeleton of an elderly
male, most of whose bones are burned black. The
skeleton was found in a grave with Burials 5 and 6.
The sex was determined from the heavy brow ridges
and nuchal markings, male sacral proportions and
ischial notch, and the robustness of the limb bones.
Cranial sutures are obliterated, and there is advanced
degenerative disease of the joints.
Fig. 54. a, Child of 2-4 years, Tc 35w; Venta Salada phase;
The skull was reconstructed from fifty-eight pieces.
with typical large tuberosity, b Tc 35w, Burial C; adult
,
Its vault is mesocranial in proportions. The supraorbi-

male; Venta Salada phase, c, Tc 35w; looted burial; adult
tal ridges are continuous. The facial region is flat with female, d, Tr 218, Burial 2; adult male; Santa Maria phase;
large zygomaxillary and malar tuberosities. The occip- treponemal infection has destroyed most of the vault.
ital region is mound shaped. The chin form is medio-
and there is
bilateral, slight gonial eversion and a
mylohyoid arch on both sides.
The jaws are toothless with resorbed alveoli, except with considerable angulation. Two rib fragments also
for three short tooth roots. Advanced attrition had show signs of healed fracture.
exposed the pulp chambers of these and resulted in Tc 50, Burial 5. The skeleton of an adult female was
apical abscesses. found with Burials 4 and 6. The sex was determined on
Stature estimated from the length of the tibia is 165.1 the basis of pelvic and cranial criteria, as well as by the
± 3.27 cm., or 5 feet, 5 inches. femoral head diameter of only 39 mm. Although vault
Most of the available joint surfaces, especially the sutures are still distinct, degenerative changes through-
apophyseal joints of the cervical vertebrae, show signs out the skeleton suggest at least middle age.
of osteoarthritis. Osteophytosis of the vertebral bodies The skull was completely reconstructed. Wormian
is most marked in the lumbar region. A compression bones occur as inclusions in the lambdoid suture as
fracture of the third thoracic vertebra has caused its well as bilateral parietal notch bones. The facial region
body to become wedge-shaped and to fuse to the is flat, with small malar tuberosity and marginal tuber-
vertebral bodv above by a large osteophyte on the left cle, but with a huge zygomaxillary tuberosity formed
side. The lumbar vertebra exhibits spondylolis-
fifth jointlyby the two bones. An occipital mound extends
thesis, the anterior slipping of its body being shown from the lambda to the inion. The chin is median in
by lipping on the periphery of the first sacral body. shape. Gonial eversion is evident, and there are bilat-
There is an old healed fracture of the right fifth eral mylohyoid arches. Apparent maxillary alveolar
metatarsal. Fractures are present at the midshaft of prognathism has resulted from the loss of many teeth.
the left radius and ulna, where healing has occurred The foramen spinosum is open medially on both sides.
95
Three left ribs underlying the scapula also show signs

of healed fractures.
Both humeri have septal apertures, but there is no
vastus notch or third trochanter.
Tc 50, Burial 6. The third component of the multiple
burial is the skeleton of an infant less than six months
of age.The skull is badly fragmented. Immaturity is
shown by the lack of fusion of the symphysis menti, the
parts of the occipital bone, the vertebral neural arches,
and the limb epiphyses. The four parts of the temporal
bone have fused, but there has been no lateral growth
of the tympanic ring.
Tc 35tc, Zone 5. This level of El Riego Cave contained
a cremation consisting of small fragments of limb,
vertebrae, and skull. The delicacy of the bones suggests
that they are those of an adult female. A burned canine
Fig. 55. The skull of Tr 218, Burial 7; adult male; Santa tooth shows evidence of cervical caries.
Maria phase; cranial index, 82.2. Note frontal slope, flat Summartj. Seven individuals of the El Riego phase
vertical occiput, “rocker’-shaped lower border of mandible. are represented from three sites: at Purron Cave the
single extended burial of an adult male; at Coxcatlan
Cave two multiple burials, one consisting of two chil-
dren whose heads had been removed and exchanged,
the other consisting of three individuals, a burned
Early temporomandibular joint arthritis is shown by elderly male, a somewhat younger female, and an in-
eroded areas on both articular eminences. fant; at El Riego Cave a cremated adult female.
Half of the teeth had been lost and the alveoli re- The skulls are uniformlv mesocranic, with a low oc-
sorbed. Advanced wear on the remaining teeth left cipital mound and the highest part of the vault occur-
oblique surfaces, exposing secondary dentin and some- ring above the parietal bosses. The faces are flat with a
times the pulp chamber. Sharp crescents of enamel re- prominent zygomaxillarv tuberosity' present even in
main on the occlusal surfaces. The palate is eroded by children.
extensions from alveolar abscesses. The unusual pattern and advanced state of the wear
Stature calculated from fibular length is 159 ± 3.24 on the teeth is probably the result of fibrous and gritty
cm., or 5 feet, 3 inches. vegetable material being pulled through the teeth
Perhaps even more than her colleague. Burial 4, this during mastication. The teeth are worn down oblique-
skeleton is riddled with degenerative and traumatic ly, leaving a crescent of enamel on the crown and
changes. Osteoarthritis is present in both hips, both often exposing the pulp chambers. Loss of teeth under
first metatarsophalangeal joints, the acromioclavicular these conditionswas extreme. Caries is not common,
joints, the left ankle joint, and at the vertebral apoph- and when present, results from the impaction of food
yseal joints. There is advanced osteophytosis in all between teeth.
regions of the spine. The very high incidence of healed fractures exhibited
Compression fractures occur throughout the ver- bv two skeletons from the same grave is noteworthy.
tebral column: C4, C5, T4, and T7 have been crushed
Coxcatlan Phase
to approximately half their normal thickness, and the
bodies" of T8 and T12 have become wedge shaped. This period, dated roughly from about 5000 b.c. to
The left scapula exhibits signs of an old fracture 3500 b.c., is one of incipient agriculture, but there is
of the bodv which healed with infection. From a large still a great reliance on wild foods. Skeletal material
central defect, three fracture lines radiate to the per- from this phase comprises only Burials 2 and 3 from
iphery of the bone. The fragments have slightly over- Purron Cave, dated about 4300 b.c., and some frag-
lapped each other, giving the region of healing a ments found at related layers of the same site.
pleated appearance. Along the fracture line are round Tc 272, Burial 2. This damaged skeleton of an adult
regular holes representing abscesses which failed to be male is rather curiously incomplete, lacking the left
contained by the thinness of the blade of the scapula. forearm and lower limb.
96
THE HUMAN SKELETONS
The skull vault differs considerably from any others

from the Tehuacan sites. It is relatively long and low,
with some heightening behind the marked parietal
bosses. The estimated cranial index is 68.3. There is a
prominent mound-shaped occiput. The supraorbital
ridges are heavy, continuous, almost straight, extend
the full width of the orbit, and blend with the supraor-
bital margin. Behind them, the postorbital region ap-
pears constricted, although the minimum frontal diam-
eter of 88 is not extremely small. The mastoids are big
and bulky with a huge suprameatal “torus” and a deep
triangular suprameatal pit. The mandible has very
strong mental eminences which result in a subsym-
physeal concavity. There is gonial eversion.
The dental picture resembles that seen in the El
Riego phase, with marked oblique attrition producing
enamel crescents on the occlusal surfaces of the teeth. Fig. 56. The skull of Tr 218, Burial 4; adult male; Santa
Maria phase. Note prominent, straight, continuous supra-
There is evidence of three abscesses, and periodontal
orbital ridge; blurred subnasal margin; alveolar prog-
disease shows in the marked recession of the alveolar
nathism; strong suprameatal crest. Mandible shows the
margin and exposure of the roots.
typical ridged and everted gonial angle.
The limb bones appear long, straight, and very ro-
bust. The linea aspera is spectacular. Stature calcu-
lated from the length of the femur is 168.8 ± 3.8 cm.
(5 feet, 6 inches). this phase and a few' pieces of lower limb bone from
Osteoarthritis appears only in the cervical apoph- Zone M of Purron Cave.
yseal joints. Thisand the peripheral lipping of the
bodies encroaches on the foramina transversaria. Two Purron Phase
ribs show signs of healed fractures. This phase, w'hich lasted from about 2300 to 1500
Tc 272, Burial 3. This burial consists of a mandible, b.c., is the least understood segment of the sequence.
five loose upper
and the fragmentary infracranial
teeth, Pottery was first produced during this period. No hu-
skeleton of an adult, probably a male. The crowns of man skeletal material has been recovered from the twm
most teeth have been worn almost flat, exposing the excavated floors.
dentin. Enamel forms a ring on the occlusal surface,
surrounding a depressed zone of dentin. The lower Ajalpan Phase
left third molar is unworn (it evidently was unopposed) This phase lasted from 1500 b.c. to about 900 b.c.
and so has a carious lesion in a deep distolingual pit on The population consisted of full-time agriculturalists
its crown. The mandible is robust and has a medio-bi- occupying small villages. The skeletal material, repre-
lateral form of chin. The only pathologv noted on the senting probably twelve individuals, comes from two
infracranial fragments is advanced lipping of one lum- sites: Coatepec (Ts 368e) and Ajalpan (Ts 204 and
bar vertebral body. Ts 204c).

Tc 272, Fragments. Fragments of adult cuboid, cal- Ts 368e, Burial 2. These remains from Coatepec con-
caneum, fibula, rib, and vertebra were found in Zones sist of a few fragments of an adult skeleton, mainly
Q Q, and P of Purron Cave.

1
,
pelvis and scapula, with evidences of burning. The
ischial notch is that of a male.
Ts 368e, Zone K. This level contained what are prob-
Abejas Phase
ably parts of two individuals: (a) A burned fragment
During this period, extending from about 3500 to of adult mandible with teeth charred and their crowns
2300 b.c., the population of the valley alternated be- broken. There is evidence of loss of teeth, both before
tween hunting camps during the dry season and pit and after death. There are also rib fragments, parts of
houses on river terraces during the wet season. An the skull vault in w'hich the bone is very thick, and a
increasing amount of food was being grown under cul- left talus. (
b) A
fragment of the left side of a subadult
tivation. Unfortunately, we have only one burial from mandible with three molar teeth. The first molar has
97
(b Fragments of adult radius, and a man-

ribs, a left
dible with chin of median form. Loss of teeth was
followed by alveolar bone resorption.
Ts 204, Zone G, Melded a frontal and a left parietal
fragment, the shaft of a left humerus, femoral condvle
fragments, and the left mandibular ramus of a sub-
adult male. The supraorbital ridges are continuous and
extend for half of the width of the orbit. Min im um
frontal diameter is 95. The root of the nasal bones is
broad. The third molar has not yet erupted. There is
slight gonial inversion and no mvlohvoid arch. The
breadth of the ramus mm. and the height is 57 mm.
is 34
Ts 204, Zone G 1 contained parts of two individ-
,
uals: (a) a frontal bone fragment with sharp supraor-

bital margins, probablv that of an adult female; ( b a
humeral head with a diameter of 0.41 mm., a femoral
condyle fragment, and a left adult male patella with a
few bony spurs extending into the quadriceps attach-
ment.
Ts 204, Zone H, contained fragments of an adult
skeleton of indeterminate sex: proximal end of right
ulna, ribs, left mandible, and loose premolars. The
Fig. 57. Two immature skulls showing different vault forms. mandibular fragment shows premortem tooth loss. The
Above: Te 50. Burial 1. Palo Blanco phase. Below: Tc 35w,
premolar crowns are worn flat and one exhibits large
Zone 2, 2—4 years old, Yenta Salada phase. Measurements
mesial cervical caries.
are given in Table 10.
Summary. If we maintain the distinctness of the
bones at each level, the fragments from components
blunted cusps, the second was lost after death, and the of the Ajalpan phase represent five individuals from
crown of the unerupted third lies in its alveolus. Coatepec and seven from Ajalpan. One skeleton from
Ts 36Se, Zone K 1
,
contained a burned left talus. each site is a subadult. The absence of child skeletons
Ts 368e ,
Zone Ks . vielded a fragment of a right frontal may be noteworthv. Evidence of burning is found in
bone and part of a left maxilla bearing three teeth. onlv one skeleton. The pattern of dental attrition has
Attrition has exposed dentin on the lateral incisor, but changed with the switch to an agricultural economv:
has only blunted the cusps of the canine and first the wear is not as severe and is no longer oblique. Loss
premolar. of teeth continues to be serious, but is now caused bv
Ts 204, Burial 1. This burial from the Ajalpan site caries.
consists of manv split fragments of long bones and a
partially restored mandible of an adult whose ro-
Santa Maria Phase
bustness suggests that he is male. The chin is of medio- During this phase, dating from before 900 b.c. to
bilateral form. Many teeth have been lost with subse- about 200 b.c.. a population of full-time farmers using
quent resorption of the alveolar process. Cusps are irrigation lived in Milages oriented around a larger
worn flat, exposing the dentin in their cores. Two pre- communitv with ceremonial structures. The skeletal
molars share approximal caries resulting from food material comes from eight burials at Quachilco and
wedged between them, and a canine has a small cervi- from certain zones at Coatepec.
cal carious lesion. Tr 218, Burial 2. This almost complete adult male
Ts 204, Zone F, contained bone fragments prob- skeleton from Quachilco is one of the most interesting
able representing two individuals Fragments of:
(a) specimens among the human materials. Much of the
ribs, the shaft of a humerus, a right patella, and a piece vault of the skull is missing. Examination of the re-
of right mandible with emptv alveoli OK'ing to post- maining areas, supplies the explanation for its fragility.
mortem tooth loss. The chin form is medio-bilateral, A widespread osteitis of the neurocranium, sparing
and there is no gonial eversion. General skeletal robust- onlv the squamous temporal, has eroded the external
ness suggests that this individual was probably a male. surface in an irregular manner and in some areas has
98
THE HUMAN SKELETONS
completely penetrated the skull. Surrounding these

diffuse lesions are unevenly deposited plaques of new
bone. The irregular “worm-eaten” appearance is typ-
ical of a treponemal infection, probably syphilis.
The chin is bilateral in form, and shows some gonial
eversion. There is a well-marked ossification of the

apical ligament’s attachment to the anterior margin of
the foramen magnum.
Attrition has exposed the dentin on the incisors and
canine teeth, but has only blunted the cusps on the
posterior teeth. Six teeth are carious, and five have
been lost and the alveoli resorbed. Crowding has re-
sulted in slight mesial rotation of the upper left central
incisor and the lingual eruption of the lower left lateral
incisor.
Stature determined from the length of the fibula is
157.3 ± Very little osteo-

3.24 cm. (5 feet, 2 inches).
phytosis found in the vertebral column except on
is
C3 and C4, which are markedly lipped and fused with

each other at body and apophyses, but not at the Fig. 58. a. Right side of the vault, Tc 272, Burial 2; adult
male; Coxcatlan phase, b, Left side of the vault, Tc 272,
laminae.
Burial 4; adult female; El Riego phase, c, d. Right and left
Tr 218, Burial 3. This is the almost complete skeleton
sides of the skull, Tr 218, Burial 2, showing the treponemal
of an eight-year-old child. The first permanent molar
destruction of the vault.
is at the occlusal plane, as are the upper four incisors.
The posterior deciduous teeth were sufficiently worn to

expose the dentin. Crowding is shown in the mesial
rotation of both upper central permanent incisors. vertebral osteophvtosis. The sacrum consists of only
Supraorbital notches are bilateral. The typical zygo- four elements and has accessory articulations with L5.
maxillary tuberosity is huge, and the chin is medio- Tr 218, Burial 5. This burial consists of the frag-
bilateral in form. mentary remains of a six-year-old child. The first per-
All vertebrae show union of the halves of their manent molar has erupted. Dentin is exposed on the
neural arches, which in the cervical and upper thoracic incisors, and there are cusp facets on the posterior
region have also fused to their centra. The anterior arch deciduous teeth. The chin is medio-bilateral, and gonial
segment of the atlas and the os apicis of the axis have eversion is prominent. There is no mylohyoid arch, but
not fused. The foramen transversarium is absent on there are accessory mandibular foramina on each side.
the left side of C6 and C7. Tr 218, Burial 6. This burial consists of teeth and
Tr 218, Burial 4. The almost complete skeleton is that bone fragments of an adult female. The chin is bilat-
of an adult male. The mesocranic skull has multiple eral, and there is gonial eversion. The anterior teeth
tiny Wormian bones in the right coronal suture. On have exposed dentin; the posterior teeth have cusp
each side is a supraorbital foramen, but there are no facets. There has been labial eruption and rotation of
grooves on the frontal bone. The supraorbital ridge is the lower right canine. The lower central incisors are
a continuous long low bar which blends with the supe- worn down severelyby a groove on their labial sur-
rior orbital margin. The chin is bilateral in form, and faces that runs downward and to the right. The enamel
there is gonial eversion. On the left side, an antero- of all teeth appears mottled.
medial spur projects into the foramen ovale. There is an anomalous foramen in the right upper
Attrition has exposed the dentin on the anterior teeth, part of the anterior surface of the manubrium.
and has blunted the cusps on posterior teeth. There are Tr 218, Burial 7. This specimen consists of the skull,
many pits on the enamel of molars; two are carious. teeth, and fragmentary infracranial skeleton of an
Crowding has resulted in the mesial rotation and labial adult male. The vault is high, short, and round, and the
eruption of the upper central incisors, and in the lingual estimated cranial index is 82.2. The supraorbital ridges
displacement of the lower left central incisor. are continuous. The occiput is mound shaped. The chin
There is acromioclavicular joint arthritis and slight is medio-bilateral in form, and there is gonial eversion.
99
femur and cranial vault and a lower premolar with

blunted cusps.
Summary. The human material from the Santa Maria
phase consists of six adults and two children from
Quaehilco, and a child and at least one further adult

from Coatepec. Certain general observations may be
made: (a) There are no evidences of burned bone. ( b
Cranial vaults are shorter, higher, and rounder, (c)
Teeth are less worn, but enamel pits and fissures have
become the site of caries. ( d ) The crowding of in-
cisors and the labial grooves on the lower incisors of
one individual may correlate with the finding of lip
plugs in this complex. ( e ) Two individuals have ac-
cessory lumbo-sacral articulations.
Palo Blanco Phase

The population from 200 b.c. to a.d. 700 had de-
veloped a more sophisticated agricultural economy and
lived in villages oriented to large hilltop sacred cities.
The skeletons from this complex consist of individual
Fig. 59. a, Posterior of the skull of child 2-4 years old, Tc burials from Coxcatlan and Purron caves, from Qua-
35w, showing ten wormian bones in the right lambdoid chilco and Coatepec, as well as scattered bones from
suture, b, Detail of left face, Tc 50, Burial 1, a child, Palo the two niches of El Riego Cave.
Blanco phase; an infraorbital suture is still distinct and
Tc 50, Burial 1. The skeleton of a fifteen-month-old
bounds an accessory infraorbital foramen, c, Internal sur-
child from Coxcatlan Cave is complete. Some soft tis-
face of the mandibular ramus, Tc 50, Burial 7, Venta Salada
sue has been desiccated and preserved.
phase, showing a mylohyoid arch over a groove which forms
at some distance from the mandibular foramen, d, A man-
The cranial vaidt is short and wide with prominent
dibular fragment from Tc 35w, Ajuereado phase, with a bossing. The posterior pole appears asymmetrical be-
double left mental foramen. cause of a more prominent left occiput, and there is a
wide supralambdoid sulcus. The zygomaxillary tuber-
osities are very large. The chin region is median in
form, and there is gonial inversion. The deciduous
Only three teeth are present, most having been dentition is complete except for unerupted second
lost before death and the alveolar bone resorbed. One molars and canines that are half way to the occlusal
of the teeth is carious. plane.
There are accessory facets between L5 and SI. The The atlas vertebra is in three segments and its for-
sacrum is damaged at the level of its fourth body so that amen transversarium is incomplete. Neither the dens
it is impossible to tell the number of sacral segments. segment nor the os apicis has fused to the axis, al-
There is slight osteophytosis of the lower lumbar though the neural arch is united to the centrum. The
region. spurs surrounding the foramen transversarium almost
Tr 218, Burial 8. The fragmentary intracranial skel- meet. The halves of the neural arch of the other cer-
eton is that of an adult, probably a male. No lower vical vertebrae have united, but the centrum remains
limb bones are represented. discrete in each. The foramen transversarium is almost
Tr 218, Burial 9. This burial consists of fragments of closed. The fifth cervical vertebra has a separate lateral
cranium, mandible, and intracranial parts of an adult element on each side which forms the external border
male. Anterior teeth have exposed dentin. The enamel of the foramen transversarium. This foramen is double
of the molars is characterized by pits and fissures, six on the right side of C7. In other regions the neural
of which have become sites of carious lesions. arches have not fused to the centra.
Ts 368, Burial 3. This fragmentary burial from Tr 218, Burial 1. The fragmentary skeleton from
Coatepec represents a two-year-old child. Quaehilco of an adult male possessed no skull or teeth.
Ts 368, Zones C, C 1 and D. Scattered human mate-
,
There are slight arthritic changes in the limb joints.
rial from these levels consists of fragments of an adult Tc 272, Burial 1. This fragmentary skeleton from
100
THE HUMAN SKELETONS
Purron Cave represents an adult female. The three

loose teeth do not show much wear, but one shows
caries.There is early osteophytosis in the lower lumbar
and the sacral region.
Ts 368, Burial 1. This burial from Coatepec is the
partial infracranial skeleton of a youth about seven-
teen years old. Elbow epiphyses have fused, but those
at the shoulder and wrist have not. Union of the head
to the shaft of the femur has begun. The symphysis
pubis is transversely ridged, and the vertebrae are com-
plete except for their secondary centers. There is a
large septal aperture on the only humerus available,
the left.
Tc 35w, Zone 2. The West Niche of El Riego Cave

contained a skeleton consisting of many small dam-
aged fragments of partially burned friable bone with
black crystalline accretions on the surface.
The age represented by the skeleton is estimated as
eleven years. There are wear facets on the cusps of the
Fig. 60. Advanced, oblique wear is characteristic
a, b, c,
firstpermanent molars and the second molars are just
of teeth of the earliest skeletons; cshows exposed dentin
beginning to erupt. The long-bone epiphyses have not
and remaining crescents of enamel, d, Teeth from a later
united. In the vertebral column the neural arches and phase show little wear; numerous enamel pits on the molar
centra have united. The axis shows recent fusion of the crowns are not carious because in life they were effectively
os apicis, but the dens segment has been united to plugged by calculus.
the body for a long time.
Tc 35iv, Zone 4. This level contained the fragmentary
burial of an adult male. The bones show varying de-
grees of burning. Tc 35w, Burial B. This is the partial skeleton of a
Tc 35e, Zones D and Fragments probably repre-
E. young adult female. The teeth show very little attrition.
senting four individuals were found in two levels of the The enamel of the molars is riddled with pits and
East Niche of El Riego Cave: (a) an adult male with fissures, most of which were plugged with a calculus
occlusal caries on a premolar, (b) an adult female with deposit, and only one of which is a site for caries. The
only slight dental attrition, (c) a newborn child, and buccal surfaces of the lower right canine and first pre-
(d) the burned remains of an adult of indeterminate molar are involved with a deep carious lesion which
sex. seems to have been preceded by a traumatic groove.
Summary. Material representing the Palo Blanco There is an area of minimal periostitis on the midshaft
phase was collected fromsix sites and consists of ten of the right tibia.
individuals, mainly fragmentary, of whom three are Tc 35w, Burial C. The almost complete skeleton and
immature and three show signs of burning. The teeth partial skull represents an adult male. The anterior
show less wear than in earlier periods and an increase surface of the zygoma is oblique rather than vertical, as
in the incidence of caries. was typical of earlier skeletons. The facial skeleton
is rugged with large malar tuberosity, marginal tuber-
Venta Salada Phase cle, and zygomaxillary tuberosity.
The most recent material comes from the Venta The dentition is complete and shows slight attrition.
Salada phase of about a.d. 700 to 1540. There were five Deep enamel pits on the molar crowns have been
burials and many partial remains from the West Niche plugged with calculus and are not carious.
of El Riego Cave and fragmentary remains from the Stature estimated from the length of the femur and
cave’s East Niche, Burial 7 from Coxcatlan Cave, and fibula is 163.8 ± 3.18 cm. (5 feet, 5 inches).
fragments representing two individuals from Coatepec. Tc 35ic, Burial D. The partial infracranial skeleton
Tc 35w, Burial A. From the West Niche of El Riego and a segment of the right mandible represent an adult
Cave came a badly weathered fragmentary infracranial female. One of three available teeth is partially de-
skeleton of an adult female. stroyed by caries. The acetabulum is shallow, elon-
101
with the squamous but not with the basilar part. All the
deciduous teeth are present. The skull vault is short,
broad, and high, with an almost vertical frontal region
and a very flat area above the inion. The left occipital
region is somewhat more prominent than the right, and
both parietals bulge laterally. There is an os inca, and
Wormian bones occur only in the right half of the
lambdoid suture, in the parietal notch, and at the as-
terion. There are a large zygomaxillary tuberosity and
an accessor)' mandibular foramen on each side.
The atlas vertebra is in four segments and the spurs
enclosing the foramen transversarium are just touching.
The dens segment of the axis is not united. The foramen
transversarium of the third cervical vertebra is incom-
plete on the right, but complete
and double on the left.
Such double foramina occur also on the left side of C5
and bilaterally on C6. No limb epiphyses have fused.
(d) Some of the fragmentary' remains are those of a
five-year-old child. The deciduous dentition is com-
Fig. 61. Evidence of crowded teeth in jaws from Tr 218, plete,and the first permanent molar crown is well
Santa Maria phase, a, Central incisors erupted labiallv and formed in its alveolus. The vertebral neural arch halves
are mesiallv rotated, b. The left central incisor is linguallv
are united with each other and some have fused to
displaced and the left lateral incisor is distallv rotated, c,
their centra. The failure of fusion of the neural arches
Left lateral incisor is tilted linguallv. d, The right canine
on the sacral segments at this age suggests spina bifida.
erupted labiallv.
No limb epiphyses have fused.
(e) The mandible and a feyv infraeranial fragments
of a six-year-old child displayed full deciduous denti-
gated to a long oval shape, and has a poorly defined tion. The first permanent molars yvere almost at the
anterior border. occlusal plane. Accessory mandibular foramina are
Tc 35w Burial E. The
,
skull fragments of tins partial present bilaterally.
skeleton of an adult male are burned black. There are Tc 35w, Zone 3. This level contained fragmentary
widespread evidences of osteoarthritis and an old healed remains of the skulls and teeth of two individuals: (a)
fracture of the shaft of the left fibula. a 13-year-old child yvhose teeth shoyved slight attrition,
Tc Zone 1. This level contained fragmentary'
35tc, and (b) an adult female yy'ith dentin exposed in the
remains of two individuals: (a) an infant less than six anterior teeth and cervical caries on a premolar.
months of age and (b) an adult female. Tc 35e, Zone B. A grave in the East Niche of El
Tc 35w, Zone 2. A great jumble of bone from this Riego Cave contained the fragmentary remains of three
level when sorted out represents five individuals, an individuals: (a) an adult female, (b) an adult male, and
adult female and four children. (c) parts of the skull and teeth, burned black, of a fif-
(a) The upper limbs of the adult female were asso- teen-month-old child.
ciated with each other in the burial, but the remaining Tc 35e, Zone C. This level of the East Niche con-
skeleton was burned and scattered. One molar tooth tained fragments representing two individuals: (a) an
shows an old fracture of a cusp and a small cervical infant less than two years old and (b) an adult female
carious lesion. There are no degenerative changes in yvith bones burned black. The front teeth of the latter
the joint surfaces of vertebrae. are yvorn to expose the dentin, and the cusps of the
( b Some of the fragments are those of a newborn posterior teeth are blunted. Two teeth yvere lost before
infant. death. One molar is carious.
(c) The skeleton of a child two to four years old has Tc 50, Burial 7. This burial from Coxcatlan Cave is
an intact skull. The elements of the temporal bone have an adult female represented by jaws and teeth, parts
united, a foramen
in the tympanic plate is almost en- of the skull yvhich are not reconstructible, and in-
growth of the plate is not complete.
closed, but lateral fracranial fragments mainly of the upper limb.
The condylar parts of the occipital bone have united The tympanic plates are slightly thickened yvith mar-
-
102
THE HUMAN SKELETONS
Table 9. Measurements of Adult Skulls, by Phase that they could be from the same family. In the East
(in mm.) Niche of the cave an adult male and female were
buried with the burned skeleton of a fifteen-month-old
Cox-
child, and a cremated adult female was buried with an
El Riego catlan Santa Maria Venta Salada
infant.
Tc 50-4 Tc 50-5 Tc 272-2 Tr 218-2 Tr 218-4 Tr 218-7 Tc35e-B Tc35w-C
The teeth of the Venta Salada phase are less worn
Sex M F M M M M F M and are characterized by enamel pits and fissures. A
Length 179.0 174.0 198.0* 171.0* 185.0* 168.0 - -
heavy calculus deposition seems to have partially pro-
Breadth 137.0 136.0 135.0* 140.0* 143.0 138.0* - -
Height 136.0 - - - 141.0* - - tected the teeth from caries.
126.0
Cranial module 147.3 148.7 - - - 149.0* - - The crania show signs of occipital flattening and a
Cranial index 76.6 78.3 68.2 81.8* 77.3* 82.2* - - more oblique slope to the facial aspect of the zygoma,
Basion-nasion 102.0 97.0 - 105.0* - 97.0 - -
although the zygomaxillary tuberosity remains promi-
Basion-
prosthion _ 99.0 _ 105.0* _ 101.0 _ _ nent.
Min. frontal 86.0 85.0 88.0 96.0* 92.0 - 98.0 98.0
Length-ht.
_ _ _ _ _
The Skulls
index 70.4 78.2 84.0*
Breadth-ht.
index 92.0 100.0 _ _ _ 102.1* _ _ Craniometry
Bizygomatic
dia. 135.0* 132.0* _ 140.0* 144.0* 138.0* _ 137.0*
With rare exceptions the crania recovered were
Total face badly damaged or fragmentary. They were recon-
height _ 100.0* _ 134.0* 131.0 113.0 _ _
- - -
structed when possible, but the resulting distortion
Facial index - 75.8* 95.8* 91.0* 81.9*
Upper face
was often sufficient to make measurements invalid.
height _ 60.0 _ 80.0 78.0 70.0 _ 68.0 Table 9 summarizes the data for the nine skulls with
Upper face
index _ 45.5 _ 57.1 54.2* 50.7* _ 49.6*
reliable measurements. The cranial indices fall into
Nasal height 54.0 44.0 - 53.0 56.0 50.0 - 50.0 the mesocranic or brachycranic ranges. More recent
Nasal breadth 21.0 23.0 - 24.0 24.0 22.0* - 27.0
Nasal index 38.9 52.3 - 45.3 42.9 44.0* - 54.0
Orbital height 33.0* 38.0 - 38.0 35.0 36.0* - 32.0
Orbital breadth 41.0* 42.0 - 40.0 40.0 42.0* - '

41.0
Orbital index 80.5* 90.5 - 95.0 87.5 85.8* - 78.2
Alveolar
breadth _ _ _ _ 62.0 _ 59.0 67.0
Alveolar length - - - - 58.0 - 51.0 55.0*
Alveolar index - - - - 107.0* - 116.0 122.0*
Estimated.
ginal foramina but no dehiscences. There are very long

mylohyoid arches on both sides of the mandible. Den-
tin is exposed centrally in the anterior teeth and in the
lingual cusps of the maxillary molars. Many enamel

pits and fissures are present on the occlusal surfaces of
molars, one of which is involved in caries. These teeth
show cervical caries. No degenerative changes are seen
in the available joint surfaces and vertebrae.
Tc 368, Zone A. This level of Coatepec contained a

few pieces of bone from two individuals, (a) a young Fig. 62. a, Virtually toothless mandible, Tc 50, Burial 4,
El Riego phase. Retained roots of both canines have exposed
child and (b) an adolescent.
pulp chambers and were surrounded by abscesses, b, Lower
Summary. From three sites we have the remains,
right dental arch, Tc 35w, Burial B, Venta Salada phase;
most of them fragmentary, of twelve adults and ten
showing carious canine and premolar, c, Approximal caries
children. Burned bone comes from three different mul-
of lower right premolars, Ts 204c, Burial 1, Ajalpan phase.
tiple burials. In Zone 2 of the West Niche of El Riego cl, Lower teeth, Tr 218, Burial 6, Santa Maria phase; central
Cave an adult female with cremated lower limbs was incisors are severely worn by a labial groove that runs
buried with four children, whose ages are so spaced downward and to the right.
103
Table 10. Measurements of Immature Skulls margins, with which they blend. Supraorbital notches
(in mm.) occur in 69 percent of twenty-six sides, the rest having
single foramina. In one case only do the supraorbital
Venta Salada Palo Blanco
Tc 35w-L2 Tc 50-1
vessels and nerve groove the surface of the frontal
(2-4 years) (15 months) bone. The infraorbital foramina appear very large and
in25 percent of the skulls in which they appear have
Length 142.0 150.0
Breadth 149.0 130.0

an accessory foramen supermedial to the usual site.
Height 107.0 115.0 In 26 percent the hypoglossal canal is divided by a
Cranial module 132.7 131.8 bony partition.
Cranial index 105.0 86.6
The commonest shape of the nasal bones is that of an
Basion-prosthion 71.0 69.0
Basion-nasion 72.0 78.0

hourglass (in 87 percent of the skulls with the facial
Min. frontal breadth 85.0 79.0* region intact) and these two bones meet each other at
Length-height index 75.4 76.6 a rounded junction in 62 percent. The nasal bones are
Breadth height index 71.9 88.4
very narrow in the earliest skulls, but are broader in
Bizygomatic diameter 101.0 92.0
Total face height 79.0 69.0

the skulls from the more recent Palo Blanco and Venta
Facial index 78.2 75.0 Salada horizons. Ninety-three percent have blurred
Upper face height 50.0 45.0 subnasal margins, as shown in Fig. .54.
Upper face index 49.5 48.9
The facial region of young and old is distinguished
Nasal height 36.0 34.0
Nasal breadth 18.0 17.0

by a prominent zygomaxillary tuberosity (see Fig. 54).
Nasal index 50.0 50.0 This and the flaring of the zygoma contribute to the
Orbital height 28.0 31.0 vertical facial profile and the straight facial contour as
32.0 32.0
Orbital breadth
seen from below, all of which is well shown in the skull
Orbital index 87.5 96.9
Ramus breadth 25.0 20.0

of Tc 50, Burial 5, illustrated in Fig. 53. This facial
Symphysis height 21.0 18.0 flatness is less prominent in more recent burials, such
Bigonial diameter 73.0 65.0 as the skulls shown in Fig. 54 b and c, where the
Ramus height 33.0 32.0
zygomaxillary tuberosity is less pronounced and the
“Approximation. facial profile is more oblique.

Of thirty tympanic plates, thickening of the lateral
margin is found in 27 percent and a dehiscence in 23
percent.
skulls tend to be rounder headed and show a slight The posterior pole of the skull is composed of an
degree of asymmetrical occipital flattening. occipital 89 percent of the sample. On the
mound in
The measurements two immature skulls are given

of interior of the occipital bone, the groove for the supe-
in Table 10 and photographs of their vault contour ap- rior sagittal sinus swings to the left on 25 percent of the
pear in Fig. 57. Te 50, Burial 1, is longer in relation to crania.
its breadth and has a bulging occipital region which is The transverse palatine suture is either straight or
more prominent on the left and is demarcated above bulges anteriorly, except for one specimen in which
by a wide supralambdoid sulcus. it bulges posteriorly. No torus palatinus is present.
Neurocranial form variations are evident in the fol- Rarer anomalies include one os inca, two skulls with
lowing skulls: Tr 218, Burial 7, has a sharply sloping a foramen spinosum which is open medially, one uni-
frontal region and an almost vertical flattened occiput. lateral occurrence of a bony spur partially dividing
Tc 272, Burial 2, the only dolichocranic individual a foramen ovale, a precondylar facet on the anterior
(cranial index 68.2), appears long and narrow as seen margin of the foramen magnum, and one ossified at-
from above, with a very prominent occipital mound, tachment of the apical ligament to this area. There are
well-marked parietal bossing, and postorbital constric- no examples of pterygo-basal bridging, paramastoid
tion. Tc 272, Burial 4, in profile has a high lump on process, metopic suture, os japonicum, or sagittal keel.
the parietals and a narrow occipital mound. Although the sample is hardly adequate, it suggests
temporal trends toward less facial flatness, broader
Morphology nasal bones, a less distinct parietal notch on the tem-
The supraorbital ridges are well marked, continuous, poral bone, and the absence of tvmpanic-plate thick-
and extend at least half way across the superior orbital ening.
104
THE HUMAN SKELETONS
The Mandibles Table 11. Measurements of Male and

Female Mandibles
Table 11 summarizes four measurements taken on (in mm.)
from four to eight male mandibles and from five to six

Male Female
female mandibles. As usual, the male measurements are
larger than the female, but taken together, the dimen- Minimum ramus breadth
Range 34-40 30-37
sions reveal no temporal trend.
Number 8 6
Generally the mandibles are quite robust with strong Mean 36.7 33.9
muscle markings, particularly at the angle. There is Height of symphysis

eversion of the angle in 55 percent of twenty individ- Range 34-41 28-40
uals and inversion in 15 percent. Gonial eversion Number 4 5
Mean 38.2 33.8

shows no secular trend (within the limits of the small
Bigonial breadth
sample, it occurs at all time levels), is not limited to
Range 92-105 86-103
males (it is found50 percent of the females), and
in
Number 6 5
is not a function of age (it is well developed in young Mean 100.4 95.2
mandibles such as that of Tr 218, Burial 5, a six year Ramus height
old). Range 57-72 55-65
Number
In most individuals, the central part of the outer 7 6
Mean 63.4 58.4

surface of the ramus bulged or roughened. The
is
external oblique line extending downward from the

anterior border of the ramus forms a rounded torus
in 71 percent of mandibles representing this area, with
no significant sex, age, or time differences in its inci- found in the fragment of an adult left mandible from
dence. The coronoid process varies in shape in cross Zone 6 of Tc 35w, illustrated in Fig. 59 d.
section. Of the sixteen mandibles with this area intact, Fig. 56 illustrates the commonest appearance of Te-
three are triangular in cross section, eight have a sharp huacan mandibles, with gonial eversion, notching of
exterior edge, three a sharp posterior edge,and two are the inferior border, and a torus on the external oblique
irregular in shape.There is no evident sex or age differ- line.
ence, but the triangular form occurs only in three early
skeletons. The Dentition
Typically in the mandibles the lower border of the
Crowding
body deeply notched anterior to the prominent gonial
is
region. This border in four mandibles is sufficiently The only evidences of crowding of teeth occur in
curved to allow the bone to rock when placed on a flat four burials from the Santa Maria phase at Quachilco:
surface. Tr 218, Burials 2, 3, 4, and 6 (Fig. 62). In each case,
Half of the mandibles have a medio-bilateral chin including Burial 3 (the skeleton of a six-year-old
form. The medial form is confined to females. Of child), there is rotation and displacement of incisors.
seven immature mandibles, three are medial in chin Burial 6 also shows labial eruption and rotation of a
form, three are medio-bilateral, and one is bilateral. lower canine. On the labial surface of the lower cen-
Four mandibles with an intact inner sur-
of the fifteen tral incisors of this individual there is a deep oblique
face of the ramus have no projecting lingula over the groove worn into the dentin of the crown. The presence
mandibular foramen. This condition does not occur of lip plugs in this phase suggests an explanation for
in the more recent skeletons. In two of the fifteen, this grooving, and perhaps for the incisor displace-
accessory mandibular foramina are present. One of ments.
these, Tc 35w, L2:5, has one large extra foramen on
Attrition
the left and two smaller ones on the right. A mylohyoid
arch appears in ten of thirty-eight sites (26 percent). The degree and pattern of wear on the teeth fall into
One specimen (Tc 50, Burial 7, Fig. 59 c) has a bridge three quite distinct categories related to cultural phase
over a mylohyoid groove which begins at a bony and so reflect dietary habits.
aperture some distance from the mandibular foramen. The “early pattern” is seen in jaws from the Ajuere-
There is only one example of double mental foramen, ado, El Riego, and Coxcatlan phases. A tough fibrous
105
the bucco-lingual plane and no enamel remains on its
occlusal surface. PM2 is sharply sloped buccally,

leaving only a small sharp lingual enamel crescent.
On the left side of the maxillary arch, PM2 slopes
buccally and all its enamel is worn aw'ay. Similarly, Ml
has no remaining enamel, but it slopes lingually. M2
and M3 show much less w'ear. Cuspal contours are still
present.
In the lower dental arch it is anterior teeth that re-
main, except for the M3, which is surrounded
tilted left
by an abscess cavity'. M2 was present but was no
Left
longer held in its socket. The central incisors are miss-
ing, but the lateral incisors are present and are mirror
images of each other, sloping sharply in the mesio-
buccal plane and with a tiny, sharp distolingual enamel
crescent as the last remnant of the incisive edge. Both
C D canines have enamel crescents on their buccal mar-
gins. Two premolars remain on the left and one on the
Fig. 63. a Sternum, Tr 218, Burial 6, Santa Maria phase,
,
right. Their enamel is completely worn aw'ay, and both
with a large oblique bony canal passing through upper lingual and buccal surfaces are convex.
right of the anterior surface of the manubrium, b, Upper The marked degree of w'ear exhibited in this early
margin of the manubrium, Tc 272, Burial 1, Palo Blanco pattern may be attributed to the coarse, abrasive diet of
phase, showing irregular, notched appearance of the calci- people who gathered a large part of their food from the
fying first costal cartilage, c. Under surface of a left rib,
natural vegetation, among which maguey and various
Tc 50, Burial 4, El Riego phase, showing expanded arthritic
cacti are prominent. The oblique occlusal surfaces
tubercle, d, Healed fractures of two ribs, Tc 272, Burial 2,
sloping in various planes w'ere the result primarily of
Coxcatlan phase.
grittv fibrous plant material being pulled obliquely
through the teethto strip aw'ay the more edible tissues.
With advanced attrition, the opening of pulp chambers,
gritty diet has severely ground away the crowns of the and the resulting loss of teeth, the changing occlusal
teeth, exposing dentin and eventually the pulp cham- relationships became secondary causes of the erratic
ber. The usual characteristics of the pattern are the appearance of the dental pattern.
sharply sloped, obliquely worn surfaces of the teeth, An intermediate pattern is found in the teeth of the
the confusing appearance of a dental arch in which Ajalpan phase. Generally, the degree of attrition of
each tooth slants in a different direction, and the sharp these full-time agriculturalists continued to be high, per-
crescents of enamel remaining on the occlusal surface haps owing in part to the abrasive contribution of the
after the rest of the crowm has been sheared away. stone mortars in which food w'as ground. There are no
These characteristics are illustrated in Fig. 60. longer, how'ever, signs of the oblique w'ear and pro-
The complexity of this pattern of attrition can best jecting enamel crescents of the earlier period. Crow'ns
be appreciated by describing the state of each tooth of premolars and molars are worn, but wom flat. There
in one of these early skulls, for example, Tc 272, Burial is some evidence that anterior teeth w'ere subjected to
2. The only upper teeth remaining are the last four on considerably less attrition than the grinding posterior
each side of the arch and the left canine, which is worn teeth. These are characteristics usually associated with
past the cemento-enamel junction. On the right side, the dentition of hunting and gathering populations in
M3 still show's cuspal contours, but the mesiolingual other areas of the New World.
cusp is w'orn, exposing dentin. M2 still has enamel- The later pattern of attrition is shared bv the indi-
covered rounded buccal cusps, but the lingual half viduals from the Santa Maria, Palo Blanco, and Venta
of the tooth is sharply tapered, exposing dentin all the —
Salada phases or with the dentition of most agricul-
way dowm uncovered lingual root. A crescent
to the tural peoples. Here, the amount of w'ear is slight. Molar
of enamel remains on the distal side of the occlusal and premolar cusps are blunted by occlusal facets. At-
surface. The tapered lingual surface formerly occluded is usually somewhat greater in the anterior teeth
trition
with the buccal slope of low'er M2. Ml is biconvex in —characteristically with a thin line of dentin exposed
106
THE HUMAN SKELETONS
in the biting edge of the incisors and, in older indi- We would, of course, anticipate a rise in the caries
viduals, a dot of dentin surrounded by the enamel of rate with the transition to agriculture, because the
the blunted canine cusp. high carbohydrate diet provides an encouraging en-
Fig. 60 d is a photograph of the teeth of an adult of vironment for oral bacteria and because the softer diet
the Venta Salada phase showing no attrition except for produces little wear on the teeth to erase pits on the
small facets on some molar cusps. enamel surface that serve as likely sites for the begin-
ning of caries.
Tooth Loss When the incidence of caries is calculated for the
Postmortem tooth loss is high in disturbed and sec- various periods, the changing pattern shown in the ac-
ondary burials, particularly in Venta Salada phase companying tabulation is found. As expected, the teeth
skulls from both niches of El Riego Cave. from early periods are rarely involved. Indeed, only
The incidence of tooth loss owing to dental pathol- two cases of caries are present in the El Riego and
ogy shows an interesting gradient in time. The num- Coxcatlan material: one is approximal caries resulting
ber of teeth lost premortem during each phase was from the traumatic impaction of food between two
totaled and expressed as a percentage of the number teeth, and the other occurs in an occlusal pit on an un-
of tooth sites in the available jaws. The accompanying opposed third molar where attrition could not eradicate
tabulation shows the high incidence of tooth loss in the a pre-carious site.
hunting and gathering economies decreasing with the
adoption of agriculture. The dramatic decrease in tooth
% of
Phase Incidence Caries
loss between the last two groups, both full-time agri-
culturists, is correlated with a significant, although not El Riego 1/24 4.2
Coxcatlan 1/38 2.6
as spectacular, decrease in caries, which will be de-
Santa Maria 15/87 17.2
scribed below.
Venta Salada 11/96 11.5
% of
Phase Incidence
Although the incidence of caries does increase with
tooth loss
the transition to agriculture, the increase is not as great
El Riego, Coxcatlan 57/137 41.6
as anticipatedfrom findings on other sites, or as would
Ajalpan 6/16 37.5
be expected from the large number of enamel pits on
Santa Maria 34/125 27.2
the molar surfaces. Then too, the decrease during the
Palo Blanco, Venta Salada 8/128 6.2
Venta Salada phase calls for an explanation. The an-
As the incidence of tooth loss changes, so does its swer probably lies in the mineral-rich water of the
etiology. In the early phases, the causal villain was a valley, the soils of which are deposited on the teeth as
combination of an abrasive diet and chewing habits a heavy calculus which effectively plugs potential car-
which initiated the sequence leading through attrition, ious sites. Most of the Venta Salada teeth come from El
traumatic occlusion, exposure of the pulp chamber, and Riego Cave and have a considerably heavier calculus
the introduction of micro-organisms, to the loss of the deposit than teeth from other sites. In preparing the
tooth from its infected alveolus. The adoption of agri- specimens, the enamel pits were found only after con-
culture provided a softer diet and substituted caries for siderable work in scaling the teeth.
attrition as the prime cause of tooth loss.
Periodontal disease, as diagnosed by porosity of the The Infracranial Skeleton
alveolar bone and recession of the alveolar margin with
Measurements
exposure of the roots of teeth, is more prevalent in the
earlier phases. As would also be expected, the incidence Accurate measurement was possible on relatively
of alveolar abscesses decreases with time. few limb bones because of damage to the delicate
articular ends.
Caries
The formulae of Trotter and Gleser (1958) for Mon-
It is difficult to devise a meaningful way of report- goloid skeletons were used for estimating stature. The
ing the incidence of caries because we cannot be cer- results aresummarized in Table 12. Since one of the
tain of the proportion of lost teeth that were shed be- seven females and one of the eight males for whom
cause of caries. In this study, the incidence is reported stature was estimated fell well outside the range of the
as a percentage of the teeth still remaining in the remaining estimates, these two figures were omitted in
mouth. arriving at the average statures for each sex. Heights
107
Table 12. Estimated Statures The scapulae typically have a convex medial border,
a blunt inferior angle, a rectangular acromion, and a
Bones Used Stature in Cm. Stature in Inches
supraspinous part that forms a low triangle. They also
FEMALE have a suprascapular notch, fairly deep in individuals
Tc 35w-A Ulna 161.3 ±4.66 63.6
from the earlier levels and almost indiscemable in
Tc 35w-D Radius 156.8 ±4.6 61.7
Tc 35w-L2:l Humerus
those of the Venta Salada period.
160.1 ±4.25 63.0
Tc 35e-B:l Humerus 156.1 ±4.25 61.4 A septal aperture is present in about half of the fe-
Tc 50-5 Fibula 159.4 ±3.24 62.8 male humeri but never in the male humeri. There are
Tc 272-1 Fibula 157.4 ±3.24 62.1
no cases of third trochanter of the femur or vastus
Tc 272-4 Ulna 167.8 ±4.66 66.0
notch on the patella. The so-called squatting facets are
MALE
Tc 35w-C Femur and fibula 163.8 ±3.18 64.5
neither so common nor so prominent as in most series
Tc 35w-E Radius 166.3 ±4.60 65.5
of American Indian tali. In half of the individuals the
Tc 50-4 Tibia 165.1 ±3.27 65.0 anterior and middle talar facets on the calcaneus are
Tr 218-2 Fibula 157.3 ±3.24 62.0
joined.
Tr 218-4 Tibia 171.8 ±3.27 67.7
Tr 218-7 Radius
Of eleven intact atlas vertebrae, three have bilateral
165.5 ±4.60 65.2
Tr 218-8 Radius 168.8 ±4.60 66.4

anomalous foramina in the posterior arch, and two have
Tc 272-2 Femur 168.8 ±3.80 66.4 a bridge over the vertebral artery. There are a few
examples of double foramen transversarium, but dam-
age has reduced the number of intact cervical verte-
brae to an insufficient sample. Two vertebral columns
for the remaining six females ranged between 156.1 have accessory' articular facets between L5 and SI.
and 161.3 cm. and averaged 158.5 cm., or 62.4 inches. Both individuals come from the Santa Maria level at
The seven male statures ranged between 163.8 and Quachilco.
171.8 cm. and averaged 167.2 cm., or 65.8 inches. If the
estimated statures of Tc 272, Burial 4, and Tr 218, Pathology
Burial 2, are included — the first being a rather tall fe- Pathological specimens from the Tehuacan sites may
male and the second a very short male the average for — be classified into four categories: trauma, infections,
each sex comes out the same, 166.0 cm., or 65.3 inches. degenerations, and congenital abnormalities.
A low intermembral index of 66 indicates that the
legs are rather long in proportion to the upper limbs.
Trauma
The brachial index is high, ranging from 74 to 80, indi- The skeletons of twenty-one adults are sufficiently
cating a relatively long forearm. For this index, Comas complete to determine the presence or absence of in-
(1960) reports a range of 73.2 to 74.5 in Europeans and jury to all of the body parts. Of these, five show evi-
76.9 to 78.2 in American Indians. The crural index of dence of healed fractures of a total of eighteen bones:
approximately 87 indicates a relatively short femur in one scapular body, shafts of two radii and one ulna,
relation to tibial length. one fibular shaft, one fifth metatarsal, five ribs, and
Platycnemic indices reflect the minimal amount of seven vertebrae.
tibial shaft flattening present. Values range from 58.7
to 74.2, with a mean of 67.1. Indices below 63.0 are
considered platycnemic. Platvmeric indices range from
71.9 to 96.0. A gradual increase in this index with time Table 13. Length of Immature Diaphyses
shows a tendency to reduction in the degree of femoral (in mm.)
shaft flattening. The mean for the El Riego period is
75.0, for the Santa Maria period 78.5, and for the Venta Tc 50-2 Tc 50-3 Tc 50-6 Tc 35w-L2 Tc 35w-L2 Tr 218-3 Ts 368-3
(5 yrs.) (under (under (2-4 yrs.) (5 yrs.) (8 yrs.) (24 mos.)
Salada period 82.2. 6 mos.) 6 mos.)
The length of the diaphvses of immature skeletons

Clavicle 81 48 43 63 _ _ -
was measured to aid in determining age. These figures -
Humerus 148 69 62 108 130 167
are given in Table 13. Radius - - 83 104 129 88
56
Ulna - -
64 59 93 114 145 95
Morphology Femur 199 78 75 141 167 228 -
Tibia 166 67 60 117 140 - 117
This section summarizes the incidence of certain
Fibula 160 63 - 113 138 188 -
variations in the bones of the infracranial skeleton.
108
THE HUMAN SKELETONS
Infection
The high incidence of mandibular abscesses follow-

ing exposure of the chamber in severely worn teeth
has already been described.
The most spectacular result of infection in this collec-
tion is the eroded skull of Tr 218, Burial 2 —the result
of treponematosis (Fig. 58 c and d). The compound
nature of the fractured scapula of Tc 50, Burial 5, is
confirmed by the multiple areas of infection along the

fracture lines. One further evidence of infection is the
slight periostitis of the tibial shaft in Tc 35w, Burial B.
Degeneration
Degenerative joint disease is evident in a very high

percentage of the adult Tehuacan skeletons. Osteophy-
tosis of the vertebral bodies is present in ten of four-
teen intact columns, seven showing advanced stages
of the process (Fig. 64). Of these, four have involve-
ment of the apophyseal joints. In Tc 50, Burial 5, arth-
Fig. 64. a, Arthritic lipping of the dens facet of the atlas ritic lipping has encroached upon the foramen trans-
and a fan-shaped apex of the dens, Tc 50,
exostosis of the versarium of cervical vertebrae with resulting occlu-
Burial 4. b, C5—7 from the same
showing advanced
burial, sion of the vertebral artery on one side.
arthritic involvement of the apophyseal joints, c, Segment
of the cervical vertebral column, Tr 218, Burial 2. Large
osteophytes project downward from the front of the body of
the axis. C3 and C4 are fused, d, Advanced arthritis and
osteophytosis of three cervical vertebrae, Tc 50, Burial 5.
Lipping from the uncovertebral joint encroaches on the
foramen transversarium and on the left side of the fifth
vertebra completely occludes it.
Of the five skeletons from the two earliest phases

represented (El Riego and Coxcatlan), four have healed
fractures: Tc 272, Burial 4, Colles’ fracture of the left
radius (Fig. 68 b ); Tc
272, Burial 2, two ribs (Fig. 63 d);
Tc 50, Burial 4, a wedge-shaped compression fracture
of the third thoracic vertebra (Fig. 65 c) and fractures
of the fifth metatarsal and shafts of the left radius and
ulna (Fig. 68 a); Tc 50, Burial 5, the left scapula (Fig.
67), three ribs, and six vertebrae, two of which have
wedge deformities (Fig. 65 d ).
In contrast, only one fracture (a left fibular shaft in
Tc 35w, Burial E) appears in the sixteen adult skeletons
Fig. 65. a, Atlas, Tr 218, Burial 4, seen from behind.
from the three later phases: Santa Maria, Palo Blanco,
The marker at left is an anomalous posterior arch fora-
in
and Venta Salada.
men; on the right, an incomplete lateral bridge over the
The marked decrease in evidences of trauma no
vertebral arterv. b, L5 of Tc 50, Burial 2, a 5-year-old child.
doubt reflects the adoption of a settled way of life and The neural arches have failed to fuse on this and L4 and on
an easier one. It should be noted that among the many all of the sacral segments: a case of spina bifida, c. Com-
fractures of the earlier periods none is of a lower-limb pression fracture of T3, Tc 50, Burial 4; an osteophyte has
long bone; a person with such an injury probably fused its body anteriorly to the vertebra above, d, Compres-
would not have made it back to the cave. sion fracture of T12, Tc 50, Burial 5.
109
been recovered. The Tehuacan series, because of its

and long span of time, cannot produce
size, condition,
valid metrical data, although the more important mor-

phological observations are available for comparison.
Unfortunately, however, very little morphological de-
scription has been published for the more recent osteo-
logical populations, although metrical data are avail-
able. A preliminary survey of some of the comparative
material will at least set the stage for future attempts
to answer these questions.
Because of its Mexican location and probable an-
tiquity, it is natural that our first comparisons should be

with Tepexpan man (De Terra et al. 1949). The accom-
panying tabulation fists the measurements in milli-
meters of the Tepexpan skull and an early male skull
from Tehuacan, Tc 50, Burial 4 (El Riego phase). The
two are very similar metrically differing chiefly in the
somewhat smaller size of Tc 50, Burial 4, and in the
broader nasal region of Tepexpan, in which it resem-
Fig. 66. a, Front ofL5 and sacrum, Tr 218, Burial 4. The
bles some of the more recent Tehuacan skulls. Neither
sacrum has onlv four elements and there are only five
has a sagittal keel, and both have heavy V-shaped
lumbar vertebrae, b, Side view of L5 and sacrum, Tr 218,
Burial 7. Both specimens show accessorv articular facets.
supraorbital ridges and a mound-shaped occipital
c , Anterior and, d, posterior views of the lumbosacral re- region. Dental conditions make it impossible to esti-
gion, Tc 50, Burial 4. There is spondvlolisthesis of L5. mate facial height and the degree of prognathism. In
both, the nasal bones appear pinched and the inferior
nasal margin is indistinctlv demarcated from the face.
Eleven of nineteen adult skeletons show the lipping,
erosion, and ebumation of osteoarthritis in some or all of T epexpan Tc 50-4
the joints. Examples are illustrated in Figs. 63 and 69. Cranial length 179.0 179.0
Cranial breadth 143.0 137.0
Congenital Abnormalities Basion-bregma 136.0 126.0
Abnormalities sufficiently severe to be considered Cranial module 152.6 147.3

Cranial index 79.9 76.6
pathological appear in three skeletons. Tc 35w, Burial
Length-height index 76.0 70.4
D, had bilateral congenital hip dislocation shown by
Breadth-height index 95.1 92.0
the shallow acetabulum with deficient margin (Fig. 69
Minimum frontal breadth 99.0 86.0
a). Spondylolisthesis of the fifth lumbar vertebra is evi-
Bizvgomatic diameter 140.0° 135.0°
dent in Burial 4 of Tc 50 (Fig. 66 c ). Spina bifida of the Orbital height 34.0 33.0°
last two lumbar and all of the sacral vertebrae is evi- Orbital breadth 40.0 41.0°
dent in Burial 2 of Tc 50 (Fig. 65 b). Orbital index 85.0 80.5°
Nasal height 49.0 54.0
Comparisons Nasal breadth 25.0 21.0
Nasal index 51.0 38.9
In comparing the Tehuacan material with that of
Ramus breadth 36.0 36.0
other sites, we would hope toanswer four questions:
Ramus height — 58.0
How do the earlier Tehuacan skeletons differ from
Bigonial diameter
—
io b0 92.0
those of recent American Indians? Do they resemble ° Estimated.
recent Mexican material more than other New World
groups? Is there similaritv between specimens of early Both have prominent zvgom axillary' and malar tuber-
man from various localities in the Americas? Do dis- osities and a vertical profile of the lateral face. The
tant affinities show in comparison with Asian skele- mandibles are similar, with low, broad ramus, gonial
tons of the same time level? It is disappointing that an- eversion, a notched inferior border of the body', and a
swers to these questions cannot yet be found. There prominent external oblique fine. Neither has a palatine
are few sites from which skeletons of early man have or mandibular torus.
110
THE HUMAN SKELETONS
Although there has been considerable tooth loss and

alveolar resorption inTepexpan man, the remaining
teeth show the advanced oblique attrition with occlusal
enamel crescents which is characteristic of the early
Tehuacan teeth.
Similar dental conditions are found in the cranial
material from Santa Maria Astahuacan (Romano 1955),
which has a radiocarbon date of 9670 b.p. ± 400 years.
A single mandible has lost a third molar and both cen-
tral incisors. Four teeth have sharp lingual enamel
crescents, while on the left first premolar attrition has
Table 14. Comparative Data on New World Skulls

(in mm.)
Length- Breadth-
Cranial height height Nasal Orbital
index index index index index
Tehuacan Tc 50-4 76.6 70.4 92.0 38.9 80.5
Tepexpan 79.9 75.9 95.1 51.0 85.0

Brown's Valley 73.5 73.6 100.0 46.5 92.8
Confins 69.2 80.0 114.0 48.9 94.2
Sauk Valley 74.2 73.3 98.6 46.5 80.5
Pun in 71.0
Fig. 68. Healed fractures of midshaft of left radius and
a,
66.0 94.0 59.6 69.2
Lagoa Santa 70.7 74.3 104.7 50.7 86.4
ulna, with angulation and considerable callus formation,
Tc 50, Burial 4. b, Healed Colles’ fracture of left radius,
Tc 272, Burial 4. c, Osteophytosis of bicipital tuberosity of
left radius of Tc 50, Burial 5. d, Pseudoepiphyses of both
first metacarpals in a child, Tc 50.
removed all the enamel and exposed the pulp cham-

ber. As in the Tehuacan specimens, oblique wear varies
in its plane from tooth to tooth: buccal, distal, disto-
lingual, and mesiolabial.

A
male skull from the same site resembles Tc 50,
Burial 4, in having V-shaped brow ridges, an occipital
mound, gonial eversion, and a blurred subnasal mar-
gin. It differs, however, in its slight sagittal keel and
in the absence of the characteristic facial flatness.
There is no metopic suture, mylohyoid arches, or mul-
tiple foramina, but there is a dehiscence in each tym-
panic plate.
Table 14 compares five metrical indices of certain
skulls from the New World to which considerable an-
tiquity has been ascribed. Metrically, the skull of Tc
50, Burial 4, resembles Tepexpan and Sauk Valley more
Fig. 67. Healed fracture of the body of the left scapula,
closely than the others. The skull of Brown’s Valley
Tc 50, Burial 5. The dorsal a, and costal surfaces, b, show
the triradiate fracture line, central defect, and multiple
Man from Minnesota (Jenks 1937) is longer and higher,
abscess holes, c, Involvement of the lower part of the with a broader nasal region and taller orbits. Although
glenoid fossa and growth of osteophytes into the triceps there is slight gonial eversion, in most respects the
tendon, d, The underlying three ribs show healed fractures. mandible does not resemble the Tehuacan pattern in
See also Fig. 64, d. that it has a tall narrow ramus and there is no thicken-
111
been published. The cultural context is the earlv Post-

classic period, with a date of approximately a.d. 800.
Because of certain similarities to the Tehuacan ma-
terial, it will be described brieflv here.
The skeleton is that of a twentv-five-year-old female.
Its stature of 158.1 ± 3.24 cm. places it within the range
of Tehuacan females. Limb proportions are as in the
Tehuacan skeletons: a relatively long forearm (brachial
index, 79.6), short femur (craral index, 68.2), and long
legs in relation to upper limbs (intermembral index
68.2). There is no flattening of the tibial shaft (plat-
ycnemic index 70.0), but the femoral flattening is
shown in a platymeric index of 74.3.
Tehuacan
El Risco Tc 50-5
Fig. 69. a. Right acetabulum, Tc 35w, Burial D, is elon- Cranial length 172.0 174.0
gated, shallow, and has a deficient anterior margin; the Cranial breadth 139.0 136.0
condition is Lipped and ebumated head of first
bilateral, b, Basion-bregma 134.0 136.0
metatarsal, Tc 50, Burial 5. c. Middle phalanx of left third Cranial module 148.4 148.7
finger of the same burial, lipped proxrmallv to accept an Cranial index 80.8 78.3
sccessorv ossicle, d. Trapezoids, Tr 218. Burial 4; at left, the Length/Height index 77.8 78.2
“toe of the boot” is expended bv a porous osteophvte. Breadth/Height index 96.4 100.0
Minimum frontal breadth 84.0 85.0
Bizvgomatic diameter 129.0° 132.0°
Orbital height 33.0 38.0
ing of the external oblique line or notching of its in-
Orbital breadth 39.0 42.0
Orbital index 84.7 90.5
The subnasal margin is sharp. Confins
ferior border.
Xasal height 50.0 44.0
man (Walter et al. 1937), which was discovered a few
Xasal breadth 26.0 23.0
miles from Lagoa Santa in Brazil, differs in having a
Xasal index 52.0 52.3
longer and higher vault, a broader nasal region, and
Ramus height 66.0 60.0°
almost square orbits. An examination of the published Ramus breadth 37.0 37.0
photographs, however, revealed certain morphological Svmphvsis height 37.0 31.0°
similarities to the Tehuacan material: a prominent Bigonial diameter 93.0 101.0
zygomaxillarv tuberosity, vertical malars flattened cor- “Estimated.
onally, and mandible with short broad ramus and
a
gonial eversion. Subnasal prognathism is spectacular. The close eraniometric similarity is illustrated in
Except in the anterior teeth, attrition is not far ad- the accompanving tabulation, which compares El
vanced. Risco with the female skull from Burial 5 of Tc 50, an
The Punin Calvarium (Sullivan and Heilman 1925) El Riego phase skeleton. Measurements are in milli-
from central Ecuador differs greatlv from Tehuacan, meters. Morphologicallv. El Risco shows the following
not onlv in its longer vault, broader nasal region, and similarities: V-shaped supraorbital ridges, an occipital
low rectangular orbits, but in its scaphoid vault and mound, blurred subnasal margins, a thickened tym-
the absence of malar verticalitv. The Sauk Valley skull panic plate, and a rather flat malar region. The Tehua-
Jenks and M ilford 1938), although similar to Tehua- can mandibular pattern is also present, with a short,
can in the indices reported, differs in showing marked broad ramus, a torus on the external oblique line, a
postorbital constriction, a more acute frontal slope, a notched inferior border, and gonial eversion. It differs
slight sagittalcrest, and a much taller mandibular in its verv wide nasal bones, a somwhat scaphoid vault,
ramus (71.5 mm. vs. 58.0 mm. in Tehuacan). and in the absence of a zygomaxillary tuberosity. There
During the excavation of the El Risco site in the is considerable alveolar prognathism. Attrition has
Valley of Mexico (Maver-Oakes 1959) a complete flattened the occlusal surface of the teeth, exposing
skeleton was recovered, a report of which has not yet dentin in the base of the cusps.
112
THE HUMAN SKELETONS
Among New World populations outside of Mexico, and concavo-straight inferior margin of mandible.
the Tehuacan morphological complex found most
is Eskimo characteristics not seen in the Tehuacan skulls
closely, strangely enough, among the Eskimo. Of nine are: sagittal keel, palatine torus, and mandibular torus.
qualities listed as characteristic of Eskimo (Oschin- Although there is considerable similarity in the mor-
sky 1964), six are typical of the earliest skulls from phology of the face, mandible, and temporal bone,
Tehuacan: narrow nasal bones, vertical malars, thick Eskimo crania are considerably larger, longer, and
tympanic plate, gonial eversion, short broad ramus, more rugged than those from Tehuacan.
REFERENCES
Anderson, J. E., 1963. Mayer-Oakes, 1959.
Ashley Montague, 1960. Oschinsky, 1964.
Comas, 1960. Romano, 1955.
Genoves, 1960, 1962, 1964. Sullivan and Heilman, 1925.
Jenks, 1937. Terra, Helmut de, 1949.
Jenks and Wilford, 1938. Trotter and Gleser, 1958.
MacNeish, 1961a, 1962, 1964b. Walter et al., 1937.
113
CHAPTER 7
Codex Borgia and the Venta Salada Phase
Robert Chadwick and Richard S. MacNeish
Archaeological surveys of the area bordering the peoples of Mexico, among them the Aztecs, Mayas,
/ \
Tehuacan Valley demonstrate the existence of Zapotecs, and Mixtecs, painted such books in late pre-
/ mremains of the terminal aboriginal culture, Hispanic times, or during the years immediately fol-
known archaeologieally as the Venta Salada phase, lowing the Spanish conquest. Some of the latter bear
within a rather well-defined region that is approxi- comments written in Spanish or Nahuatl to explain
mately coterminous with the pre-Conquest domain of the pictures.
the Senorio de Teotitlan del Camino, an independent A few of these folding books have, for reasons of
political enclave within the Aztec empire. For this content or been linked by some scholars with
style,
reason, we assume documentary information
that that part of Mexico in which the Tehuacan Valley lies
about the people of the Tehuacan Valley and the and at least one has been linked with the Tehuacan
Senorio de Teotitlan may supplement our knowledge Valley itself (Seler 1963: 101-103). The relevant books
of the Venta Salada phase. consist of six manuscripts dealing with ritual calendri-
Despite the importance of the area both politically cal matters which have been called “the Borgia
and ceremonially at the time of the Conquest, docu- group” because the Borgia Codex is chief among them.
mentary material dealing directly with Tehuacan or the The other five are the codices Laud, Fejervary-Mayer,
Senorio is scant (Nicholson 1955: 117). The valley was Cospi, Vaticanus 3773, and Mexican Manuscript Num-
not the subject of intensive ethnographic investigations ber 20 of the Biblioteque Nationale in Paris (Robertson
such as those conducted by Burgoa in Oaxaca or by 1966: 298). The Codex Porfirio Diaz has
reverse of
Sahagun in the Valley of Mexico. Some information been placed same group by Nicholson (in press).
in the
about the region can, however, be garnered from such Besides a common hieroglyphic system and a common
sixteenth-century sources as Ixtlilxochitl, Torquemada, concern with ritual matters, a distinctive graphic style
Motolinia, Roman y Zamora, Herrera, and the Rcla- — shared with a group of genealogical and historical
ciones geogrdfcas of Coxcatlan and of Teotitlan del codices of acknowledged Mixtec derivation —has pro-
Camino. Twentieth-century studies by Eduard Seler, vided grounds for classifying these manuscripts to-
Francisco del Paso y Troncoso, Joaquin Paredes Co- gether and for attributing to them a Mixtec origin
lin, Gorgonio Gil, Robert H. Barlow, Carmen Cook de (Robertson 1966: 309).
Leonard, Eduardo Noguera, Henry B. Nicholson, and During a perusal of the codices of the Borgia group
Alfonso Caso have brought to light data that bear we noted that pottery vessels depicted in Codex Borgia
upon the ancient history of the Tehuacan Valley. are similar to types characteristic of the Venta Salada
Better for our purposes than either ancient or mod- phase. Further, most of the depicted Venta Salada pot-
ern secondary sources are the pictorial manuscripts or tery types occur in the general region of the Senorio
“codices” dealing with history and religion that were de Teotitlan, and are characteristic of that region and
painted by the Indians themselves. Many indigenous cultural phase. They do not occur farther south in the
114
CODEX BORGIA AND THE VENTA SALADA PHASE
contemporaneous Monte Alban V ceramic complex in

Oaxaca, or to the north in Cholulteca I-III of Puebla
and adjacent Tlaxcala, or in Aztec periods of the Val-
ley of Mexico. Such pottery types have not been dis-
covered east of the Tehuacan Valley in such cultural
manifestations as Cempoala I-IV, Isla de los Sacrificios,
or Upper Cerro de las Mesas in Veracruz. No excava-
tions have yet been carried on to the west of the Te-
huacan Valley.
Accordingly, we began a comparison of material
items depicted in Codex Borgia and artifacts represen-
tative of the material culture of the Venta Salada
phase. As a result, we have come to the conclusion
that the Borgia and the Venta Salada phase are closely
related. It seems probable that Codex Borgia originat-
ed in the Senorio de Teotitlan, with the culture char-
acterized as the Venta Salada phase, and quite possi-
bly within the Tehuacan Valley itself. If our assump-
tions are correct, the scant documentary information Fig. 70. Codex
Variations of the glyph Flint Knife from
concerning the Venta Salada phase can be augmented Borgia, pis. 2 and and similar glyphs on sherds of
15,
by data derived from one of the most ornate and com- Teotitlan Incised pottery from an important Venta Salada
site.
plete prehistoric religious documents of central Mexico
— Codex Borgia.
Before we present the evidence for connecting origin. Seler first thought that it derived from the Gulf
Codex Borgia with the Venta Salada culture of the Coast but later came to believe that was painted in
it
Tehuacan region, we should say a few words about the area of Tehuacan-Coxcatlan-Teotitlan del Camino.
the codex itself. The earliest knowledge concerning it Caso, struck by similarities between the Borgia and
dates from the late eighteenth century, when it was altar paintings discovered at Tizatlan, Tlaxcala —par-
owned by Stephano Cardinal Borgia in Veletri, Italy. ticularly a representation of the god Tezcatlipoca — was
Through him, it came to the Library for the Propaga- convinced of a Tlaxcala-Puebla origin. Cholula re-
tion of the Faith in Rome. Details of the manuscript ceives strong support from other authors, and still oth-
were first published by Baron von Humboldt in 1813 ers point to the Gulf Coast as a possible provenience.
and its first was by Lord Kingsborough in 1830.
edition We believe that the strongest arguments are those sup-
The first truly important study of the document was porting a Nlixtec attribution, but to review all the evi-
undertaken by the German scholar Eduard Seler dence for or against the various theories is beyond the
(1904-1909). Recent scholars, including Caso, Nichol- scope of this paper. Our major purpose is to point out
son, Robertson, and others, have added further to the observations that have brought us to believe that
the analysis of the codex. Codex Borgia was painted by people whose material
The Borgia manuscript is painted on both sides of culture conforms to the Venta Salada phase within
a strip of animal hide, probably deerskin, made of the area of distribution of that culture, the Senorio of
fourteen separate pieces giving a total length of about Teotitlan del Camino, in which the Tehuacan Valley
ten meters and folded into thirty-nine leaves each 26.5 is included.
by 27.0 centimeters (Lehmann 1905: 251). The codex is A glyph for Flint Knife engraved on a potsherd
for the most part a book of rituals and ceremonies and from a large Venta Salada site (Tr 75) near Tilapa
probably served as a manual for priests and diviners constitutes significant evidence of a close connection
(Nicholson: in press). Although its glyphs have been between Codex Borgia and the Venta Salada phase of
deciphered, some of its complex ritual scenes have not the Tehuacan cultural sequence. The symbol engraved
yet been fully interpreted. The codex itself (CB 1963) on this sherd (Fig. 70, left) is identical to one used
and a Spanish translation of Seler’s original commen- throughout the codex and featured as the central orna-
tary (Seler 1963) have recently been published in ment on the breast of the god Tezcatlipoca in Borgia
Mexico. plate 17. Although codices of the Borgia group and the
Various authors have speculated as to its place of Mixtec genealogical codices indicate the day Flint
115
BORGIA GROUP
MEXICAN MS 20
OTHER CODICES
NUTTALL
Fig. 71. Glyphs for Flint Knife from various codices.
116
Knife by a number of variations of the glyph, all dis- tangles and the rim of the vessel. In a very few cases
playing a general resemblance but differing greatly in the incised element consists of a stepped design filled
detail, this particular form can be found only in Codex with crosshatching.
Borgia. A second sherd from the same Venta Salada site The double-outlined rectangles or triangles, usually
(Fig. 70, right) also is engraved with a glyph. This with a circle in the center, are also used in other ways
glyph resembles somewhat another Flint symbol in the in the codex: in plates 12-19, 24-26, 51, 55-59, 61-72,
Borgia but is perhaps closer to variations appearing in and 76 as part of the decorative devices on garments;
Codex Laud and illustrated in Fig. 71, together with in plate 21 as part of the decoration of a ball-court
glyphs from several related codices. marker; in plates 23, 40, and 71 in solar disks: in plates
Another indication that the Borgia may have been 25, 60, 64, 71, and 76 in shields; in plate 49 in a tree;
painted in the area of Venta Salada culture is the and in plates 51, 55, and 56 on staffs. Although this
probable date symbol on the sculptured stone temala- inventory is by no means exhaustive, its importance lies
catl, or gladiatorial stone, from Tehuacan, now in the in the fact that dominant elements of design on the
regional museum in Puebla. This large stone undoubted- ceramic type, Teotitlan Incised, are also among the
ly came from a site on Cerro de la Mesa, north of town. most popular designs in Codex Borgia.
It bears a “ray” symbol that in form closely resembles These designs in the Borgia appear on pots of two
the year symbol in Codex Borgia (pis. 27, 28, 51, 52, 71), colors —
yellow-brown and gray, the colors of Teotitlan
which is the only manuscript of the Borgia group to Incised, although this typeis more streaky black than
show symbol (Nicholson: in press). The

this particular gray. Thus the close connection between the Borgia
symbol differs from the usual “A-O” version of the Mix- pots and archaeological pottery of the type Teotitlan
teca in that it lacks the “O” element interlacing the Incised is demonstrated bv the use not only of similar
“A” or “ray” element. According to Nicholson, this ren- decorative elements but also of similar colors.
dition also resembles early forms of the symbol found The correspondence between the forms of vessels
at Teotihuacan, Texmelincan, and occasionally, Mon- shown in the Borgia Codex and those of archaeological
te Alban. The similarity between the representation examples of Teotitlan Incised from the Venta Salada
of this symbol on a sculpture from a Venta Salada site phase is remarkably close. These similarities are seen
and representations of the same symbol in Codex in the range of forms and appendages and the relative
Borgia suggests a close relationship between the man- frequency of each. The most popular form of Teotitlan
uscript and the Tehuacan Valley. Incised vessel, both in archaeological examples and in
Other Venta Salada artifacts, of which the most con- Borgia drawings, is a tripod flat-bottomed bowl with
vincing are various pottery types, also provide evi- outdaring walls. Hemispherical bowls or cups are rela-
dence of a connection between the Venta Salada phase and flat-bottomed, flaring-necked
tively rare, ollas with
and Codex Borgia. The diagnostic Venta Salada pot- two loop handles are extremelv rare.
tery type, Teotitlan Incised, seems to be dominant Some types of vessel supports shown in the codex
among pottery forms depicted in Codex Borgia. Ac- —stepped, animal head, and dumbbell or “bulb”—are
cording to present archaeological evidence, this pottery identical with three styles found in the pottery of the
is not a resident type outside the region of the former Venta Salada phase. A preliminary survey shows that
Senorio de Teotitlan del Camino. It has not been re- this inventory of vessel supports is seen only in the
corded for Cholula in central Puebla, Cempoala in Borgia Codex and in no other.
central Veracruz, or Monte Alban V in Oaxaca. The The ceramic type Coxcatlan Red, occurring as flat-
most common decoration of Teotitlan Incised consists bottomed bowls with distinctive stepped feet and ver-
of two or three straight parallel lines forming a hori- may be seen on Borgia
tical to slightly flaring walls,
zontal band. Often two sets of horizontal parallel lines plates 8 and 60. Red wares with stepped feet are not
separated by vertical lines form rectangles. Sometimes, to our knowledge found outside the Venta Salada
instead of vertical lines, diagonal lines divide the bands culture province.
into diamonds or triangles. One or more small carved Among themore common pottery types in Codex
(
circles, squares, or spirals appear in about half the Borgia is one which “V” elements decorate a vessel
in
areas so delimited. The circle in a rectangle or diamond of tall cylindrical form (pis. 6 and 18). We believe the
drawn with double lines is the most common decoration V’s represent the conical applique of Coxcatlan Coarse,
!
on the Borgia pots. In a few instances, in both the a type in which large cylindrical vessel forms are com-
Borgia and Teotitlan Incised, a band of “U” elements, mon. The codex also shows this type with spherical
lying horizontally, is placed between the band of rec- feet (pis. 11, 14, 23, 63, and 65). Examples of this pot-
117
Fig. 72. Sherds of Teotitlan Incised and vessels from Codex Borgia, pis. 3, 4, 7 0, 67, 69, 65.
tery and depictions of it in the Borgia usually have
a white surface finish which is sometimes decorated
with red bands. Archaeological specimens of similar
vessels have been reported only from Cholula and the
Venta Salada region.
Another distinctive pottery type, Coxcatlan Black on
Orange (yellow in the codex), appears as flat-bottomed,
out-flaring bowls, occasionally with loop handles, and
is characterized by black stepped designs, usually filled
with crosshatching. A classic example of this type is
shown in the middle panel, right side, of Borgia plate

57.
Less striking resemblances are seen in three other
wares. One of these is Coxcatlan Gray. A bowl of this
type, with decorated stepped feet, may be shown in
plate 57. Archaeological examples, such as the one

shown have a design stamped on the
in Fig. 75, usually
bottom of the interior. The second type is Coxcatlan
Polychrome. Archaeological specimens of this type
have creamy white backgrounds decorated with or-
ange or brown designs overlaid with fine red linear
motifs. The type seems to be shown, albeit with some-
what different colors, on Borgia plate 61, upper center.
The third ware is a vaselike vessel which has a red
stepped design on a cream background (pi. 44) as does
our type Coxcatlan Bed-on-Cream.
The appearance in Codex Borgia of two vessel forms
characteristic of Venta Salada pottery may also be sig-
nificant. In Borgia plates 1-8, 26, and 47, for example,
the pedestal-base form characteristic of Venta Salada
pottery is seen. The second distinctive form is the xantil,
an anthropomorphic, polychrome vessel, usually with a
disproportionately large head on a small body. The
mouth is usually open, so that when the hollow figure
is set over an incense burner, smoke issues from it.
Paddock (1966: 229) illustrates several of these vessels

from the vicinity of Teotitlan del Camino, and Noguera
(1940: pi. 6) illustrates a xantil he found near Calipan.
Xantiles are known from Cholula, Veracruz, and other
regions, but because of their frequency of occurrence
in the Tehuacan Valley, we consider them to be a
diagnostic trait of the Venta Salada phase. At the bot-
tom of plate 38 of the Borgia a vessel which may be
a xantil is depicted in the form of the rain god Tlaloc.
This rendering has points in common with the Tlaloc
brasero from the Tehuacan Valley described and il-
lustrated by Bernal (1965b). A xantil that Seler ac-

quired in Teotitlan del Camino and the face of the god
Xochipilli in Codex Borgia displayed so many simi-
larities that Seler was led to believe that the codex was
Fig. 73. Vessel supports from Venta Salada pottery and painted in the vicinity of Teotitlan (1963: I, 101-103).
from Codex Borgia, pis. 57, 69, 1, 2. Perhaps the only completely new item that the
119
Fig. 74. Sherds of Coxcatlan Coarse with conical nodes and vessels with similar decora-
tion from Codex Borgia, pis. 18 and 63.
Borgia can add to the Venta Salada ceramic complex is Salada phase. If we were to apply to the Borgia pots
the clay pipe. On plate 58, middle row, right panel, the same methods and techniques we use to classify
the goddess Xochiquetzal holds in her right hand what the thousands of archaeological specimens, we would
appears to be a smoking elbow pipe with a vasifonn emerge with attribute clusters forming the following
bowl. Venta Salada pottery types: Teotitlan Incised, Cox-
The Borgia Codex depicts many features of ceramics catlan Red, Coxcatlan Coarse, Coxcatlan Black-on-
common to it and the ceramic complex of the Venta Orange, Coxcatlan Gray, Coxcatlan Red-on-Cream,
120
£
ir
7
^ ]
Fig. 75. Above: Bowl of Coxcatlan Gray (shown half-size) and a bowl of similar form
from Codex Borgia, pi. 57. Below: Sherd of Coxcatlan Polychrome and a vessel from
Codex Borgia, pi. 61.
121
foJ’A
44, with similar designs and eolors.

Sherds of Coxeatlan Red-on-Cream and a vessel
from Codex Borgia, pi.
Fig. 76.
122
CODEX BORGIA AND THE YF.NTA SAL ADA PHASE
Fig. 79. Buildings with flat parapet, parapet with round or

stepped merlons, and with rings shown projecting from
sides and from back. Codex Borgia, pis. 45, 13, 50, 68.
Coxcatlan Polychrome, and Tehuacan Xantiles. Thus,

if we considered each depiction of pottery in Codex
Borgia a ceramic vessel, we would have little hesitation
in considering the codex a component of the Venta
Fig. 77. Tlaloc brasero from the Tehuacan region (courtesy
Salada phase. On the basis of Coxcatlan Polychrome,
Instituto Nacional de Antropologia e Historia).
Coxcatlan Red-on-Cream, Coxcatlan Coarse, Coxcat-
lan Black-on-Orange, and Tehuacan Xantiles, one
might seriate the ceramic collection and thus place
the Borgia in the latter half of the Venta Salada pe-
riod. The dominance of Teotitlan Incised and the three
other types would tend to place it early in the last
half of the period, say a.d. 1100 to 1300.
Let us now consider representatives of buildings
and structures shown in Codex Borgia. House or
temple platforms are of two main types, either simple
truncated pyramids with from one to five steps lead-
ing to the summit (pis. 10, 12-14) or pyramids con-
structed in three to six steps, each with its panel and
battered base (pis. 35, 37, 42). Many are shown with
red and white bands or panels hinting that they were
plaster covered and then painted. The use of the
platform as a base for a building persists to this day.
Five kinds of buildings are shown on such platforms,
most of them in cross section. Floors are formed by the
top of the pyramid. Walls, and what appear to be slabs
Fig. 78. This building, possibly a chapel, at Tehuixtla, on which the roofs rest, are shown as narrow' rectangles.
Puebla, illustrates the surviving use of stone platforms. The floor and slab extend a short w'ay beyond the
123
elevation (pis. 62, 63, 65, 66) or in side elevation (pis.

25, 26, 33-35, 55, 61, 74).
The last type of house is the same shape as those
discussed above. The back wall, however, is not merely
a rectangular block but a series of two or three separate
courses, each wider than the one below it (pis. 5-8, 12,
18, 55, 61). The painting again suggests that these
structureswere plastered and then painted. Most com-
monly thev are painted in red and outlined in white,
although a few are all red, some are black with gray
borders, some display red disks against a black back-
ground, and one has yellow and blue disks on a red
background.
Fig. 80. Buildings and rooms. Codex Borgia, pis. 62, 66, 61,
55, 46, 14.
wall. Those buildings shown in section through the IrrriffTHTTTVfrwffliiniiiTrrriTTHrfTHTl^tftiTfMwl

entrance doorway and back wall have no front wall,
’
and the structure assumes the form of an inverted "L

turned in either direction. In this view the slab at the I I 1 1
1*
1 I Li lLLLLI 1 '
1
1
m |H m 1 1 1 iii itt i ill liimi]
top may, in reality, have been a lintel. The artist

may have been unable to conceive of a span without lo QQQ OnOO
the support of a
Many
lintel.
houses have a curious ringlike or looplike

Most frequently the
projection fixed to an outside wall.
loop is shown as though projecting from the back
o m
wall (pis. 1, 10, 12, 14, 37, 47, 49-51, 64, 68). Only
twice shown as projecting from either side of a
is it
temple shown in front elevation (pis. 13 and 49), and Fig. 81. Buildings with “eared” roofs. Codex Borgia, pis.
not at all on the remaining houses, whether in front 51, 49.
124
CODEX BORGIA AND THE VEXTA SALADA PHASE
Fig. 82. An “eared” or Chocho-Popoloca roof in Acatepec, Puebla.
Roofs in a variety of forms cover these buildings, porches. This type of roof is shown without an external
which may be temples or houses. All seem to be set ridgepole (pis. 3, 4, 12, 61) as often as with one (pis.
on temple platforms. Roofs are either peaked or flat. 2, 8, 46, 55). The least common type of roof shown
The latter are less common and seem to be of two in the codex (pis. 14, 40) appears to have been a conical
general types: those with flat parapets (pi. 45), and structure, also with projecting eaves.
those whose parapets are topped with “battlements.” Another, and very distinctive, type is the Chocho-
Those with “battlements” include three subtvpes, con- Popoloca These are gabled roofs, the upper por-
roof.
; sisting of those with rounded merlons or almenas (pi. tion of which overhangs the gable-filling thatch at each
45), with stepped merlons (pi. 13), or stepped pyram- end and rather resembles an ear. The gable of this
idal merlons (pis. 50, 51, 68). Some of these flat-topped projecting upper part of the roof appears to be closed
roofs seem to have been painted and probably were by vertical flaps (pis. 34, 37, 40, 49, 51). Plate 49, lower
plastered. left, lacks the vertical flap at the end of the gable but
Thatch, the most common roofing material, may be is otherwise similar.

applied over a frame made in different styles. Most A connection between Codex Borgia and the vicinity
common are simple roofs looking like the cross section of the Tehuacan Valley seems to exist in these depic-
of a truncated triangle; these have an external false tions of “eared” thatched roofs. According to Cook de
i ridgepole depicted at the summit (pis. 4.5-47, 55, 66). Leonard, this tvpe of roof is one of the outstanding
A few of the same form lack the external ridgepole (pis. characteristics of the Chocho-Popoloca-speaking peo-
18, 62, 64, 65). Almost as common are gabled roofs which ple to the west of the valley (1953: fig. 58). It is appar-
fall away to lower flattened eaves that may represent ently not found among the Mixtecs in Oaxaca, or the
mens, and between roofs shown in the Borgia and

modern roof types of the Tehuacan Valley and the
region immediately to the west. There also are simi-
laritiesbetween other forms of artifacts shown in Co-
dex Borgia and archaeological types known from the
Venta Salada phase. Unfortunately the material cul-
ture of the surrounding area is so poorly known that it
is difficult to assess the value of these artifacts as evi-
dence of a solely Venta Salada material culture.
We have discussed the evidence to connect the Bor-
gia Codex specifically with the Tehuacan Valley. We
believe the relationships summarized below indicate
that the codex was painted in the Tehuacan Valley,
1 JA * '• by people with the material culture of the Venta Salada
Fig. 83. Village of Tehuixtla, Puebla, showing “eared”
phase: (1) Two distinctive Borgia glyphs —that for
roofs.
Flint and an unusual version of the A-O year symbol
— occur on Venta Salada artifacts. (2) Pottery types
native to the Tehuacan Valley and diagnostic of the
Venta Salada phase are depicted in the codex. (3)
Chochos in the Temazulipan-Tejupan-Coixtlahuaca Houses with unusual eared gabled roofs, like those
area (Cook de Leonard 1953; Paddock, personal com- on Popoloca houses to the w est of the Tehuacan Val-
r
munication, 1965). Its depiction in the Borgia may help ley, are shown in the Borgia. (4) The cult of the planet
to eliminate Tlaxcala, the Cholula region, and most of Venus and its attendant ceremonies treated in Codex
the Mixteca Alta as well, from consideration as places Borgia wêre practiced in the Tehuacan Valley.
in which the codex may have originated. Roofs similar
in style to those of the Chocho-Popoloea are shown in Let us take the information contained in the Borgia
a Mixtec pictorial, the Codex Sanchez Solis, whose ex- Codex, as well as other evidence concerning the Tehua-
act provenience is unknown, but this does not neces- can region, and attempt supplement the cultural
to
sarily render our thesis untenable. Although the codices data derived from the archaeology of the Venta Salada
Bodley, Selden, Nuttall, and Vienna show roofs phase.
similar in form to many in the Borgia, the distinctive Documentary sources add little to the information
eared roof of the Chocho-Popoloca area does not ap- regarding the local subsistence pattern that we have
pear in any of them. derived archaeologically (see Chapter 15 below ). In ?
Codex Borgia contains more elaborate references to terms of sustenance, the documents mention com,
the —
Venus cycle a count of which Aztec documents beans, and squash, but none of the other domesticated
make only scant mention (Robertson 1963: 162) than — plants of which we found archaeological evidence, such
other ritual codices of its group. There is documentary as peanuts, guava, sapote, and avocado. Except for
evidence of the importance of the Venus cult in Teo- maguey and Opuntia, even less information exists
titlan del Camino (Roman
y Zamora 1897; I, 170). about wild food stuffs. Although there are depictions
According to Krickeberg (1956: 262-63), the Tehuacan of a number of animals in the Borgia, it contains no
Valley seems to have been the only locality in Meso- specific information about meat as food. Data concern-
america outside of the Maya region where calcvdations ing food preparation is equally slim in the documents.
of the cycle of the planet Venus were commonly made. The Borgia paintings do show the old goddess of the
This seems a further indication of the close connection sky grinding corn with a cylindrical mano on a tripod
of the Borgia with the Tehuacan Valley, for the Borgia metate (pi. 9) very similar to our archaeological exam-
is one of three manuscripts which contain a correction ples and goddesses of water and maize making nix-
of the Venus cycle (Bever 1965: 291-93). The other tamal (pi. 43, lower left and right.)
two are the probable Cuicatec codex, the Porfirio Diaz, Other pictures show pots containing fermenting
and the Mixtec genealogical codex, the Nuttall (ibid., pulque (pi. 2} or vessels in which things, mainly deities
291). or their human representations, are being boiled (pi. 7).

We have noted similarities between ceramic forms People making fire by friction, with wmoden imple-
pictured in the Borgia and actual archaeological speci- ments like those we found, appear in plates 49-51. As
126
CODEX BORGIA AND THE YEXTA SALADA PHASE
Fig. 84. Grinding com, making fire, and netting fish. Codex Borgia, pis. 9, 51, 13.
Fig. 85. Ball court with decorated stone rings, sacrificial victim and the god Tezcatlipoca. Codex Borgia, pi.
be shown, all this data and much more can

will later show ball courts without rings (Bernal 1965<z: 807).
be derived from analysis of the archaeological collec- This is further evidence of a tie between the Borgia
tions. and the Venta Salada phase.
Subsistence activities are not extensively recorded The Borgia’s rather stylized pictures of flint knives,
in the ritual manuscripts. The Borgia does include blades, flint spikes, and large atlatl points tell us no
drawings that show speared deer (pis. 22, 52), wild more about the flint-knapping industry than we have
animals probably marked for sacrifice by the pennants learned from a study of the archaeological materials.
they bear (pis. 13, 20, 21; Seler 1963: I, 219), and Nowhere in the Borgia is bow and
there a picture of the
jaguars stuck with flint knives (pis. 10, 24), but these arrow, in spite of the fact that we found many small
tell little of hunting practices. The picture of a man points which, in an archaeological collection, are nor-
catching fish with a dip-net shaped like a lacrosse mally classified as arrow points, and in spite of the
stick (pi. 13) is revealing and closely parallels a similar fact that Motolinia (1903) speaks of the people of the
scene in Codex Vaticanus 3773, plate 32. Tehuacan Valley as great archers.
In the Borgia there are drawings of ball courts in the The only form of bone implement in the codex is a
shape of a flattened “H.” Bings shown for the ball dagger (possibly made from a femur) that is some-
courts appear to be of stone (pis. 21, 35, 40, 42) and times shown being poked into the eye of a sacrificial
resemble remains of rings found on late sites in the victim (pis. 10, 16), held symbolically with an agave
Tehuacan Valley. They contrast significantly with spine (pis. 22, 48, 59), or in other circumstances (pi.
ringless courts in Oaxaca and the Mixteca. Except for 22). Archaeologists usually call this tool a long bone
Codex Nuttall (pi. 74), Mixtec genealogical codices awl.
127
Fig. 86. Industries and artifacts: conch shell horn, clay spindle whorls, wooden atlatl and dart shafts, necklace of shell
tinklers, wooden throne, and a merchant with woven headband, tump line of Z-twist cord, and twilled container for his
goods. Codex Borgia, pis. 9, 64, 16, 54, 6, 55. Notice differences in male dress.
Few representatives of ground-stone tools are to be what possibly were round shell gorgets. All of these,
found in the codex or in early documents. Only square- also, are shown in the Borgia.
based 1, .3-5, 12, 59), manos and metates
celts (pis. Graphic evidence from Codex Borgia supplements
and possiblv stone bowls (pi. 9) are shown.
(pis. 9. 43), knowledge derived from the archaeological collection
These resemble Venta Salada tools. of perishable artifacts. Here one reallv derives more
Shell-working is represented by conch-shell horns information from documentary sources than from exca-
(pis. 9, 24), triangular shell gorgets (pis. 55, 57), gear- vated materials.
like multipointed round gorgets (pis. 22, 25), and oliva- An extensive wood-working industry is shown in the
shell gorgets. Although we found none of the foregoing Borgia. Paintings of atlatls (for example, see pi. 25) and
types in the excavations, we did find pierced oliva atlatl darts or lance shafts probably made of wood
shells that could have served as tinkler bracelets (pi. occur throughout the manuscript. Equally common are
55) as well as necklaces (pis. 55, 56, 64); earrings (pi. long triangular wooden handles for knives or daggers
64); disk and spherical shell beads; and fragments of and wooden handles for celts or axes (for example, pi.
128
most lintels, and probably for all the furniture, for

example, the throne in plate 6.
One class of object called “banderolas” in the Span-

ish translation of Seler (1963) occurs throughout the
Borgia; undoubtedly it had a wooden handle. Seler
believed that the flag portion was of bark cloth. He
also believed that the paper ornament worn bv sacri-
ficial was of the same material (1963: I, 111,
victims
right). Bark cloth and bark beaters appear among
archaeological remains of the Venta Salada phase.
Closer study of the Borgia would undoubtedly disclose
other tools of wood, but even this superficial survey
gives conclusive evidence of a wood-working industry.
Among the perishable artifacts depicted in the
Borgia, textiles are perhaps the most frequently shown.
The merchants in plate 55 are carrying their wares in
twilled baskets and are using tump lines for the pur-
pose. On plate 19 there is a painting of what appears
to be a petate. Two-ply cord, probably of coarse-fibered
yarn, is shown as both Z-twisted (pi. 3, section 15, bot-
tom; pi. 6, section 39) and S-twisted (pi. 3, section 15,
top). Spindle whorls are shown in two general forms,
truncated cones (pi. 16) and disk-shaped or plano-
convex disks (pi. 12). All these items appear among the
artifacts of the Venta Salada phase.
Male clothing, generally speaking, is less ornate than
female attire. The predominant article of male clothing
is the loin cloth. This garment, probably nothing more
Fig. 87. Ornately dressed women. Codex Borgia, pis. 59,
than a length of cloth made on a back-strap loom, was
16, 9.
worn between the legs and wrapped around the waist.
The front end passes over the wrap and hangs below
19). The handles are usually painted yellow or brown, the knees, while the back end is usually tied, perhaps
apparently to indicate that they were of wood. Other with a reef knot, its free end often hanging to the
wooden implements include clubs shaped like cricket ankles. Some ornately dressed males wear decorative
bats with inserted obsidian blades; a cleaver-like “hoe” cloths about the hips and fringed capes, perhaps
with which the god Tlaloc is cultivating a cornfield (pi. poncho-like garments, that hang from the neck to the
20; cf. Volume II of this series, fig. 138); small round waist (pi. 9). A few other men wear what seems to be
“shields” (pi. 17) which are shown in the hands of in- a cape over the shoulders.
dividuals; “batons” in the hands of priests (pi. 1); long Women’s clothing shows more decoration, and the
canelike staffs that are carried by merchants (pis. 4, width of some skirts and capes suggests that these
21); and “palos de sonajas” (rattle staffs), a staff some- objects were made either by sewing together strips
times striped and with a pointed end, which possibly woven on a back-strap loom, or perhaps by using cloth
was used for music (pis. 49, 56, 73). Certainly the fire made on a wide loom. Women usually wear very dec-
drills and the hearth shown in the lower panels of orative skirts that extend to or slightly below the
plates 49, 50, 51, and 52 were of wood. On plate 18, knees. Equally ornate are shoulder capes and ponchos.
the sun god, Tonatiuh, seems to be holding a wooden Rectangular ponchos quechquemitl ) hang from the
(
spoon or ladle, and on plate 3, the goddess of fire neck to the waist with pointed corners showing. Other
appears to be carrying a gourd rattle. On plate 70 there ponchos show a rounded sinuous edge.
may also he a depiction of a digging stick, in this in- Exotically dressed personages of both sexes often
stance being used to push someone down into a tank wear woven sandals. Like those of archaeological ex-
of water. Wood must have been used in the construc- amples, the woven sole is usually short, so short that
tion of some architectural features, such as roofs and the toes are shown extending fully beyond it. The
129
may be Many figures wear

a lip plug, are depicted.
necklaces, and although these seem to be of only five
or six basic kinds, about a dozen types of gorgets or
pendants may be attached to them. Five types of brace-
lets, some with appendages, and twelve kinds of leg
bands or garters worn just below the knee appear in

the drawings.
Codices give evidence of other technological ac-
tivities, many which were in the hands of artisans
of
whose specialtywas one product. Obviously not the
least of these were the artists who painted not only
Codex Borgia, but also other codices, murals, figurines,
Fig. 88. Sandals with different types of ties and heel sup- and so on. Highly evolved styles, skills, and conventions
ports shown in Codex Borgia. testify to a schooling and discipline developed over
many generations by a closely knit guild of expert
craftsmen. Their consummate skill has made it possible
for us to recognize in their work the artifacts which we
sandals also have what seems to be a woven rectangu- have known only from excavated fragments. Further
lar heel cup that extends up to the ankle strap. Meth- study of the Borgia may increase our knowledge of the
ods of tying straps vary. Some are tied with a single religious and ceremonial life of the people who are re-
sash at the front, some with a double sash or a looped sponsible for the Venta Salada material remains.
sash with a strap around the loop. The sandals with a Although we cannot with certainty place Codex
small round tab over the instep shown throughout the Borgia in the Tehuacan Valley, we have been able to
Borgia differ from those shown in other codices of the point out certain striking resemblances between draw-
group. Excavated specimens appear to duplicate the ings in the codex, on the one hand, and archaeological
Borgia types, especially the sandals with the small tab traits and a modern form of roof, on the
distinctive
or loop over the instep (Fig. 88). other. Detailed excavation of the numerous Classic and
Even some of the headdresses, such as the one Postclassic sites within the boundaries of the Tehuacan
Quetzalcoatl wears on plate 9, include some woven Valley was beyond the purposes and scope of the Te-
parts. Generally speaking, however, the headdresses huacan Archaeological-Botanical Project, which was
are so complex and so varied that they cannot be de- designed to seek out the beginnings of agriculture and
scribed in this context. If they represent the handiwork civilization through establishment of a complete ar-
of any industry other than that of mask-makers, mil- chaeological sequence as explained in Chapter 1. Cer-
liners, or hatters, it is probably that of feather-workers tainty the close parallels between Codex Borgia and
and jewelers. the Venta Salada phase of that sequence point to the
The jewelry- or ornament-making industry is repre- need for further investigations in both the valley and
sented by more than parts of headdresses. Over fifteen the adjoining area to the west. A more complete rec-
varieties of ear ornaments are shown, ranging from ord of painted representations almost certainly exists
simple disk plugs to ones with a large variety of dang- in some of the large Venta Salada sites in the valley,
ling objects, including even a human hand (pi. 13). and we confidentlv expect that they will confirm our
Almost as many types of nose plugs, and even what hypothesis.
REFERENCES
Bernal, 1965a, 1965b. Codex Cospiano (Bologna), 1898.
Beyer, 1965. Codex Fejervary-Mayer, 1901-1902.
Caso, 1927. Codex Bodley, I960.
Cook, 1953. Codex Borgia, 1904-1909, 1963.
130
Codex Laud. Motolinia, 1903.

Codex Nuttall, 1902. Nicholson, 1955, in press.
Codex Selden II, 1964. Noguera, 1940.
Codex Vaticanus 3773 (Vatieanus B), 1902-1903. Paddock, 1966b.
Codex Vindobonensis, 1929. Robertson, 1963, 1964, 1966.
Krickeberg, 1965. Roman y Zamora, 1897.
Lehmann, 1905. Seler, 1904-1909, 1963.
CHAPTER 8
Vertebrate Fauna and Hunting Patterns
Kent V. Flannery
n 1962 and 1963 I conducted zoological investigations 1. The Valley Center and Broad Travertine Ter-
of two types in the Tehuacan Valley. One involved races on the West Side
I analysis of animal bones from archaeological sites,
Subarea 1, whose fertile, deep alluvial soil is the best
spanning the period from perhaps 9000 or 10,000 B.c.
agricultural land in the valley, has been intensively cul-
to a.d. 1500. The other consisted of the accumulation of
tivated for the last two thousand years, making it
skeletons of present-day fauna from the region as a
difficult to determine the “original” flora. Similar
comparative study group by means of which the pre-
stretches in the valley of the Rio Calapilla and some
historic bones could be identified. In conjunction with
uncultivated patches in the Tehuacan area how-
itself,
the latter study, notes weremade on the modern fauna,
ever, suggest that this was probably mesquite-grass-
and the stomach contents of animals collected were an-
land, at least in areas of higher water table such as the
alyzed by Eric O. Callen of Macdonald College of
Ajalpan plain. Besides mesquite (
Prosopis jtdifiora),
McGill University, whose notes are published at the
other wild tree legumes grow in the region, as well as
end of this chapter. Since the zoological research was
small shrubs like “coyotomate” (
Castela tortuosa)
primarily concerned with the animals eaten by the pre-
which bear fruit in the rainy season (
June-September).
historic inhabitants of the Tehuacan Valley, our infor-
Our best look at the more common edible fauna of the
mation is incomplete for most of the smaller or less
area actually comes from village remains from the first
commonly eaten vertebrates; for other animals we have
and second millennia b.c. (Ajalpan and Santa Maria
reliably dated evidence going back some ten thousand
phases), before intensive irrigated cultivation be-
years.
gan. The following lists give the most frequent animals
Within the larger environmental unit of the Tehua-
in the area.
can Valley, local differences in rainfall, drainage,
exposure, and the nature of the bedrock have combined MAMMALS
to produce four main local environments or subareas Sylvilagus audubonii
and a few specialized niches. These subareas, with their S. cunicularius
various geological features and differing plant and an- Heterogeomys sp.
imal communities, were originally outlined by James Dipodomys phillipsii
Schoenwetter and are generally recognized by the Liomys irroratas
several scholars contributing to this volume, although Reithrodontomys fulvescens
we do not refer to them in identical terms. The four Peromyscus spp.
subareas and the specialized niches within them, as re- Canis latrans
flected by the faunal remains and my own observations Urocyon cinereoargenteus
in 1962-63, are described below. Mephitis macroum
132
San Lorenzo
ZOOLOGICAL COLLECTING LOCALITIES
Tehuacan
El Riego Cave • Modern (1962-3)
-«yl Prehistoric (archeological remains)
O
—r
0"^==—
Q/
ir
O Q Coriyon •• • v 1 •
.
* .» n
.V.° ° • 0 -
_
-AREA
o • ••: ••• • •. *. Coxca 1 a n Cavev

1
o
« !
°
° °
'’•••^•• ;
^|||||lllllll
'sw^w
Pur b n r 8,
o D d
’•V'^^:lv^:V VvHlAbe as Caves
° 0
:
o o C O 0
SUBS
llllllllllll
;,&Tilapa*,J
'4j8k
AREA
in
^K§%% "Sfeotitlan i(
Valley center S broad terraces on west sidet

Wti,
-y
:
:
o 0 0
^ Limestone-travertine slopes on west edge of valley, ^%*Y.\v *!"•:
!•*/•';
t . .
•VV*V'
"'-S'-S. Broader alluvial slopes on east edge of valley
Narrow canyons 8 dissected alluvial slopes on

southeast edge of valley
Humid
Alluvial
river bottoms
canyon floors in Sub-area 2
%
Area of El Riego springs
Barren saline steppe
-Hfv
NHF
Fig. 89. Ecological subareas of the Tehuacan Valley.
133
Abundant deer and acorns (Autumn);
maguey (year-round);
line, trickling through only a small area of their wide
gravel beds. At Venta Salada, west of Coxcatlan, the
Rio Salado flows only thirty centimeters deep. This is
however, to produce a “river-arroyo” vegeta-

sufficient,
tion somewhat lusher than any other in the valley; parts
of the area between the five-meter-high bluffs of the
river bank are very densely thicketed. Many river-
side areas are known as “quebrachales” (Miranda
1948:346), dominated by dense stands of the “que-
bracho” ( Acacia unijuga ) and the “chintoborrego”
Vallesia glabra), a tall fruit tree growing only in
('
regions of high water table. Also present are thickets

Phragmites communis), as well
of a local river cane (
as reeds and “palo de agua” which grow on sandy

gravel bars in the river itself. There are numerous
sloughs along the lower Rio Salado where marsh weeds
are added to the plant community.
Fig. 90. Tehuacan Valiev,
Idealized cross section of the
This river-arroyo habitat almost rivals the Coxcatlan
showing some of the environmental zones and the seasons
in which the fauna exploited various resources. After Coe
region ( Subarea 3 ) in terms of animal resources, for
and Flannery 1964. it supports not only large populations of many of the
species found in other parts of the Tehuacan Valley,
but also a unique fauna occurring nowhere else in the
Lynx rufus area. The only turtle known today from the Tehuacan
Odocoileus virginianus Valley is a mud turtle (Kinosternon integrum inhabit-
BIRDS ing the Rio Salado and its tributaries. The green iguana
Colinus virginianus (Iguana iguana), a tropical form which is an unex-
Zenaida asiatica pected resident of the arid valley, appears to be
ColumbigaUina passerina restricted to the humid river-bank vegetation of this
Corvus corax specialized habitat. The cinnamon teal (Anas cya-

Caprimulgus ridgivayi noptera feeds in the sloughs on the lower Rio Salado
Chordciles acutipennis from November through Februarv, and killdeer ( Char -
REPTILES adrius vociferus) are also common along the river bank
Ameiva undulata in the diy season. Besides these localized forms,
Phrynosorna spp. extremely high populations of opossum and cottontail
Ctenosaura pectinate (especially Sylvilagus cunicularius) occur in the river
arrovos, and archaeological remains from the area sug-
Specialized Niche: The Arroyos of the Rios Salado, gest that collared peccary may once have been plentiful
Zapotitlan, and Calapilla. This environmental milieu there. Doves and quail are among the birds feeding
within Subarea 1 is defined as the gravel beds of the in this habitat during the rainy season, when the
rivers mentioned above, the vegetation lining their chintoborrego bears and thousands of weed seeds
fruit
banks, and the first few meters of their flood plains. Bv are ripe. Deer are still known from the floodplain of the
far the most humid niche within the valley system, this Rio Calapilla, where a pair were sighted in July 1962.
was probably a more extensive habitat in pre-Col- The following forms are common in this specialized
umbian times, before cheek-dams were built to hold niche.

back the water of the upper Rio Salado for irrigation. MAMMALS
Today the river has water only in the lower half of the Didelphis marsupialis
valley (below its confluence with the Barranca de los (and probably Marmosa canescens
Mangos and the Rio Zapotitlan), but formerly its north- Sylvilagus audubonii
ward extent must have been much greater. Previously S. cunicularius
it also received water from the El Riego mineral Liomys irroratus
springs, much of which is now commercially bottled. Rcithrodontomys fulvescens
All three of these rivers are shallow and mildly alka- Peromyscus spp.
134
VERTEBRATE FAUNA AND HUNTING PATTERNS
Urocyon cinereoargenteus
Procyon lotor
Pecari tajacu
Odocoileus virginianus
BIRDS
Anas cyanoptera (winter)
Colinus virginianus
Charadrius vociferus (winter)
Zenaida asiatica
Casmerodius albus
Formerly:
Grus canadensis
REPTILES
Iguana iguana
Kinosternon integrum
Specialized Niche: The Short-grass Steppe West of

Venta Salada. West of the confluence of the Rio Zapo-
titlan and Rio Salado, near the Venta Salada railroad
stop, lies an area roughly two kilometers by one kil-

ometer devoid of all but short-grass cover and an
occasional tree legume. High salinity and extremely
shallow soil may be contributing factors in forming Fig. 91. Subarea 2. Level top of La Mesa, looking toward
this barren plain, for a major salt-producing site of the Sierra de Zongolica.
Postclassic period is located on the edge of the steppe,
which occupies a flat limestone terrace three to five
meters above the arrovo of the Rio Salado. The salt
industry depended on seeps from saline springs in the
2. The Limestone-Travertine Slopes and Steppes
cliffs to the west, which presumably have saturated the
on the West Edge of the Valley
whole area. Here occurs one of the few populations of
white-sided jack rabbit ( Lepus callotis we observed
) In Subarea 2 the rainfall is particularly low, the
in the Tehuacan Valley, a region normally dominated ground is barren limestone or travertine with little or
by cottontails. Kangaroo rats are numerous on the open no and the plant community includes forms which
soil,
plain, nighthawks nest under the scarce leguminous can extract water from porous rock. The tree cover,
trees, and hawks and owls scour the area for small such as mesquite, vanishes and is replaced by barrel
rodents. This specialized niche within Subarea 1 is cactus, yucca, lechugilla, and “mala mujer” ( Jatropha
located close enough to the Rio Salado arroyo so that urens). Drainage is rapid (either through surface
the fauna of the latter area (especially opossum and run-off or rapid penetration of the travertine) and
cottontail) can also venture into it nocturnally when ground cover sparse, making this faunally one of the
the grass sprouts are young and the mesquite trees poorest subareas in the valley. Rock squirrels live in
bear seed pods. The most common forms are the fol- the bare cliff faces overlooking the alluvial plains, and
lowing. wood rats inhabit the steep limestone canyons leading
MAMMALS back toward Zapotitlan.
Lepus callotis White-winged doves feed on cactus fruit in the area,
D podomi/s ph illipsi
i and hawks are common; there is little to attract quail
Liomijs irroratus or ground doves, however. Large race-runner lizards
Peromyscus spp. (Ameiva undulata) speed over the bare rock during
Mephitis macroura the rainy season (May-November), but disappear in
BIRDS the winter, along with most of the insects on which
Caprimulgus ridgwayi they feed.The following incomplete lists give only the
Chordeiles acutipenn is animals commonly sighted by us in this subarea, or re-
Buteo sp. covered from local archaeological sites.
Fig. 92. El Riego Cave; sparse vegetation on the top and face of La Mesa contrasts with lush growth along its base.
MAMMALS they have the ability to extract water from porous rock,
Lepus callotis as does much of the flora of Subarea 2. Radical differ-
Sylvilagus audubonii ences exist between the canyon walls or rim and the
Citellus (Spermophilus) mexicanus floor, where water flows occasionally in the rainy season
Feromyscus spp. and may have flowed steadily in the Pleistocene. This
Neotoma spp. specialized niche is described by Paul C. Mangelsdorf
Spilogalc augustifrons as an extremely probable habitat for wild com; in fact,
BIBDS cobs of morphologically wild com were excavated from
Zenaida asiotica San Marcos rock shelter in Tecorral Canyon (see
Buteo sp. below, Chapter 9). Included in the faunal remains from
REPTILES the same rock shelter and one adjoining it were the
Ameiva undulata forms listed below, which are probably more common
in Tecorral Canyon than in the rest of Subarea 2.
Specialized Niche: The Floors of Tecorral Canyon MAMMALS

and Related Deep Barrancas. This isolated niche, re- Sylvilagus audubonii
stricted to the floor of the wide canyons in the travertine S. cunicularius
cliffs west of the valley, is an area of well-developed Cratogeomys sp.
alluvial soils with mesquite and other “wet area” plants Dipodomys phillipsii
combined with patches of grass. This plant community Liomys irroratus
is cut off from the rest of the alluvial areas in Tehuacan Peromyscus spp.
by the barren travertine walls which surround it, Sigmodon hispidus
through which few plants can successfully pass unless Urocyon cinereoargenteus
136
Fig. 93. Subarea 3. Thorn-scrub and cactus forest during the wet season. Fig. 135 shows dry-season aspect of same area.
Photo C. E. Smith, Jr.
Mephitis macroura agriculture on the plain of Tehuacan. There is evidence

Pecari tajacu that in the past the flow may have been even greater
BIRDS (see above, Chapter 5).
Colinus virginianus The soil on the Tehuacan plain is alluvial sediment,
Zenaida asiatica and it combines with the high water table along the
El Riego cliffs to produce a belt of atypically lush vege-
tation. Mesquite, guaje, ciruela, and other trees al-
Specialized Niche: The El Riego Springs and Adjoin- ternate with occasional dense weed patches along the
ing Areas of the Valley Floor. The Cerro de la Mesa streams and irrigation canals, and the spring site of San
is porous travertine bearing for the most part the typical Lorenzo supports the largest single tree in the valley,
yucca and lechugilla vegetation of Subarea 2. The cliffs a “tide” or cypress which is clearly many centuries old.
where it descends to the plain of Techuacan are barren Large prickly pear and organ cactus flank the cliffs, and
white rock to which agaves, cacti, and an occasional scattered patches of grass afford cover for cottontails,
scrawny tree cling, and soil is virtually nonexistent. At small rodents, and bobwhites. Raccoon and opossum
the base of the cliffs, however, tremendous quantities tracks line the banks of the irrigation canals, and rock
of subsurfacewater burst out through a series of large squirrels nest in the cliffs. meter or so above the
Only a
mineral springs which formerly drained into the Rio valley floor, where the alluvial soil vanishes, the unusual
Salado. Today these springs support five bottling works El Riego vegetation gives way to more typical Subarea
— Tehuacan, Penafiel, San Lorenzo, Garci-Crespo, and 2 succulents and agaves, making this a region of inter-
El Riego —with enough water over for irrigation
left esting diversity.
137
Most common today, or in the bone remains from ley offered prehistoric man more hunting and plant-
El Riego Cave, are the following animals. collecting resources than this subarea, with its abun-
MAMMALS dant cover for all kinds of animals and its high
Didelphis marsupialis proportion of vegetal foods.
Marmosa canescens Tall tree legumes and the poclxote tree (
Ceiba
Lepus callotis parvifolia make up part of the canopy. Present also
)
Sylvilagus audubonii are wild fruit trees like chupandilla, cosahuico, and
S. cunicularius ciruela; pod-bearing trees whose seeds are eaten (guaje
Heterogeomys sp. bianco, guaje Colorado, garabatillo); countless thorny
Dipodomys phillipsii shrubs; and several kinds of large cactus whose fruits
Liorm/s irroratus are edible, including the prickly pear Opuntia sp.),
(
Reithrodontomys fulvescens pitahaya ( Lemaireocereus sp.), organo, and tetecho.

Peromyscus spp. The food supply changes seasonally, with cactus
Baiomys musculus fruits appearing in April and lasting until July, tree
Sigmodon hispidus legumes bearing pods in the rainy season (June to Au-
Urocyon cinereoargenteus gust), and wild fruits like chupandilla coming ripe in
Procyon lotor September. Even in the middle of the dry season (De-
Mephitis macroura cember-February) there are bushes like the tem-
Pecari tajacn pixquistle (Buinelia sp). whose fruit can be eaten.
Odoco ileus vi rgi n ian us Animal populations in the area are correspondingly
From caves: high; most of the coyotes and white-tailed deer remain-
Eptesicus fuscus ing in the Tehuacan Valley are to be found in this
Neotoma alleni subarea. Along wet barrancas, opossum, raccoon, and
N. mexicana ring-tailed cat (Bassariscus astutus) occur, and striped,
BIRDS spotted, and hog-nosed skunks frequent the forested
Day: talus slopes. Owing to the abundant ground cover, cot-
C ol in us virgin ianus tontails of two species ( Sylvilagus cunicularius and S.
Zenaida asiatica audubonii) are common. On a single evening (July
ColumbigaUina passer ina 28, 1962) between 6:20 and 6:45, we saw five cotton-
Night: tails active along a narrow one-kilometer stretch of trail.
C aprim ulgus ridgicayi The impossibility of telling young S. cunicularius from
C hordeiles acutipen n is adult S. audubonii at a distance just at dusk prevents
Ti/to alba our giving relative proportions of these species, but ar-
In addition, a night trip to the El Riego cliffs afforded chaeological remains from levels laid down during the
us our only glimpse, on July 27, 1962, of the unusual modern climatic regime suggest that 70 percent of the
elf owl ( Micrathene ivhitneyi) roosting high on an cottontails in the Coxcatlan area are Audubon cotton-
organ-pipe cactus. tails.
The most common mammals we observed at night

3. The Broader Alluvial Slopes in this subarea were, in order of frequency, kangaroo
on the East Edge of the Valley rats, deer mice, cottontails, striped skunks, and opos-
Subarea 3 consists of gravelly downwash and talus sums. Birds normally seen in the evenings were night-
slopes where precipitation has formed a vast dendritic jars, white-winged doves, and an occasional owl. By
network of arroyos and some subsurface flows. A few day, much of the game is hidden, but black iguanas
arroyos, like the Barranca de los Mangos, near Calipan, ( Ctenosaura pectinata and white-winged doves ( Zen-
have water even during the dry season. Here the vege- aida asiatica ) are still seen regularly. The archaeolog-
tation of the valley, in Smith’s words, “most nearly ical evidence showed that cotton rats ( Sigmodon
approaches a forest" (Smith, 1965/;: 117), with a broad hispidus), spotted skunks (Spilogale augustifrons ), and
canopy closing one and one-half to two meters above gray foxes (Urocyon cinereoargenteus) had been nu-
the ground. The flora is thickest along arroyo edges and merous in the area for thousands of years, but we never
thins out in other areas, with patches of grass occurring personally observed them. The list of animals fairly
in the gaps. Probably no part of the Tehuacan Val- common in this subarea, past and present, is extensive.
138
MAMMALS stinging nettle (mala mujer). Tall tetecho, prickly pear,

Present throughout area, or of unknown habitat: and organ cacti occur, and occasional tree legumes are
Alarmosa canescens seen on alluvial fans. The arroyo walls are barren rock,
Artibeus sp. but thefloors support a low thorn vegetation. Resources
Eptesicus fuscus do not approach those of Subarea 3, and barrancas are
Sylvilagus audubonii normally dry.
S. cunicularius White-winged doves, ameiva lizards, and black
Cratogeomys sp. iguanas are the most frequently seen animals during
Dipodomys phillipsii the day; at night, striped skunks, Audubon cottontails,
Liomys irroratus wood and kangaroo rats are fairly common. The
rats,
Peromyscus spp. (esp. melanophrys) lack of ground cover would seem to inhibit the Mexican
Sigmodon hispidus cottontail (Sylvilagus cunicularius), for few are found.
Canis latrans The following animals were present.
Urocyon cinereoargenteus MAMMALS
Spilogale augustifrons Sylvilagus audubonii
Mephitis macroura Cratogeomys sp.
Conepatus mesoleucus D ipodomys phill ipsii
Felis concolor Liomys irroratus
Lynx rufus Peromyscus spp. (esp. melanophrys)
Pecari tajacu Sigmodon hispidus
Odocoileus Virginia mis Urocyon cinereoargenteus
Present along watercourses: Spilogale augustifrons
Didelph is marsupialis Mephitis macroura
Bassariscus astutus Probably Lynx rufus
Procyon lotor From caves:
From caves: Eptesicus fuscus
Neotoma alleni Neotoma alleni
COMMON BIRDS In low frequency:
Day: Sylvilagus cunicularius
Colinus virginianus Odocoileus virginianus
Zenaida asiatica BIRDS
Columbigallina passerina Zenaida asiatica
Night: Caprimulgus ridgwayi
Caprimulgus ridgwayi Chordeiles acutipennis
Tyto alba Tyto alba
REPTILES REPTILES
Ctenosaura pectinata Ctenosaura pcctinata
Ameiva undulata Ameiva undulata
4. Narrow Canyons and Dissected Alluvial Slopes Faunal remains from archaeological sites in the
on the Southeast Edge of the Valley Tehuacan Valleyraise two points of interest regarding
Subarea 4 might just as accurately have been con- the prehistoric climate of the area. First, a somewhat
sidered a niche within Subarea 3, for it lies within and colder late Pleistocene or immediately post-Pleistocene
These canyons have a lower
alternates with the latter. climate, ending before 7000 b.c., is suggested by re-
canopy (closing about one meter) and no grassy

at mains of certain types of extinct fauna in the lowest
patches such as occur in the Coxcatlan area. The best- levels of Coxcatlan Cave. Second, since the establish-
known collection locus in Subarea 4 is the Arroyo Len- ment of the present climatic regime and the disappear-
cho Diego, where long faunal sequences were present ance of the now-extinct forms at the date mentioned
in Purron and Abejas caves. The arroyo has sparse above, there seem to have been no fluctuations which
ground cover, mainly scatterings of lechugilla and altered the fauna significantly.
139
The Early Ajuereado Climatic Phase Gopher tortoise (Goplierus cf. herlandieri), 102
The lowest four zones (XXV-XXVIII) of Coxcatlan fragments

Quail, other than Bobwhite, 2 fragments
Cave in Subarea 3 yielded a fauna which differs from
all subsequent levels in all sites excavated by the 2. The following forms, common in the Coxcatlan
—
Tehuacan Project including later Ajuereado phase area today, are totally absent in the Early Ajuereado
levels, Zones XXIII and XXIV of Coxcatlan Cave.

remains:
I would not be at all surprised to learn that a hiatus

White-tailed deer ( Odocoileus virginianus
of some magnitude separated Zones XXIV and XXV Collared peccary ( Pecari tajacu )
at the latter site. Black iguana ( Ctenosaura pectinata )
Since some 1200 identifiable bone fragments occur in
Cotton rat (
Sigmodon hispidus )
the four zones of what I am calling the “Early” Kangaroo Dipodomys phillipsii )
rat (
Ajuereado phase, the distinctness of the fauna cannot 3. The following forms are present today and were
simply be due to sampling error. Both human hunting also represented in the Early Ajuereado debris:
practices and true environmental differences are re- Audubon cottontail (Sylvilagus audubonii), less
flected in the bone collections. The Earlv Ajuereado than 100 fragments
fauna includes: (1) animals which are now extinct
Coyote ( Canis latrans), 2 fragments
everywhere in the Western Hemisphere, and (2) ani- Spotted skunk Spilogale sp.), 4 fragments
(
mals whose range today lies considerably to the north Striped skunk (
Mephitis sp.), 4 fragments
of Tehuacan, in cooler (and often drier) regions. In Hog-nosed skunk ( Conepatus sp.), 1 fragment
spite of the large sample of bones, the Early Ajuereado Ring-tailed cat (
Bassariscus sp.), 4 fragments
collections lack some of the species which are most Rock squirrel (
Citellus mexicanus), 4 fragments
common in the area today, including all of the semi- Wood Neotoma spp.), 27 fragments
rat (
tropical forms. And perhaps most significantly, the Deer mouse Peromyseus spp.), 21 fragments
(
small rodent fauna of these lowest levels has a very Spiny mouse ( Liornys sp.), 1 fragment
different aspect from that of all subsequent levels in
all 1.sites in the sequence.
The Early Ajuereado Rodents
The precise age of these “Early Ajuereado” deposits
is not known, but on the basis of radiocarbon deter- The small wild mice and rats of the Early Ajuereado
minations from later Ajuereado levels, they must ante- levels, like virtually all the rodents from caves in the
date 7000 b.c. As far as the fauna itself goes, the de- Tehuacan Valley, appear to have been brought into
posits could be as earlv as 10,000 b.c. The period must Coxcatlan Cave in the form of owl pellets. For those
be Late Peistocene or earliest post-Pleistocene, con- who are not familiar with owl physiology, let us briefly
temporary with the mammoth-kill sites in the valleys of explain this process.
Mexico and Puebla. Whatever their actual date, the Numerous of the local owls, especially the barn owl
Early Ajuereado faunal remains belong to a time when Ti/to alba, use the caves of the Tehuacan Valley as
the Coxcatlan area was not the hot, semitropical, nesting places. From these vantage points, the owls
thorn-scrub forest it is today. flyout at night and search for small rodents, finding
and swallowing whole as many as they can. There is
no evidence that owls discriminate against certain
The Emit/ Ajuereado Fauna species of rodents, and hence their diet constitutes a
Included in the bone remains from Zones XXV- random sample of the local rodent population. Owls
XXVIII are the following forms, no longer present in swallow mice whole; the fleshy parts are dissolved by
the Tehuacan region: stomach acids, and then the fur forms a cocoon around
Horse ( Equus sp. indet.), 7 fragments the clean skeleton. This cocoon, known as an “owl pel-
Antelope (ef. Antilocapra americana), 45 frag- let,” is then regurgitated while the owl is roosting.
ments Some caves in southern Mexico are covered by a caqêt
Large jack rabbits ( Lepus sp. indet.), more than of disintegrating pellets.
700 fragments Owls ccupied the Tehuacan caves during the pe-
Large fox, more rugose than Urocyon, 49 frag- riods between human occupations and left behind a
ments stratified series of disintegrating pellets which were
Small ground squirrel, chipmunk or prairie-dog eventually incorporated into the human refuse. These
size, 4 fragments constitute a series of random samples of the available
140
1-
rodent population during all periods of prehistory, an
< UJ
1- a: V) UJ almost too-perfect situation for statistical analysis.
< o 3 CO
3
a: o t- Because small rodents are more sensitive climatic
o < o
z a: oc indicators than most mammals, especially the larger
o < o we found
h- o >-
z o a: ones, it useful to characterize the prehistoric
t- z o UJ environment by calculating the percentage each genus
o < a UJ
o CO $ Q of rodent contributed to the whole sample from the
owl pellets. Each subarea of the Tehuacan Valley has
its own peculiar set of percentages, but the main five
genera of small rodents are the same everywhere: only

the proportions are different. From the Late Ajuereado
phase until the present day, as revealed by owl pellet
remains, the proportions of one type of rodent to an-
—
other have remained fairly constant a good indication
that little environmental change has taken place.
Percentages under present environmental S igmodon hispidus, the cotton rat, has increased, prob-
conditions (7000 BC - 1500 AD), ably in response to increased agriculture and second-
valley as a whole growth vegetation, but that is a not-unexpected situa-
tion; its range is still expanding northward across the
LTnited States into areas previously unoccupied (Hall
and Kelson 1959: 671).
The small rodent population of the Early Ajuereado
phase, however, is quite different from today’s and
quite different from all other prehistoric populations
observed in owl pellet samples. Fig. 94 shows this
in graphic form, and Table 15 and the briefer tabula-
Late
Percentages under present environmental
Present Climatic Regime Pleistocene
conditions (7000 BC - 1500 AD), O
Coxcatlan area only All sites Tc 50 only Tc50 only
Cotton rat (
Sigmodon ) 55 (225)* 29 (93)
Kangaroo rat ( Dipodomys ) 27 (64) 17 (31)
Spiny mouse ( Liomys 35 (138) 12 (22) 1 (1)
Wood rat ( Neotoma ) 33 (74) 9 (17) 7 (27)
Deer mouse ( Peromyscus) 17 (37) 6 (7) 11 (21)
° Figures without parentheses are minimum numbers of rodents; figures in
parentheses are numbers of identified bones.
tion drawn from it give the figures on which the graph
is based. We
have compared the Early Ajuereado ro-
dents, on the one hand, with the owl pellets from Cox-
catlan Cave (Tc 50) only, and on the other hand, with
the entire sample from all Tehuacan sites.
Some 225 identifiable bones of cotton rat appeared
in post-Pleistocene levels at Tehuacan. This is not
Percentages during Late Pleistocene,
surprising, for Sigmodon hispidus, wherever it occurs,
Coxcatlan area
tends to be the most numerous wild rodent in the
region (Hall and Kelson, loc. cit.). It is today perhaps
the most common wild rodent of Mexico and the south-
Fig. 94. Comparison of populations of small rodents. Per- eastern United States. Yet, out of a sample of 50 iden-
centages are based on minimum numbers of individuals of tifiable rodent bones from the Early Ajuereado phase,
each genus. Sigmodon is totally lacking.
141
Table 15. Small Rodent Remains from the Tehuacan Sites by Zone and Phase
Deer Mouse Wood Rat Spiny Mouse
N.
N. alleni mexicana Unspecified
corpse corpse
mandible mandible
palate palate palate palate
fragments fragments fragments
or
of or or
of
mandible
or mandible
individuals mummified innominate individuals individuals mummified innominate

innominate
skull skull skull

total ramus femur
-O other total
skull femur
-O other total ramus femur
tibia other
r. r.
Tc 35e, A 2 2 2 1 3 1 1 1 2 5 1 2 1 1
Tc 50, 1
Tc 50, II 1 1 1
Tc 35w, 3 1 2 1 1
Tc 35e, B 1 1 1
Tc 50, III 1 1 1
Tc 255, A 1 1 1 2 4 1 1 2 1 2 1 1
Tc 35e, C 1 4 1 1 1 1 2 9 2 2 1 3 1 3 6 3 2 1
Venta Salada totals 4 7 3 1 2 1 7 19 1 4 2 2 2 7 1 8 15 4 5 3 2 1
Tc 35e, C-D (owl pellets) 3 18 3 4 4 4 3 3 10 4 4 2 10 88 10 15 19 19 20 5

Tc 35e, D 1 3 1 1 1
Tc 255, B 1 5 1 1 1 1 1
Tc 35e, E 1 1 1
Tc 50, IV 1 1 1
Tc 254, B 1 1 1 1 4 2 1 1 1 4 1 1 2
Tc 50, V 1 3 1 1 1
Tc 272, B 1 2 1 1 1 2 1 1
Tc 272, E 1 1 1
Tc 50, VI 2 4 2 2
Tc 272, F 1 1 1
Palo Blanco totals 2 3 1 1 1 4 12 2 3 1 1 3 2 8 17 1 2 2 3 3 6

Tc 50, VII 1 2 1 1 2 6 1 1 4 1 3 1 1 1
Santa Marla totals 1 2 1 1 2 6 1 1 4 1 3 1 1 1

1
Tc 254, C 1 1 1
Ts 204, H 1 3 2 1
Ajalpan totals 1 1 1 1 3 2 1
Tc 272, L 1 2 2
Tc 50, VIII 2 2 2 1 2 1 1
Tc 50, IX 1 1 1
Tc 272, N
Tc 254, D 1 1 1 2 3 1 2 1 1 1
Tc 50, X 1 1 1 2 4 2 2 2 4 2 2
Tc 272, P 1 4 1 2 1
Abejas totals 2 2 1 1 8 14 4 5 2 3 5 9 1 5 3
Tc 50, XI 1 1 1 1 3 2 1
Tc 254, E 1 1 1 4 5 4 1
Tc 254, F 1 2 1 1
Tc 50, XII
Tc 272, Q
Coxcatlan totals 2 2 1 1 6 10 4 1 3 1 1
Tc 50, XIV
Tc 50, XV 1 2 2
Tc 307, F
Tc 35w, 5
Tc 272, S 1 r 1
Tc 50, XVIII 1 1 1
Tc 50, XIX 1 1 1 1 1 1 1 1 1
Tc 50, XXII 1 1 1
El Riego totals 3 3 2 1 2 2 1 1 2 3 2 1
Tc 35w,
Tc 50, XXIV
Late Ajuereado totals
Totals, Present Climatic Conditions 17 37 1 6 8 4 12 5 1 33 74 3 18 2 2 9 25 9 6 35 138 1 16 23 23 35 34 6
Tc 255, C (Mixed El Riego and Early
Ajuereado) 2 3 2 1 1 1 1
Tc 50, XXV 1 1 1 1 4 1 2 1
Tc 50, XXVI 4 8 4 4 3 9 2? 1? 5 1 1 1 1
Tc 50, XXVIII 6 12 7 2 2 1 3 14 1 2 2? 3 3 3
Early Ajuereado totals 11 21 11 7 2 1 7 27 2 2 4? 3? 8 4 4 1 1 1
Totals, all periods 28 58 1 6 19 4 19 7 2 42 104 5 20 6? 7? 9 34 13 10 37 140 1 16 23 23 35 36 6
142
(Minimum Numbers of Individuals Derived from Identified Fragments)
Kangaroo Rat Cotton Rat Harvest Mouse Pygmy Mouse
1 I
1- 1 1 1a J 1a !
Ia ~
5 S S
, , I o o ° 5 °
j
llilljll III
individuals
I 1 1 i 1 1 1 I 1
1 2 Tc 35e, A
Tc 50,
3 7 2 5 Tc 50, II
Tc 35w, 3
Tc 35e, B
2 4 2 2 2 8 2 2 2 1 Tc 50, III
1 1 3 10 4 4 1 Tc 255, A
1 2 2 10 13 11 Tc 35e, C
8 17 3 8 12 52 5 11 17 11 3 Venta Salada totals
3 21 3 5 5 63 7 16 9 12 8 1 3 1 1 2 1 Tc 35e, C-D (owl pellets)
2 Tc 35e, D
1 Tc 255, B
Tc 35e, E
15 Tc 50, IV
9 Tc 254, B
10 Tc 50, V
3 Tc 272, B
Tc 272, E
5 25 5 7 Tc 50, VI
Tc 272, F
20 65 6 13 7 22 13 4 Palo Blanco totals
1 Tc 50, VII
Santa Maria totals
Tc 254, C‘
Ts 204, H
Ajalpan totals
Tc 272, L
1 1 Tc 50, VIII
Tc 50, IX
1 5 Tc 272, N
1 2 1 1 Tc 254, D
1 1 2 5 Tc 50, X
1 1 Tc 272, P
1 1 7 18 1 2 2 3 1 Abejas totals
Tc 50, XI
12 11 Tc 254, E
Tc 254, F
1 1 1 Tc 50, XII
Tc 272, Q
2 3 1 1 1 1 2 Coxcatlan totals
l 1 Tc 50, XIV
1 Tc 50, XV
1 Tc 307, F
1 Tc 35w, 5
Tc 272, S
1 1 1 Tc 50, XVIII
Tc 50, XIX
2 Tc 50, XXII
5 4 El Riego totals
Tc 35w, 6
1 1 Tc 50, XXIV
1 1 Late Ajuereado totals
27 64 3 10 28 18 1 55 225 20 40 33 70 45 17 1 3 1 1 2 1 Totals, Present Climatic Conditions
Tc 255, C (Mixed El Riego and Early

1 1 Ajuereado)
Tc 50, XXV
Tc 50, XXVI
Tc 50, XXVIII
Early Ajuereado totals
28 65 3 10 28 19 1 55 225 20 40 33 70 45 17 1 3 2 Totals,
1 1 all periods
143
The second most common small rodent in the Cox- rabbits and wood rats like green leaves of mesquite,
catlan area, and one of the most common in the valley and wood rats are particularly abundant in groves of
as a whole, is the kangaroo rat, Dipodonujs. It is also prickly pear (Schmidt-Nielsen 1964: 14.3—45). The ab-
totally lacking in the Early Ajuereado sample. sence of deer and possibly Mexican cottontail suggests
The deer mouse, a prairie form which was never that there was no real “forest” cover in the vicinity of
particularly abundant in post-Pleistocene owl pellets Coxcatlan, and the absence of iguana suggests that
from the Tehuacan caves, constitutes 60 percent of the winters were more severe than they are today.
small rodents from the Early Ajuereado phase. Byers (Chapter 4 above) has suggested that during
To check for the significance of these differences in the last glacial period there may have been an equator-
percentages of small rodents, we ran a Difference of ward displacement and subtropical
of the tropical
Proportions Test on Sigmodon, Peromyscus, and Neo- zones, giving southern Puebla a cooler and drier cli-
toma. The test indicated that the chances that our mate like that of San Juan sin Agua, in San Luis Potosi.
Early Ajuereado rodent sample and our post-Pleis- Such a possibility is indeed supported by the Early
tocene rodent samples from the Coxcatlan area came Ajuereado fauna. Faunal remains, however, are a poor
from the same population was less than 0.01 percent. second choice to pollen analysis, and it is to be greatly
In other words, the rodent population, and presumably regretted that pollen was not preserved, thus render-
the environment as well, were significantly different ing impossible a palynological study of these early
during the Early Ajuereado phase. levels on which to base a somewhat more firm con-
clusion.
A Reconstruction of the
Emit/ Ajuereado Environment The Modern Climatic Regime
Archaeologists have, on the average, tended to pic- Bv at least 6500 b.c., according to radiocarbon dates,
ture the Late Pleistocene as a cooler and moister pe- a completely modern fauna was present in the Tehua-
riod. Thisis not what is reflected in the Earlv Ajuere- can Valley. These dates are in accord with those from
ado fauna. If anything, it suggests that the Tehuacan Frightful Cave, Coahuila (Taylor 1956), which show
environment may have been cooler and still more arid fully modern fauna by 6900 b.c. (M-191). The last of
than today. the extinct animals, of course, may not have been gone
One way to reconstruct the environment is to look for long; horse at Gypsum Cave, Nevada, was dated at
an area which today contains a similar set of mammal 6567 b.c., andWhitewater Draw, Arizona, occurred
at
species: large jack rabbits, antelope, small ground in levels dated 6240 and 5472 b.c. (see Hester 1960:
at
squirrels, gopher tortoise, lots of wood rats, and prairie Table 5 for summary).
deer mice. Today, such a complex can be found in the None of the lower levels at Coxcatlan Cave show a
very arid interior plains of northern Mexico. Parts of period of overlap during which outgoing antelope and
western San Luis Potosi, eastern Zacatecas, and west- incoming white-tailed deer shared the Tehuacan Val-
ern Coahuila supported large antelope herds until the ley. Either there is a real hiatus between Zones XXIV
1800’s (Leopold 1959: 518), and are still rich in the —

and XXV which seems not unlikely or else we are —
smaller fauna characteristic, of the Early Ajuereado dealing with Hester’s "Tvpe 1” extinction, a “swift and
phase. total decline” for horse and antelope (1960: 58). Smaller
This area today open temperate steppe, with mes-

is forms such as fox and turtle were probably “replaced”
quite in the alluvial bottomlands and prickly pear by their modern counterparts, but the larger species
on rocky slopes. In contrast to the Coxcatlan area, it vanished so abruptly that we are inclined to suspect
receives winter frosts and lacks the dense vegetation new environment was
that the unsuitability of their
of the arid tropical thorn forest. Its relative lack of augmented bv pressure from human hunting. As Leo-
underbrush and its open expanses of plains make it bet- pold points out (1959: 523), antelope are “not particu-
ter habitat for antelope and jack rabbit than for white- larlv adept at keeping away from people, and hunting
tailed deer and cottontail, although the latter occur without regulation or control can easily exceed the
in the area. productivity of the herds.” The last of the dwindling
This is how
would tend to reconstruct the Early
I antelope bands may simply have fallen victim to man,
Ajuereado environment. The valley floor would have and the absence of deer bones in Earlv Ajuereado levels
been open steppe, grazed by horse and antelope, but may only mean that prehistoric man found that hunting
to judge by the lack of cotton rats, it could not have antelope on the open plain was easier and more pro-
had a very rich cover of tall weeds or grass. Both jack ductive than hunting deer in the bushy arroyos. At
144
the Levi Rock Shelter in Texas, white-tailed deer were quent in the dry canyon areas or barren limestone
alreadv present at 8000 b.c. in levels containing extinct plains.
horse (Alexander 1963). Habits and diet: Callen’s analysis of stomach con-
Once established, the present-day climatic regime tents of our specimen (see end of chapter) suggests a
held fairly constant. One good evidence of this is that rainy-season diet including mesquite beans and insects
after about 7000 b.c., as we have pointed out, the small of various kinds. These shy animals spend the day hid-
rodents seem to have remained the same until the time den in the underbrush and at night fall easy victim to
of the Conquest. The five main forms are present from traps or jack-light hunters.
El Riego times on, which we take to mean that the basic Prehistoric distribution: Didelphis is represented as
ecology of the Tehuacan Valley has changed very little far back as about 6000 b. c. by a maxilla fragment from
in the last 8500 years or so. Vegetal remains from the Zone XVII of Coxcatlan Cave (Subarea 3). It is most
caves reportedly suggest the same stability of climate abundant in the remains from El Riego Cave, especially
(see below, Chapter 12). during the Palo Rlanco and Venta Salada phases from
200 b.c. to the Spanish Conquest.
Systematics of the Tehuacan Fauna
What follows is not an exhaustive list of the fauna of
Marmosa canescens
(murine opossum; raton tlacuache)
the Tehuacan Valley. The was to
object of this study
compile data only on the edible mammals, reptiles, and Modern specimens collected: None.
birds of the region, and to concentrate especially on Present-day distribution: Unknown.
the habits and ecology of those species most relied up- Prehistoric distribution: The two archaeological oc-
on by prehistoric man. Hence the “systematics” pre- currences of this tiny opossum suggested that they were
sented below are heavily skewed in favor of forms brought into the cave by owls. A mandible and three
which served as food for the Indians of the Tehuacan limb bones found in a group of owl pellets in El Riego
Valley. Cave dated to just before a.d. 700. Another mandible
There are many more species of bats, insectivores, was found with a group of bat and small rodent re-
and possibly murine opossums than we collected, and mains in Zone II, Coxcatlan Cave (Venta Salada
we dealt only with the most common cricetid rodents. phase).
Among birds and reptiles, only a handful of species fell
within the scope of our project. The list below presents Order Chiroptera
the animals recovered in the order in which they are
Family Phyllostomidae
usually listed by zoologists; that is, phylogenetically.
Artibeus sp. (probably jamaicencis)
Mammals (fruit-eating bat; local name unknown)
Order Marsupialia Modern specimens collected: None.
Present-day distribution: Unknown, but probably
Family Didelphiidae ranges over all of the Tehuacan Valley.
Di del phis ma rsup ia l is Prehistoric distribution: One complete skull of
(opossum; tlacuache) Artibeus was found in Zone VI, Coxcatlan Cave (Palo
Rlanco phase).
Modern specimens collected: One adult male.
I Measurements: Total length, 830 mm. Tail, 355 mm. Family Vespertilionidae
Hind foot, 66 mm. Ear, 58 mm. Weight, 2.39 kg.
Eptesicus fuscus
Present-day distribution: Opossums are distributed
throughout the valley wherever there is even a trickle
(brown-bat; local name unknown)
of water. They are particularly common in the Rio Modern specimens collected: Four specimens col-
Salado-Rio Zapotitlan arroyos and the spring-fed irri- lected in August 1962 from caves in the El Riego cliffs
gation canal system along the El Riego cliffs. Our were identified as Eptesicus fuscus by Dr. Rernardo
specimen, shot near Venta Salada at midnight, June Villa R. of the Instituto de Riologia, Universidad Na-
21, 1962, was only one of dozens observed during night cional Autonoma de Mexico.
trips to the riverside in that area. Tracks are common Present-day distribution: Ranges over the whole of
at El Riego, where the animal was also well represented the valley.
prehistoric-ally. Opossums are, however, not at all fre- Prehistoric distribution: A complete skull of Eptes-
145
icus was found in Zone II, Coxcatlan Cave (Venta together, because the character of the ecological sub-
Salada phase). areas of the valley is somewhat reflected in the propor-
tions of these two species one to another. The archae-

ological remains have afforded us a sample of cotton-
Order Lagomorpha
tails spanning many centuries in all subareas.
Family Leporidae The Audubon cottontail is an essentially arid tem-
perate species whose range to the north reaches Mon-
Lepus callotis
tana (Hall and Kelson 1959: 267). The larger Mexican
(white-sided jack rabbit; liebre)
cottontail is more of an arid-tropical form, with a range
Modern specimens collected: None. extending south to the Isthmus of Tehuantepec and
Present-day distribution: Jack rabbits are restricted reaching only as far north as the state of Hidalgo, ex-
to open areas in the Tehuacan Valley, such as the bar- cept for an arm extending up the Pacific coast lowlands
ren saline steppe west of Venta Salada or the extensive
to Sinaloa (Leopold 1959: Fig. 133). The ranges of these
cleared areas on top of the Mesa de El Riego. They two species overlap in the Tehuacan Valley, where
are also found in the barren parts of the Zapotitlan hills
both are abundant and have been so for the last 9000
and apparently occurred in moderate numbers on the years. It would be interesting to know what factor
valley floor itself before intensive agriculture and set- enables both cottontails to live in the same area with-
tlement confined them to the less inhabited areas. out competing, but so far we have no evidence except
Habits and diet: We observed jack rabbits foraging hints (see below) that there may be a dietary difference.
on the Venta Salada steppe at midnight, July 28, 1962.
About five individuals were seen in an area roughly a Sijlvilagus audubonii
kilometer square. The vegetation in the area is very (Audubon cottontail; conejo chiquito)
short grass (kept so by sheep and goats) resting on a Modern specimens collected: Two males and one
limestone-travertine terrace with almost no interven- female.
ing soil. There are occasional tree legumes and weed Measurements (range of three specimens): Total
patches in rain-filled depressions. length, 345-365 mm. Tail, 30-41 mm. Hind foot, 80-90
Prehistoric distribution: Lepus callotis is not known mm. Ear, 63-65 mm. Weight, 0. 8-0.9 kg.
with certainty before the Coxcatlan phase. Its remains Present-day distribution: Audubon cottontails are
occur Zone XI of Coxcatlan Cave and Zone 4 of El
in
found throughout the valley, their populations being
Riego Cave, West Niche, dating to roughly 4000 b.c. highest in relation to Mexican cottontails in areas of
Jack rabbit remains are virtually lacking in thorn-forest most barren ground and poorest soils, as in Subareas
zones such as Subarea 3. 2 and 4. The specimens we collected came from the
Venta Salada plain, the Ajalpan plain, and fields near
Other Jack Rabbits
Calipan. A pregnant female, with two embryos, was
Numbers of very large rabbits, as yet unidentified, taken June 27, 1962, at Venta Salada.
occurred in the Tehuacan Valley in association with Habits and diet: During the rainy season, 1962, we
extinct horse, antelope, small ground squirrels, and observed local Audubon cottontails eating the young
other fauna dating to the Early Ajuereado phase grass that sprouted after heavy rain. Callen’s study,
(Zones XXV-XXVIII, Coxcatlan Cave). These rabbits appended to this chapter, shows that they also ate
are too large to be cottontails; occasional diagnostic corn plants and small fruits.
fragments suggest that they are jack rabbits, but not
Lepus callotis. Because of the lack of comparative ma- Sijlvilagus cunicularius
terial, especially post-cranial, these rabbits have not (Mexican cottontail; conejo grande)
yet been identified, but we wonder if they may not be- Modern specimens collected: One male and two fe-
long to the group of black-tailed jack rabbits still asso- males.

ciated with antelope and small ground squirrel on the Measurements (of two adults, dimensions of male
arid northern plains of Mexico. given first): Total length, 445 and 482 mm. Tail,
52 and 53 mm. Hind foot, 95 and 100 mm. Ear, 75 and

Sijlvilagus spp. , 86 mm. Weight, 1.59 kg.
There are two species of Sijlvilagus in the Tehuacan Present-day distribution: Mexican cottontails are
Valley, the Mexican cottontail and the Audubon cotton- found throughout the valley, populations being high-
tail. Their prehistoric distributions will be discussed est, in relation to Audubon cottontails, in areas of
146
densest ground cover and best-developed alluvial soils, Modern specimens collected: One male.
such as the Rio Salado arroyo and most of Subarea 3. Measurements: Total length, 480 mm. Tail, 205 mm.
Our specimens all came from the Rio Salado floodplain Hind foot, 59 mm. Ear, 24 mm. Weight, 0.80 kg.
near Venta Salada; one female, taken June 21, 1962, Present-day distribution: There are no tree squirrels
was pregnant, with four embryos. in the Tehuacan Valley, but rock squirrels in small
Habits and diet: These animals are nocturnal for- numbers occur along the travertine cliff faces which
agers for grass as well as seeds and pods of tree legumes mark off each step-like descent in altitude made by
( see below ) Our evidence is very scanty, but we won-
. the Tehuacan Valley as one travels south. We collected
der if may not avoid competition with
these cottontails one specimen from the El Riego cliffs and found a par-
S. audubonii by eating somewhat different things with- tial skeleton of another on the Ajalpan plain at Qua-
in the same environment. They are, as mentioned chileo, not far from the rocky bluff where the Rio
above, particularly common on the kind of deep al- Zapotitlan enters the valley.
luvial soils where mesquite and other tree legumes Habits and diet: The specimen we collected was
abound. shot in the afternoon, June 26, 1962. Its stomach con-
Prehistoric distribution of cottontails and jack rab- tents (see below) revealed it had been eating cactus
bits: Some idea of the suitability of each subarea as a fruit and mesquite beans, as well as other plants.
habitat for Mexican or Audubon cottontails is given Three limb bones of a large
Prehistoric distribution:
by the prehistoric remains. Assuming that our cotton- ground squirrel, probably Citellus, were found in Zones
tail sample for each period is random, it would appear XXV and XXVI of Coxcatlan Cave in association with
that higher proportions of Mexican cottontails were extinct horse, turtle, and antelope. The only other re-
present in Subarea 1 (both Ajalpan and El Riego lo- mains of this animal were limb bones in Zone A of El
calities) than in Subareas 2, 3, and 4. The accompany- Riego Cave, East Niche (Venta Salada phase). Rock
ing tabulation ranks each of these habitats on the basis squirrels are still common in the El Riego area.
of percentages of Audubon and Mexican cottontails
in the whole sample from a given prehistoric locality,
Small Ground Squirrel
under present-day climatic conditions (7000 b.c.-a.d.
1500). The gradient of Mexican cottontails, from 47 There are no small ground squirrels in the Tehua-
percent of the sample in the Ajalpan plain, near the can Valley today, but a small Sciurid (which may be
Rio Salado, to only 16 percent in the dry Arroyo Lencho a prairie dog or a chipmunk) occurred in the Tehua-
Diego, suggests that S. cunicularius was always more can Valley during the Early Ajuereado phase, in asso-
at home on the alluvial valley floor, while S. audubonii ciation with extinct horse and antelope.
predominated heavily in areas of sparsest grass-and-
weed-patch ground cover. Subarea 2 would probably
Family Geomyidae
have had the lowest Mexican cottontail population,
were it not that our sample comes from Tecorral Can- Gophers have never before been reported from the
yon, a specialized and moister, more bushy niche with- Tehuacan Valley. Archaeological remains, however, in-
in that subarea. dicate that gophers have been present in the area for
7000 years. Prehistoric specimens fall into two cate-
gories: a large gopher whose bones are found in sites
Percent of Cottontail Sample on the deep alluvial soils of the central valley floor,
and a smaller gopher occurring in the shallower al-
Habitat Audubon Mexican
Subareal (Ajalpan plain) 53 47 luvium of narrow canyon floors in Subareas 2, 3, and
Subarea 1 (El Riego) 55 45 4. According to Senor Ticul Alvarez of the Instituto
Subarea 2 (Tecorral Canyon) 65 35 de Biologfa, who has examined these remains, it is
Subarea 3 ( Coxcatlan 72 28 probable that two unrecorded species of gophers are

Subarea 4 (Arroyo Lencho Diego) 84 16 involved. His tentative identifications are as follows:
Heterogeomys sp. Found at sites in the Ajalpan and
Order Rodentia El Riego plains, generally in Subarea 1. This is the
larger of the two gophers.
Family Sciuridae
Cratogeomys sp. Found in caves located on dry bar-
Citellus (
Spermophilus ) variegatus rancas with shallow alluvial bottoms such as occur
(grey rock squirrel; ardilla) in Subareas 2, 3, and 4.
147
Family Heteromyidae Coxcatlan region Liomys was contemporary with ex-

Dipodomys phillipsii
tinct horse and antelope.
(kangaroo rat; raton canguro)
Family Cricetidae
Modern specimens collected: Two females.
Measurements: Total length, 255 and 260 mm. Tail, Reithrodontomys fidvescens
(harvest mouse; local name unknown)
150 and 170 mm. Hind foot, 37 and 39 mm. Ear, 13 mm.
Weight, 38 and 42 gms. Modern specimens collected: None.
Present-day distribution: This one of the more
is Present-day distribution: Known to occur from
common mammals in the Tehuacan Valley, both on the Teotitlan to Tepanco, north of Tehuacan (Hall and
valley floor and in the thorny barranca areas. Night Kelson 1959: 592-93).
after night we observed these small rodents foraging Prehistoric distribution: Reithrodontomys is found
for seeds. We found their burrows on the barren Venta in owl pellets from El Riego Cave, dating to about a.d.
Salada steppe, the hills in front of Coxcatlan Cave, and 700.
west of Ajalpan on an open alluvial plain with occa-
sional mesquite trees. Our specimens, both lactating Peromyscus spp.
females, were collected from Subareas 1 and 3 on July Peromyscus were found
Remains of in virtually all
4 and 14, 1962.
caves in all subareas, but it was not always possible
Habits and Kangaroo rats are apparently active
diet: to determine which species was present. Both P. leuco-
all year at Tehuacan; we observed them in the hearts
pus and P. melanophrys are definitely known to be
of the rainy season (July) and dry season (January). present in the Tehuacan area (Miller and Kellogg 1955:
They were mainly active between 9:00 p.m. and mid- 491, 503), and it is probable that many other species
night, but some could still be seen at 3:00 a.m. on fa- Peromyscus are more
Remains
occur there also. of
vorable nights. common in the levels with extinct horse and antelope
Prehistoric distribution: This rodent has been in the than in any of the more recent levels. They appear in
Valley since at least 6800 b.c. (Zone XXIV, Coxcatlan small frequency, however, in all subareas throughout
Cave). It has always been most common in Subarea the archaeological sequence (see Table 15).
3 (see Table 15).
Peromyscus melanophrys
(deer mouse; local name unknown)
Liomys irroratus
Modern specimens collected: One female.
(spiny mouse; local name unknown)
Measurements: Total length, 240 mm. Tail, 140 mm.
Modern specimens collected: Three. Hind foot, 24 mm. Ear 24 mm. Weight 38 gms.
Measurements: Tail, 125 mm. Hind foot, 28 mm. Ear, Present-day distribution: This mouse is known from
14 mm. Length and weight could not be taken.
Tehuacan, Teotitlan, and Cuicatlan (Miller and Kel-
Present-day distribution: The three specimens were logg, Joe. cit.).
taken from the stomach of a female bobcat (Lynx Habits anddiet: Deer mice were observed mainly
rufus) shot at midnight, July 1962, near Venta
8, between 9:00 and 11:00 on moonless nights, and even
Salada.They constitute our only record of Liomys. then their presence was unpredictable. On some nights
Habits and diet: The three mice had apparently (for example, July 23, 1962) they seemed to be every-
been active at midnight on the alluvial plain near the where, especially in the branches of mesquite trees
Rio Salado, for their cheek-pouches contained seeds ( Dipodomys was abundant on this same
terrestrially
of various kinds, among them mesquite and prickly night). On mouse could be
other nights, hardly a deer
pear (see below). seen. Our specimen collected near Coxcatlan Cave at
Prehistoric distribution: Archaeological specimens midnight August 2, already had a stomach full of in-
of spiny mice are found in all caves located near alluvial sects (see below), which it had been hunting on a
soils which will support tree legumes. Highest
popu- mesquite-and-cactus alluvial apron at the head of a
lations probably existed Subarea 1, both on the
in long barranca leading down from the cliffs to the Rio
Ajalpan plain and on the well-watered alluvium near Salado.
El Riego. Liomys was scarce or absent in Subarea 2
except for alluvial canyon floors (see Table 15). Scat- Baiomys musculus
tered specimens were found in all subareas. In the (pigmy mouse; local name unknown)
148
Modern specimens collected: None. but is present in deposits later than that date in El Riego
Present-day distribution: Known from Teotitlan and Cave in Subarea 1.
Tepanco in the Tehuacan Valley (Hall and Kelson

1959: 661).
Neotoma alleni
Prehistoric distribution: Our only find of Baiomys (Allen’s wood rat; local name unknown)
was a palate and a mandible in owl pellets from El Modern specimens collected: None.
Riego Cave, dating to about a.d. 700. Present-day distribution: N. alleni may be the most
common wood rat in the southern or arid-tropical part
S igmodon hispidus of the Tehuacan Valley; it is outnumbered in the north-
(cotton rat; local name unknown) ern end by N. mexicana, according to data from upper-
Modern specimens collected: None.

most levels in the Tehuacan caves. It has been pre-
Present-day distribution: Throughout the valley;
viously collected at Tehuacan and Teotitlan (Hall and
especially common in the southern half.
Kelson 1959: 706).
Prehistoric distribution: Cotton rats were present in Prehistoric distribution: Allen’s wood rat has been
the Tehuacan Valley almost from the beginning of the present in Subarea 3 since the Early Ajuereado phase.
present climatic regime. Their remains go back to Zone Its remains occur throughout the valley, in all sub-
6 in the West Niche of El Riego Cave (before 6500 areas, from the time of extinct horse and antelope un-
til the Spanish Conquest.
b.c.) and Zone Q in Purron Cave (before 4300 b.c.). Re-
mains are not really common, however, until about The accompanying tabulation shows the suitability
1000 b.c., after which time they are abundant in Sub- of various habitats within the Tehuacan Valley for the
areas 3 and 4 (see Table 15). Owl pellets dating to about five main genera of small rodents recovered from the
a.d. 700 (see below, p. 156) suggest that during that
caves. The figures represent the percentage each genus
period cotton rats may have constituted about 28 per- constitutes of the total sample of the five genera from
cent of the small rodents in the vicinity of El Riego archaeological sites in each subarea for the period 7000
Cave. b.c. to a.d. 1500.
Neotoma spp. Cotton Deer Spintj Wood Kanga,

Habitat rat mouse mouse rat rat
Wood rats were presentTehuacan Valley as
in the
Subarea 1
far back as our record goes; they are particularly com-
(El Riego) 25 12 35 16 12
mon in canyon areas (see Table 15). At least two dif- Subarea 2
ferent species are known through archaeological re- (Teeorral
mains. Canyon) 29 13 10 38 10
Subarea 3
Neotoma mexicana (Coxcatlan) 39 8 17 12 24
(Mexican wood rat; local name unknown) Subarea 4
( Arroyo
Modern specimens collected: None. Lencho Diego) 43 7 14 29 7

Present-day distribution: N. mexicana may be the
most common wood rat in the northern or arid-temper- Order Carnivora
ate part of the valley, but it is largely replaced or out- Family Canidae
numbered by N. alleni in the southern part, according
Canis la Ira ns
to evidence from the uppermost levels of the caves.
(coyote)
Prehistoric distribution: During the Early Ajuereado
phase, N. mexicana reached Subareas 3 and 4 in large Modern specimens collected: One male and one fe-
numbers, contemporary with extinct horse and ante- male.

lope; also possibly found in association with ante-
it is Measurements (dimensions of male are given first):
lope bones in the lowest levels of Teeorral Cave Total length, 1130 and 1016 mm. 300 and 280 mm.
Tail,
(Subarea 2). After the transition to present-day cli- Hind foot, 190 and 183 mm. Ear, not measurable; hides
matic conditions, as the southern half of the valley had been cut off. Weight, not taken; specimens had
gradually became arid-tropical, N. mexicana may have been fleshed and gutted.
retreated to the north, for it is absent from cave de- Present-day distribution: In the Cerro El Zapote area
posits laid down since 7000 b.c. in Subareas 3 and 4 of Subarea 3 reportedly five coyotes per month are
149
sighted. They are hunted on the hills east of Chapulco Present-day distribution: A and mandible
fresh skull
today. Members of the Tehuacan Project saw coyote of this fox were collected in May 1962 along the trail
feces at various places and observed a coyote cross the from the highway to Coxcatlan Cave (Subarea 3). A
highway near Zapotitlan in April 1963. We were also hide was purchased in the Ajalpan market in June
offered a young specimen from the Rio Salado thickets 1962 by Arturo Arvide of the Tehuacan Project. The
near Ajalpan, captured in Julv of 1962. Probablv the only live specimen sighted was seen at 8:00 a.m. by its
major coyote populations surviving in the valley today burrow near the banks of the Rio Zapotitlan, just up-
are in Subareas 3 and 4. stream from its confluence with the Rio Salado. All
Habits and diet: The two coyotes we collected were three occurrences suggest that dens are probably in
shot on successive nights during the week prior to the more nearly forested parts of the valley, either in
December 24, 1962, while they were attempting to thorn forest (Subarea 3) or in the quebrachales and
break into a sheep pen in the thorn forest on the slopes cane thickets along the main watercourses, but the
of Cerro El Zapote, near Coxcatlan. animal ranges over the whole valley at night.
Prehistoric distribution: Coyotes were present in vir- Habits and diet: Mainly nocturnal and extremely
tually all subareas in prehistoric time; however, their shy, this fox is probably much more common in the
remains are more frequent most ancient levels.

in the Tehuacan Valley than our limited records show. We
In Subarea 3 and perhaps in Tecorral Canyon, their got no stomach contents from collected specimens, but
remains are found with those of extinct horse and these foxes are relatively omnivorous (Leopold 1959:
antelope. 410). They probably make use of the abundant ro-
dents and wild fruits in Subarea 3 and the river arroyos.
Canis familiaris
Prehistoric distribution: Remains which are clearly
(domestic dog; perro)
those of Urocyon cinereoargenteus are found in sites
Modern specimens collected: one female, from Cox- in all subareas dating from at least 6500 b.c. to the
catlan. Spanish Conquest.
Hind foot, 164 mm. Ear, 100 mm. Weight, 8 kg. Extinct Fox
Present-day distribution: Throughout the valley.
Remains of a fox whose bones are considerably
Prehistoric distribution: Domestic dog first appears
larger and more rugose than those of Urocyon were
in theTehuacan sequence at the start of the late pre-
found in Early Ajuereado levels, contemporary with
ceramic Abejas phase. Our earliest records of it are
extinct horse and antelope.
teeth and mandibles, dating to about 3200 b.c. in Cox-
catlan Cave and to about 2500 b.c. in Purron Cave.
Family Procyonidae
Dogs increase in frequency during the early and mid-
Bassariscus astutus
dle Formative periods, after which they are common
in all levels in all subareas. In most cases there is abun-
(ring-tailed cat; cacomixtle)
dant evidence that they were being eaten, at least as Modern specimens collected: One female.
early as 1500 b.c. That dogs are definitely not present Measurements: Head and body, 430 mm. Tail, 330
earlier, in spite of the large sample of bone remains mm. (bitten off in a fight). Hind foot, 79 mm. Ear, 45
from the period before 3000 b.c., suggests that dogs mm. Weight, 1.59 kg.
were not kept in the Tehuacan V alley until a level of Present-day distribution: Our only record of ring-
considerable agricultural efficiencv had been reached. tailed cat during 1962 was the specimen mentioned
They are, in other words, characteristic of the early above, shot at night on June 30 by a farmer in the Ba-
village farming stage but ofno previous period. The rranca de los Mangos near Calipan. The animal had
romantic concept that dog was man’s faithful com- apparently descended to the edge of the valley floor
panion during the stage when man was a hunter and from the foothills above Calipan and was feeding on
gatherer seems to be completely inaccurate as far as mesquite beans and insects (see below). Nothing fur-
the Tehuacan Valley is concerned. ther is known of the distribution of this comparatively
rare nocturnal animal in the Tehuacan area.
Urocyon cinereoargenteus The only archaeological
(gray fox; zorro gris)
record of Bassariscus comes from the Early Ajuereado
Modern specimens collected: One, incomplete; no phase (Zone XXVI of Coxcatlan Cave), found in asso-
measurements. ciation with extinct fauna.
150
Procyon Jot or Present-day distribution: Striped skunks are common

(raccoon; mapache) in all parts of the Tehuacan Valley, but nowhere did
they seem quite as abundant to us as in the center of
Modern specimens collected: One male.
the valley and in the thorn-scrub forest near Coxcatlan
Measurements: Total length, 800 mm. Tail, 2<0 mm.
Cave. Hardly a night trip to Subarea 3 failed to pro-
Hind foot, 120 mm. Ear, 60 mm. Weight, 5.0 kg.
duce a sighting of one or two skunks, whose immediate
Present-day distribution: The distribution of rac-
reaction to our flashlights was to bound off into the
coons in the valley follows the watercourses regardless
brush with a warning wave of the tail. Our specimen,
of subarea. We observed their foot tracks along wet
pregnant with two fetuses, was collected near Calipan,
barrancas both in forested mountain areas and between
June 22, 1962.
nearly barren limestone cliffs. They are not uncommon
Habits and Beetles and beetle larvae appar-
diet:
in the El Riego springs area. We collected a specimen
ently make up good percentage of this animal’s diet
a
from the Barranca de los Mangos near Calipan, where
(see below), although according to Leopold (1959:
the animals reportedly collect when the cultivated fruit
459), it is less dependent on these foods than is the hog-
trees are bearing. Raccoon is not one of the most prev-
nosed skunk. We found Mephitis foraging by night
alent animals in the region, however, probably be-
along the cut-away banks of dry arroyos and on the
cause of the extensive waterless stretches in the valley.
talus slopes below cliffs.
Habits and diet: The male we collected was dis-
covered at 5:00 a.m. on July 2, 1962. His intestinal

Conepatus mesoleucus
contents indicated a dinner of cactus fruit and insects
(hog-nosed skunk; local name unknown)
(see below).
Prehistoric distribution: Our oldest record of Modern specimens collected: None.
Procyon in the Tehuacan area comes from Zone X of Present-day distribution: Unknown, but probably
Coxcatlan Cave Subarea 3, dated to about 3200 b.c.
in
more common in the pine and oak forest above the
valley than in the valley itself.
Certainly raccoons must have been present earlier, but
not being as common as some of the other small mam-
Conepatus is represented by
mals, they probably were not as often taken by pre-

mandibles and other fragments in Coxcatlan Cave
historic hunters. Remains were frequently found in
during the Early Ajuereado phase. Scattered finds
village sites located near the Rio Salado.

occur throughout the sequence, especially in Subarea 3.
Family Felidae
Family Mustelidae
Felis concolor
Spilogale augustifrons
(puma or mountain lion; leon)
(spotted skunk; zorrillo manchado)
Modern specimens collected: None.
Modern specimens collected: None. Present-day distribution: Unknown, but Frederick
Present-day distribution: Unknown, but recent
A. Peterson of the Tehuacan Project sighted a puma in
archaeological remains suggest Spilogale is far com-
the thorn forest near Coxcatlan Cave in 1961. Today
moner in the canyon areas of Subareas 2, 3, and 4 the puma is almost extinct in the area.
than in the center of the valley, where Mephitis is
Prehistoric distribution: Puma bones were found in
prevalent.
Zone XXII of Coxcatlan Cave (El Riego phase) and in
Prehistoric distribution: Spilogale is a contemporary Zone D of Las Canoas village site, near Ajalpan ( Santa
1
of extinct horse and antelope in the lowest levels of Maria phase).

Coxcatlan Cave. It is present during all phases in all
caves located in canyon areas, especially Coxcatlan and Lynx rufus

Tecorral. It is not known from any of the sites located ( bobcat; gato montes
in the Apalpan plain.
Modern specimens collected: One female.
Mephitis macroura
Hind foot, 148 mm. Ear, 69 mm. Weight 4.55 kg.
(hooded or striped skunk; zorrillo rayado)
Present-day and prehistoric distribution: Nothing is
Modern specimens collected: One female. known of the present-day distribution of the bobcat in
Measurements: Total length, 615 mm. Tail, 330 mm. the Tehuacan region beyond the occurrence of the
Hind foot, 60 mm. Ear, 30 mm. Weight, 0.9 kg. specimen mentioned above. This would seem to be
151
one of the most southerly occurrences of Li/nx rufus on seem to have been in the Rio Salado thickets, the El
record (cf. Hall and Kelson 1959: 970). Leopold (1959: Riego area, and Subarea 3.
fig. 182) does not extend its range farther south than the
Family Cervioae
Valley of Mexico, and he comments (p. 484) on the
puzzling scarcity of bobcats in south-central Mexico in Odocoileus vi rg i n ia n us
spite of the occurrence there of habitats resembling (white-tailed deer; venado cola blanca)
those in which this species thrives in northern Mexico: Modern specimens collected: None.
areas combining brush, rocky canyons, and abundant Present-day distribution: White-tailed deer are ex-
rodents. The Tehuacan Valley is such an area, and bone tremely rare in the Tehuacan Valley today, owing to
fragments from Zone XXI of Coxcatlan Cave (El Riego the pressure of year-round hunting. Relict populations
phase) indicate that bobcats have been present there exist in thesurrounding hills, however, especially in
8000 years. Subareas 1 and 3, with their
for at least Subarea 3 and in the valley of the Rio Calapilla. To the
high rodent populations, probably were and are its south, across the Oaxaca border near Teotitlan del
preferred habitat. Camino, hunters in December of 1962 reportedly killed
Habits and diet: The female bobcat we collected was four or five.
shot at Venta Salada northeast of the confluence of the Prehistoric distribution: That Odocoileus was once
Rio Salado and Rio Zapotitlan. It had been hunting common in the valley is well documented archaeologi-
spiny mice Liomys irroratus), and three of the latter
(
callv. Over 1500 identifiable fragments of deer, the re-
were recovered from its stomach, each neatly bitten in mains of at least 174 individuals, testify to its presence
half (see above). as far back as 7000 b.c., by which time it had apparently
Order Perissodactyla
replaced antelope as the most common artiodactyl of
the area. Highest prehistoric deer populations appear
Family Equidae
to have been in the forested river arroyos and in the
Equus sp. (extinct horse) thorn-forest cover of the Coxcatlan-Calipan area, as
well as the area of the El Riego springs; they were
Remains of horse occur in
the lower three zones (XXVI-XXVIII) of Coxcatlan evidently scarce in Subarea 2, even in the wetter bar-
rancas.
Cave, and date to before 7000 b.c. Some of these re-
mains match bones of extinct horse from the Pleisto- Family Antilocapridae
cene gravels in the Valsequillo area of the Valley of
Remains of antelope were present in the four lowest
Puebla (Arenillas locality), collected by Professor Juan
zones (XXV-XXVIII) of Coxcatlan Cave. Because we
Armenta.
had no well-studied comparative material at hand in
Order Artiodactyla Tehuacan, we could not determine whether the re-
Family Tayassuidae mains were those of now-extinct Pleistocene antelope
or simply represented a southerly subspecies of the
Pecari ( Taijassu ) tajacu
American prong-horn Antilocapra umericana, but it is
(collared peccary; javali, coche de monte)
probably the latter. These antelope had apparently dis-
Modern specimens collected: One male from the appeared from the Tehuacan Valley by 7000 b.c.
valley of the Rio Calapilla.
Reptiles
Measurements: No measurements were taken by the
hunters, who ate the flesh but saved us the bones (De- The Tehuacan Valley has a rich variety of reptiles,
cember 20, 1962). but only those eaten by man today or apparently eaten
Present-day distribution: The collared peccary is vir- prehistorically were studied. Hence this short list is
tually extinct within the Tehuacan Valley floor area to- culturally determined, and in no sense reflects the
day, but survives in the nearly uninhabited drainage of reptilian ecology of the valley.
the Rio Calapilla.

Order Chelmia
Prehistoric distribution: Archaeological records
show the peccary present, though never really abun-
Family Kinosternidae
dant, for 8000 years or more in all subareas with suffi- Kinosternon integrum
cient surface water. Highest prehistoric populations (mud turtle; tortuga, galapago)
152
Modern specimens collected: One male and one fe- Order Squamata
male.
Family Iguanidae
Measurements ( dimensions of male are given first )
Length of carapace, 155 and 135 mm. Weight with Iguana iguana
shell, 0.5 and 0.5 kg. Weight of turtle alone, 0.35 and (
green iguana, iguana verde)
0.30 kg. Modern specimens collected: One male.
Present-day distribution: As far as was able to
I Measurements: Total length, 1160 mm. Tail, 780 mm.
establish, this is the only turtle living in the Tehuacan Hind foot. 145 mm. Weight, 2.45 kg.
Valley today. It is restricted to the Rio Salado and its Present-day distribution: This is a basically tropical-
tributaries, just as is the green iguana (see below). One forest species whose presence in the arid Tehuacan
of our specimens, a female, was collected in the Rio Valley is quite surprising. It is restricted to the arroyos
Salado at 2:00 p.m. on July 27, 1962, during a of the Rio Salado, Rio Zapotitlan, and Rio Calapilla,
which time it was actively engaged
light drizzle, at in presumably using them as the corridor through which
eating water bugs. The male turtle was recovered at it originally entered the valley from the Veracruz coast,
5:00 p.m. on July 28 while buried in the mud at the by way of the Rio Santo Domingo. To the best of my
bottom of a small pond on a spring-fed tributary stream knowledge, this iguana is not known as far north as the
of the Rio Salado, near Venta Salada. city of Tehuacan under present-day conditions. Today,
Habits and diet: The female had been eating aquatic the green iguana survives in the southern part of the
insects and snails (see below). The male had been feed- valley, confining its activities to the most humid local
ing on leafv plants around the margins of the tributary environments, where it lives on the moist vegetation
pond. Local villagers report that they have seen these of the river bank.
animals leave the water to eat carrion, but I cannot Habits and diet: The specimen we collected near the
document this. The female was gravid, with 11 eggs. Rio Zapotitlan was active at noon June 20, 1962. Its
The male (an elderly specimen) had a thick layer of stomach contained only the leaves of tree legumes.
mossy vegetation adhering to the top of his carapace, Prehistoric distribution: Unknown.
which may aid in camouflage.
Ctenosaura pectinata
Prehistoric distribution: These small turtles have
(black iguana, ground iguana; iguana negra, iguana
been used as a source of food since about 5000 b.c. in
pinta)
the vicinity of the Rio Salado and its tributaries. This
includes Tecorral Canyon, which is presently dry for Modern specimens collected: Three.
most of the year, but which, on the basis of abundant Measurements (one adult specimen): Total length,
prehistoric mud-turtle remains, may have had a more 790 mm. Tail, 500 mm. Hind foot, 75 mm. Weight, 1 kg.
reliable flow of water in ancient times. Turtle scutes Present-day distribution: This is the most common
are most frequent in sites near Ajalpan, along the banks iguana in the Tehuacan Valley. One of our specimens
of the Rio Salado, especially during the period from was caught near Ajalpan, while foraging along the al-
1500 to 500 b.c. luvial plain among mesquite trees. Ctenosaura is not
restricted to wet areas as is Iguana and ranges over

most of the southern part of the valley, especially in
Pleistocene Turtle?
the dry canyons around Purron Cave ( Subarea 4 ) and
Remains of a much larger turtle were abundant in Coxcatlan Cave (Subarea 3). It is also found in the Rio
the four lowest zones of Cave (XXV-
Coxcatlan Salado thickets west of Coxcatlan, and at the base of
XXVIII) and in Zone C of Tecorral Cave in association the travertine cliffs on the west edge of the valley n£ar
with antelope. Much of the fragmentary material has the point where the Rio Zapotitlan breaks through to
not been firmly identified, but the homy cover from enter the valley floor. These animals are common from
an anterior vertebral scute found in a later level (and Ajalpan south, but we have yet to discover them above
presumably redeposited) at San Marcos Cave appears the 1600-meter contour.
to be Gopherus berlandieri. The Texas gopher tortoise Habits and diet: These iguanas are active mainly
is precisely the type of turtle one would expect to share during the rainy season, at which time they roam
a steppe environment with horse, jack rabbit, and ante- widely; they can also be observed late in the dry sea-
lope in the Early Ajuereado phase. son, when they tend to stay closer to their burrows.
153
They bite when cornered and sometimes will attempt conditions. Their bones are found in Zone XXIV at Cox-
escape by climbing low trees, although they seem to catlan Cave (before 6500 b.c.). From then until the
be mainly terrestrial. Our Ajalpan specimen was active Conquest, remains of Ameiva are found in all subareas,
at 3:30 p.m., June 18, 1962. The amazing range of diet and are useful rainy-season indicators. The types of
revealed by Callen’s study (see below), including fruit, parts discarded and the kind of charring observed on
insects, and spiders, makes the omnivorous character some of the bones indicate that these lizards were eaten
of this animal seem one key to its widespread distribu- by the valley’s prehistoric inhabitants.
tion. Comments by some authors (Smith 1946: 34;
Schmidt and Inger 1957: 118) that Ctenosaura is ex- Birds
clusively vegetarian would seem not to apply to this This study makes no attempt to do justice to the
Mexican species. Ditmars (1940: 28), in his description varied bird life at Tehuacan. Since our principal prob-
of Ctenosaura multispinis, states that captive black lem was the diet of prehistoric man, we concerned our-
iguanas “will take young birds and small rodents and selves only with the local game birds, or other birds
are surprisingly agile in catching the prey ... In a wild occasionally eaten in the area.
state [black iguanas] are undoubtedly carnivorous to
a considerable extent.” Family Anatidae
This iguana has several color phases. Although it is
Anas cyanoptera
almost always black when adult, it is often piebald or
(cinnamon teal; pato)
covered with gray-white or pinkish-gray patches when
young. Because of this, local villagers believe that there Modern specimens collected: One female.
are two ground-dwelling iguanas in the Tehuacan Measurements: Folded wing, 177 mm. Tail, 58 mm.
Valley. Tarsus, 34 mm. Bill, 40 mm. Weight, 340 gms.
Prehistoric distribution: Black iguanas appear in Present-day distribution: Ducks are scarce today in
archaeological deposits in Purron Cave (Subarea 4)
the Tehuacan Valley, this being the only species with
and Coxcatlan Cave (Subarea 3) as far back as 500 b.c. which the present-day residents are familiar. One
specimen donated to the project was shot in flight near
Family Teiidae the Pueblo Nuevo dam on the Rio Salado at ten o’clock
Ameiva undulata on a slightly overcast winter morning, December 21,

(large race-runner lizard; lagartija) 1962. The area backed up by the dam is a pond sur-
rounded by swampy alluvium full of marsh grass,
Modern specimens collected: Five.
palo dc agua, and other plants. It is flanked by mud
Measurements (one adult): Total length, 386 mm.
flats which end at the three-meter-high banks of earth
Tail, 250 mm. blind foot, 50 mm. Weight, 110 gms.
and gravel setting the river off from the nearby hills.
Present-day distribution: During the summer these
The ducks share this environment with the American
race-runners are abundant throughout the valley, espe-
egret Casmerodius albus).
(
ciallyon barren rock faces such as the El Riego cliffs
Habits and diet: These ducks arrive in the Tehuacan
below La Mesa. They can also be observed scurrying
Valley in November, and the few that survive the win-
along the valley floor in open areas like the Ajalpan
ter hunting season leave sometime in February. The
plain.
largest flock we saw contained fifty individuals, but
Habits and diet: These race-runners seem to be
groups of five to ten are far more frequent.
active in the valley during the rainy season, from May Unknown.
to September, when their insect food sources are most
abundant. On a single day in July one may see hun-
Family Phasianidae
dreds on the El Riego cliffs, and on a return visit to the
Colinus virginianus
same spot in December may not see a single individual.
When (bobwhite quail; codorniz)
present, the race-runners are most active during
the hot part of the day, skimming over the rocks with Modern specimens collected: One male.
blinding speed. Stomachs which I examined indicate Measurements: Folded wing, 110 mm. Tail, 62 mm.
a diet of small insects and arachnids, including June Tarsus, 33mm. Bill, 15 mm. Weight, 150 gms.
beetles and pseudoscorpions. Present-day distribution: Bobwhites are common in
Prehistoric distribution: These lizards have been weed patches along the Rio Salado and in areas where
present since the beginning of present-day climatic ground cover is fairly dense, such as the underbrush
154
around the El Riego springs or the floor of the Tecorral branches of the river-bank trees,from one end of the
Canyon. Our specimen is from Ajalpan. valley to the other. We collected specimens from
Habits and diet: These cpiail are active daily in thickets along the Rio Salado, from the thorn forest
overgrown cornfields and other weedy areas, collecting at Coxcatlan Cave, and from open cornfields south of
seeds of various kinds (see below). Ajalpan. Local villagers collected scores of others for
Prehistoric distribution: Quail have been eaten dur- their own tables.
ing all of Tehuacan prehistory, but specimens indis- Habits and diet: White- winged doves are active all
tinguishable from the bobwhite do not occur until year in Tehuacan Valley, for surprisingly long
the
Zone XXIII of Coxcatlan Cave (about 6500 b.c. ). Their stretches daily. They eat a wide variety of foods (see
remains are found through all subsequent periods, par- below). Specimens we collected in the Rio Salado que-
ticularly at sites in Subareas 1 and 3. brachales at noon on June 20 had been eating tree
fruits; others had been eating cactus fruits; and one,
Family Meleagridae captured as late as 10:00 p.m. on July 25, had been
Meleagris gallopavo eating cutworms. The doves seemed to be almost as
(domestic turkey; pavo) active at 10:00 that night as they had been at 10:00
in the morning.
Modem specimens collected: One male.
Present-day distribution: Virtually everywhere in the
Prehistoric distribution: Z enaida has figured in the
Our specimen
food supply of man for almost as long as we have any
valley. from Coxcatlan.
is
record of human habitation in the Tehuacan Valley.
Domestic turkey bones first
Zone XXII of Coxcatlan Cave (after 6500 b.c.) con-
appear in the Tehuacan sequence early in the Palo
tained a humerus of this genus, and there are many
Blanco phase, about a.d. 180 (Zone VI, Coxcatlan
other occurrences at all time periods in all subareas.
Cave). Thereafter turkeys increase in number, and by
Most of these dove bone fragments match perfectly
the time of the Conquest constitute about 10 percent
of the animals eaten in the valley, according to archae-
with modern skeletons of Z enaida asiatica, and it may
be that most (if not all) of them are white- winged doves
ological remains. There is no evidence of the wild tur-
or some closely related species.
key at Tehuacan during any period, and it is unlikely
that its range ever extended that far south (see Leopold
1959: 268-71). Coin mb igalli na passer i na

(ground dove; paloma canela, “torito”)
Family Charadriidae Modern specimens collected: One from Venta Sa-
Charadrius vociferus lada.
(killdeer; chichicuilote) Present-day distribution: These doves are common
Modem throughout the Tehuacan area, not only on the valley
specimens collected: One from the Rio Sa-
floor but also in the streets and plazas of the villages.
lado.
Prehistoric distribution: Fragments of ground dove
Present-day distribution: Killdeer are common along
the banks of the Rio Salado during the winter, and are
were uncovered in Zone XV of Coxcatlan Cave (about
occasionally killed for food by local villagers. 5000 b.c.). Later finds are sporadic and cannot always
lie specifically assigned to ColumhigaUina passerina.
Prehistoric distribution: Unknown.
Family Columbiidae Family Tytonidae

Z enaida asiatica Tyto alba
(white-winged dove; paloma de alas blancas) (bam owl; teeolote grande)
Modern specimens collected: Five. Modern specimens collected: None.

Measurements (average Folded wing 162
of five): Present-day distribution: Throughout the valley,
mm. Tail, 107 mm. Tarsus, 26 mm. Bill, 24 mm. Weight nesting occasionally in caves.
150 gms. Habits and diet: A large clutch of owl pellets was
Present-day distribution: These doves are among the found in the East Niche of El Riego Cave, interpolated
most common birds in the valley and are found in every between Zones D and C and dating to just before a.d.
environmental subarea. They can be seen roosting on 700. Their contents give us a look at the diet of barn
cacti in the most barren areas or perched high in the owls in the El Riego area at that time. The remains in-
155
eluded ten spiny mice ( Liomys irroratus), eight cotton Chordeiles is also common in the valley, but no speci-
rats (
Sigmodcm hispidus), three deer mice (
Peromys - mens were collected.
cus spp.), three kangaroo rats (
Dipodomys phillipsii),
Family Corvidae
three wood rats (
Neotoma spp.), one harvest mouse
( Reithrodontomys fulvescens), one pigmy mouse Corvus corax
( Baiomys musetdus), one murine opossum ( Marmosa (raven; cacalote)
canescens), one juvenile gopher, two juvenile cotton-
Modern specimens collected: One.
tails ( Sijlvilagus sp.), and one small bird. Since the cli-
Measurements: Folded wing, 405 mm. Tail, 221 mm.
matic conditions about a.d. 700 were not strikingly dif- Tarsus, 82 mm. Bill, 76 mm. Weight, 1 kg.
ferent from those of today, the modem diet of Tyto in
Present-day distribution: Very common throughout
the El Riego area is probably similar to the list above. the valley.
Prehistoric distribution: A and tarsometatarsus
tibia
Habits and diet: Our specimen, collected June 20,
of Tyto alba were found in Zone XI of Coxcatlan Cave 1962, just south of Ajalpan, had been eating com, liz-
(roughly 4000 b.c.). Owl pellets probably attributable to ards, and a variety of insects (see below).
Tyto are common in El Riego Cave, and in most cases Prehistoric distribution: Remains of ravens from as
it seemed likely that this owl was the animal respon-
early as 1500 b.c. were recovered at village sites in the
sible for bringing small rodents into the caves.
Ajalpan area.
Family Caprimulgidae The Pre-Columbian Faunal Remains

Caprimvlgus ridgicayi Nearly 11,000 bones or fragments of bones were re-
(buff-collared nightjar, Ridgway’s whippoorwill; le- covered from the archaeological excavations in the Te-
chuza) huacan Valley. These bones included 4,713 skeletal
Modem specimens collected: One. components which could be identified both as to type
Measurements: Folded wing, 160 mm. Tail, dam- of animal and part of skeleton. The remaining frag-
aged. Tarsus, 24 mm. 22 mm. Weight ments were too incomplete to identify, being mostly
Bill, 43 gms.
splinters or flakes from the shafts of long bones which
Present-dav distribution: This bird is a very common
nocturnal inhabitant of the valley, from El Riego to preserved no portions of the articular surface or any
Coxcatlan. During the day other diagnostic attribute. Fairly accurate rough
it can often be surprised
while nesting on the ground under mesquite trees and counts were kept on these unidentifiable fragments,
bushes along the valley floor. and each was examined for signs of butchering
technique.
Habits and diet: One specimen collected at Coxcat-
lan Cave at dusk (7:10 p.m.) on July 30, 1962, had been The 4,713 identified components represented the re-
eating a variety of local insects (see below). mains of at least 1,013 individual animals (see below).
Prehistoric distribution: Unknown. Most of thesewere animals eaten bv the prehistoric

inhabitants of the valley, some were small rodents de-
Chordeiles acutipennis posited in the caves via owl pellets, and others were
(lesser nighthawk; lechuza) fully-articulated uneaten animals which had either
Minimum No. of Unidentifiable

no. of skeletal fragments
animals components ( round nos.)
Animal food remains
Coxcatlan Cave and Terrace 381 2,555 4,000
Purron and Abejas caves 79 233 300
Tecorral and San Marcos caves 57 180 100
El Riego Cave (East and West) 109 416 250
Ajalpan, Coatepec, Las Canoas,
and Quachilco 178 683 1,100
Animals not used as food
Small rodents (all sites)
Other intrusive animals
192
17
597
49 —
150
Totals 1,013 4,713 5,900
156
been buried in the cave or died there of natural causes. juvenile animals, while the single right was from an
Some, in the uppermost levels of the caves, had been aged individual, and so on.
completely mummified by the dry Tehuacan climate. For deer, we used studies like those of Villa (1954),
Counts of all these categories of animals are given in which allowed us to measure age on the basis of tooth
the accompanying tabulation. and antler development. Thus we could reason with
All identifiable fragments of bone were checked justification that the many-tined antler of a three-year-
against one or more of the following collections: (1) old buck, found in a given level, could not belong to
skeletons of animals which I collected in the Tehuacan the deer represented in the same level by a mandible
Valley and other regions of southern Mexico and in with an as-yet-unerupted first molar, since this tooth
Guatemala; (2) skeletal material in the collections of comes through at an age of six months; two deer had
the Instituto de Biologia, Universidad Naeional Auto- to be present. We were cautious with antlers, knowing
noma de Mexico, Mexico City; (3) skeletons of Mexican that man might bring back shed specimens to the cave
animals in the collections of the Chicago Natural His- for tool-making, and only accepted an antler if we
tory Museum; and (4) fossil collections from Valse- could find fragments of its unshed base. Shed antlers
quillo, Puebla, at the Museo de la Revolucion in the were not counted in the animal remains. Thus our
city of Puebla.® estimates of “minimum number of individuals” for
Once all the identifiable hones from a given natural each species are the result of a number of approaches
level or zone at a particular site had been cheeked, an to the problem, done with care and (we hope) over-
effort was made to determine the minimum number of looking as little as possible.
individuals present for each species. This was done by Still other methods were used in order to determine
the method worked out by White (1953: 397), who the season of the year during which a given zone was
suggests: deposited. Some animals, like the large race-runner liz-
ards (
Ameiva common along the El Riego cliffs, are
)
The method I have used on butchering
in the studies
active only during the rainy season; when their bones
technique is to separate the most abundant element of the
were present, we guessed that the level had been laid
species found (usually the distal end of the tibia) into
right and left components and use the greater number as
down sometime between June and September. Animals
the unit of calculation. This may introduce a slight error
probably traceable to winter or dry-season occupa-
on the conservative side because, without the expenditure tions included the sandhill crane, which today visits
of a great deal of time with small return, we cannot be Mexico from November to April and summers in the
sure all of the lefts match all of the rights. United States and Canada, although in the past it may
have nested in Mexico (see Leopold 1959: 281; Martin
This method is coming
be used increasingly by other
to del Campo 1944).
faunal analysts (see Lundelius’ study in Alexander Other seasonal evidences were derived from deer
1963: 513), and we gave it only a slight modification: antlers and even deer fetuses, the latter of which were
we expended the “great deal of time with small return” not uncommon in Coxcatlan Cave. Villa’s studies on
in order to see if all the lefts matched all the rights.
time of antler drop by deer from Mexico show, as
Frequently they did not, and our figures were hence
pointed out by Leopold (1959: 512), that “the first bucks
changed. The most common difference observed was
to lose their antlers seemed to be those from the south
—
one of age that is, the lefts might all be from very
— —
Oaxaca and Guerrero where the breeding season is
earlier.” February, March, and possibly early April
appear to be the usual months during which Villa’s
° I was greatly aided in this study by other zoologists, in-
southern specimens lost their antlers; new ones began
cluding specialists in small rodents and birds, two categories
with which I was unfamiliar. Professor Bernardo Villa R. kindly to appear during April and May. Through the rainy
allowed me access to collections at the Instituto de Biologia, season the antlers were in velvet, and by the end of
where Sr. Ticul Alvarez identified many of the small rodents
from the Tehuacan caves. Two visiting American ornithologists,
October they had reached their “forma decidua defin-
Allen Phillips and Robert Dickerman, took time from their busy itiva,” fully grown and ossified right to the end of the
schedules to identify bird bones. Study of the Chicago Natural tines (Villa 1954: 459). Regarding bones of fetal deer,
History Museum collections was made possible by Drs. William
Turnbull and Philip Hershkovitz, and more of our Tehuacan
we relied on Leopold’s assembly of figures (1959: 510-
rodents were identified there by Dr. Joseph Moore. Professor 11), which suggest from Chihuahua to Guerrero
that
Juan Armenta of the Universidad Autonoma de Puebla allowed the time of fawning may
vary from June to September.
us to examine fossil mammals from the Valsequillo gravels. My
grateful thanks go to these men for the role they played in mak-
Roughly the same time period would be the one to
ing this paper possible. which late-term fetuses might lie attributed, and we
157
have accordingly done so, especially since there was a Perhaps eighteen of these are small rabbits which ap-
good correlation between bones of fetal deer and bones pear to be Audubon cottontails. The others are much
of Ameiva lizards in rainy-season levels. larger, and certainly are jack rabbits, but since they
One final word of caution for the reader: the per- differ from the forms found in the Tehuacan Valley
centages given for certain species of animals in the today, I did not feel secure enough in my identification
total remains from a level, used in the discussion below, to break them down into fine categories. Hence, all
are not to be taken too literally. They are calculated these animals are presented in Table 16 simply as “Late
from very small samples and should serve only as a Pleistocene rabbits,” with the comment that they are
rough guide to the activities and food preferences of a mostly jack rabbits. Present but not so abundantly rep-
given period. resented are coyotes, foxes, skunks of three kinds ( spot-
ted being the most common, with striped and hog-
The Sequence from Coxcatlan Cave nosed skunks in second and third places), ring-tailed
and Terrace, Subarea 3 cats, large rock squirrels, small ground squirrels (which
For a variety of reasons, Coxcatlan Cave lies in per- may be either prairie dogs or chipmunks), and quail
haps the best hunting locality in the Tehuacan Valley, and other birds. One hundred and two scutes, repre-
and the bone remains reflect this. From its vantage senting at least seven turtles of a large type no longer
point in the Cerro Agujereado, 975 meters above sea present in the area, also appeared in the debris, as did
level, this big rock shelter looks north over a rolling fragments of snake and lizard. Less than 10 percent of
mountain talus covered with the densest thorn forest the individual animals were horses and antelopes, the
observed in the whole valley system. To the east is a large fauna usually so abundant in “kill sites” from this
high sierra, also covered with scrub thorn, which gives period (see, for example, Hester 1960).
way to pine and oak forest some 500 meters above the The finding of great piles of foot bones of jack rab-
cave. To the west, over a low ridge of hills, lie the bit in a few excavated squares of the cave suggests that
valley floor and the wide arroyo of the Rio Salado, these extremities were often trimmed off and discarded
hidden by quebrachales and dense stands of cane. all in one corner of the occupation floor. Antelope were
Hunters camped at Coxcatlan Cave could have for- probably brought back intact to the rock shelter, judg-
aged in the Venta Salada area and returned home in ing by the variety of bones found, which included ribs,
half a day, or ascended to the pine and oak vegetation vertebrae, inner ears, and other head components. We
belt to cut maguey or hunt deer and peccary. The plat- are not equally sure that horses were brought back
form of the cave gives an unobstructed view of many intact, since our sample of bones of this species is too
square miles of varied country. Some brief mention of small to be conclusive. Long bones of the larger ani-
the fauna of Subarea 3 has been made above and we mals are not badly smashed as they often are in
will only repeat here that almost all the animals found later periods, and many of the long bones of the smaller
in the Coxcatlan area occur in the immediate vicinity of animals are completely unbroken, although occasion-
the cave. ally they are charred.
Earh/ Ajucreado Phase Late Ajuereado Phase
Irregular occupation levels in the yellow rock-dust This period is characterized by tools of the same
at the bottom of the shelter yielded the bones of ani- type as Early Ajuereado, but a completely modern
mals either extinct or no longer present in the Tehua- fauna seems to be present. White-tailed deer are now
can Valley. The implications of this fauna in terms of being hunted and represent about 40 percent of the
paleo-climate have been discussed above. This was a individual animals recovered. Rabbits have dropped
period which in most of North America is rather in- to 30 percent, and all specimens are the Audubon and
adequately labeled “paleo-Indian," there being usually Mexican cottontails common in the valley today. Rob-
so little evidence concerning human subsistence pat- white quail, ameiva lizards, and hog-nosed skunks
tern during the period that no more descriptive term round out the remains from Late Ajuereado levels.
can be used. The lower four zones of Coxcatlan Cave The presence of hardened antler tines (in their Oeto-
show that during this period, which occurred before ber-February condition), fragments of fetal deer (prob-
7000 b.c. and possibly as early as 10,000 or 9000 b.c., ably from the May to August period), and such rainy-
man in the Tehuacan Valley ate considerable amounts season indicators as ameiva lizard suggest that the
of small game. Sixty-nine of the individual animals Late Ajuereado occupations at Coxcatlan Cave rep-
recovered from the Early Ajuereado phase are rabbits. resent camps made during several seasons of the year.
158
El Riego Phase ments of a fetus carried by one of the does, which indi-
cated a late rainy-season camp; remains of cottontail
This is the most abundantly represented phase in and fox were also found in this half. In the western part
Subarea 3, with ten zones from Coxeatlan Cave and of the same zone lay the debris from another rainy-
Terrace. Over 400 identified bones of deer, represent- season camp where deer, cottontails, and birds had
ing the remains of at least forty-two individuals of all been eaten. Zone XIV was an extremely large occupa-
ages and both sexes, show that white-tails were about tion divided into an east section, and upper-west and
40 percent of the animals eaten in this area during the lower-west sections. All seasons of the year are repre-
El Riego phase. Cottontails constituted 35 percent of sented, with dry-season indicators mostly confined to
the individual animals, and we have evidence of three the bottom part of the zone, and spring or rainy-season
peccary, three foxes, both spotted and hooded skunks, indicators present principally in the upper part.
one mountain lion, one bobcat, and two opossums. A There is more evidence of the prehistoric butchering
mud was brought to the cave (possibly from the
turtle technique in El Riego phase levels than in any other
Rio Salado), and ameiva lizards were collected, prob- phase in this subarea. In the case of deer, virtually
ably in the vicinity of the cave. Several doves were every part of the body is represented, from long bones
eaten, as well as quail and songbirds. In Zones XVIII and hooves down to skull components, which are sel-
and XX, in a rainy-season context, fragments of large dom preserved in Tehuacan sites fragments of zygo- —
fish were found. Since no fish of this size exists in the matic arch, hyoid, squamosal, occiput, rim of the orbit,
Tehuacan Valley today, we can neither identify it nor and others. Also present are bits of sternum, rib, verte-
explain it, except to hazard a guess that the Rio Salado bra (even some from the base of the tail), and blade of
flowed deeper then than it does today, or that its the scapula, which has convinced us that the entire
sloughs were more extensive. carcass was brought intact to the rock shelter. The long
All seasons are represented in the faunal debris, and bones of the deer during phase (and especially be-
this
some of the occupations must have been quite long. tween Zones XIX and XIV) are smashed to a degree
Zone XXI of Coxeatlan Cave contained fragments of that could not have happened accidentally, even under
deer fetus and hardened antler, suggesting a camp conditions of heavy rock fall. Hundreds of fragments
made during the late rainy and/or early dry seasons. are no larger than a toothpick, and all have been
Zones XX
and XIX both have enough ameiva lizard re- splintered by heavy percussion blows. This bone-
mains to brand them as rainy-season occupations. splintering technique is unlike anything found in the
Three more zones from this phase were so large that other phases at Coxeatlan Cave, and it undoubtedly
they could be subdivided into separate occupations reflects extensive use of bone marrow. Most deer long
which took place within a short span of years. Zone bones show some sign of roasting before the marrow
XVIII had an extremely large drv-season occupation w'as extracted.
in its east end, with the remains of three deer of differ- Our hypothetical reconstruction of El Riego butcher-
ent ages, two cottontails, a skunk, and a songbird, while ing technique, based on fragments of bone, cut marks,
the west half of the same zone, a rainy-season camp, type of skeletal components found and their condition,
had fish, lizard, dove, three cottontails, and a spotted is as follows. First, the entire carcass of a deer was
skunk, as well as three deer whose remains indicated brought back to the cave and laid out on the floor. The
they had been brought intact to the rock shelter. hide w'as then removed, the last few 7
tail vertebrae be-
The pattern in Zone XVI was somewhat similar. ing carried away with Next the animal was gutted
it.
Debris in the west end of the cave shows that this part and, if it was a pregnant doe, the fetus removed. If
of the zone was a dry-season camp, with bones and the deer was a buck, the antlers were removed by
hardened antler tines of deer and many fragments of breaking them away from the frontal bone, and later
cottontail. The east half of this same zone had exten- (sometimes) these were cut up into antler hammers and
sive refuse from all seasons of the year, including tine pressure -flakers. The shoulder joint must have been
ameiva lizards and fragments of fetal deer (May to cut through with heavy cleavers, for the upper end of
September), full hardened antlers ripped from the the humerus is virtually never found. A blow just above
frontal bone (October to February), and a spongy, the distal end of the metapodial separated the feet,
only partially-developed antler which had been in vel- which w'ere often discarded as a unit. After the meat
vet when the deer was killed (April to June?). had been cut off, the long bones were roasted and then
In Zone XV, the east half of the rock shelter showed systematically pounded into splinters with heavy stone
the butchered remains of at least three deer, plus frag- tools so the marrow could be completely extracted.
159
Table 16. Food Animals from Coxcatlan Cave and Terrace (Subarea 3) by Zone and Phase
Late
Early Ajuer- Sta. Palo Venta
Ajuereado eado El Riego Coxcatlan Abejas Maria Blanco Salada
XXVIII
XXVII
XXVI
XXV
XXIV
XXIII
XXII
XXI
XVIII
XVII LLJ XIII VIII
CO
XVI
XX XIX
Q XV XIV XII
XI X C_> IX VII
< VI
V IV III II
50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
LO LTD LO Totals
CO CO
Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc h~ Tc Tc Tc Tc Tc Tc £ Tc Tc Tc f— Tc Tc Tc Tc Tc Tc
Extinct Horse (indivs.) 1 1 1 3
Fragments 2 1 4 7
Antelope (indivs.) 1 1 2 3 7
Ramus, mandible 1 1
Teeth 2 3 5
Petrous bone 3 4 7
Scapula 1 1
Radius 1 1
Innominate 1 2 3
Tibia 1 1
Metapodial 3 1 4
Phalanx 2 9 4 15
Other 1 6 7
White-tailed Deer (indivs.) 2 2 2 2 2 4 6 1 9 2 6 8 2 3 9 4 1 3 2 4 1 2 2 2 1 2 1 85
Antler 1 1 1 1 1 1 4 2 3 2 5 1 2 3 28
Teeth 2 1 3 8 1 5 7 2 1 1 31
Ramus, mandible 1 1 2 4 5 5 4 7 1 3 6 4 2 1 4 1 2 2 2 57
Petrous bone 7 2 1 6 4 11 2 2 8 2 2 1 1 1 50
Scapula 1 2 3 1 1 2 1 2 2 1 2 1 19
Humerus 1 3 1 1 4 4 3 1 4 2 1 3 1 29
Radius 1 1 2 1 1 2 7 1 16 1 5 6 1 2 9 1 2 3 2 2 2 68
Ulna 2 1 2 3 1 8 7 3 3 3 9 6 2 1 1 1 53
Innominate 2 1 1 1 1 2 3 11
Femur 1 1 2 3 2 4 1 1 3 1 2 1 1 1 1 1 1 27
Tibia 1 1 2 3 3 1 6 1 4 10 1 4 6 1 2 1 3 2 1 1 54
Calcaneum 1 1 2 1 2 5 5 1 1 3 3 25
Astragalus 1 1 1 2 5 1 7 2 4 1 10 5 1 3 4 3 51
Naviculocuboid 1 2 1 3 1 1 4 1 1 2 2 1 20
Metapodial 2 4 3 4 3 3 10 3 15 11 17 3 6 22 6 1 5 7 8 6 2 1 1 1 144
Phalanx 1 1 1 4 1 2 1 6 3 15 10 4 10 8 2 2 2 73
Vertebra 2 2 1 3 1 1 1 1 1 5 6 3 15 8 1 6 2 6 7 1 73
Rib 1 1 2 3 5 1 3 2 1 2 5 5 2 2 6 6 3 50
Other 2 2 2 3 3 7 4 7 6 5 6 6 5 2 1 3 1 65
Frags, foetus 1 2 2 4 3 1 3 1 2 2 21
Peccary (indivs.) 1 1 1 1 1 1 1 7
Maxilla 1 1 2
Ramus, mandible 2 2
Scapula 1 1
Humerus 1 1
Tibia 1 1
Astragalus 1 1 2
Metapodial 1 1 2
Phalanx 1 1 2
Other 2 2
Late Pleistocene Rabbits (indivs.) 8 5 40 16 69
Maxilla 4 2 21 2 29
Ramus, mandible 4 1 24 4 33
Humerus 8 5 42 16 71
Radius 8 3 18 14 43
Ulna 9 22 7 38
Innominate 11 7 63 16 97
Femur 4 3 35 16 58
Tibia 3 5 47 20 75
Calcaneum 8 2 27 29 66
Other 45 5 100 140

290
10 4 6 3 6 2 2 1 7 4 2 2 2 2 72
Recent Cottontails (indivs.) 2 1 2 2 2 3 5 1 1
Maxilla 1 1 3 2 2 3 1 1 8 2 24
Ramus, mandible 1 1 1 4 6 4 7 1 4 2 2 1 3 2 1 40
3 14
Humerus 1 1 1 1 1 1 1 3 1
Radius 1 1 2 1 1 6
2 1 11
Ulna 1 1 1 1 1 1 1 1
Innominate 1 1 1 3 4 1 6 1 4 1 4 1 4 2 2 1 2 39
5
Late
Early Ajuer- Sta. Palo Venta
Ajuereado eado El Riego Coxcatlan Abejas Maria Blanco Salada
XXVIII
XXVII
XXVI XXIII XVIII
XXIV XXII XVII
XXV XXI VIII
XX XIX XVI D-E XV XIV XIII XII XI X C IX
VII A,B VI V IV III II 1
50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 51, 50, 50, 50, 50, 50, 50, 51, 50, 50, 50, 51, 50, 50, 50, 50, 50, 50,
Totals,
Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Ts Tc Tc Tc Tc Tc Tc Ts Tc Tc Tc Ts Tc Tc Tc Tc Tc Tc
Recent Cottontails (cont.)

Femur
Tibia
1
1
1 1
1
112
3 1 5
12
2 3
1 5
3 3
1 4
3
1 2 1 2 27
25
Calcaneum 1 1 1 1 4
Other 1 1 4 11 3 1 5 3 1 6 1 2 3 10 2 1 1 1 57
Recent Jackrabbit (indivs.) 1 1
Fragments 1 1
Coyote (indivs.) 1 1 2
Fragments 1 1 2
Fox (indivs.) 113 1 1 1 1 1 1 1 1 1 1 1 16

Fragments 9 5 26 9 2 1 1 2 1 3 3 1 2 3 68
Skunk (indivs.) 2 4 1 2 1 1 1 1 2 1 1 1 18
Fragments 3 6 1 2 2 3 3 1 4 2 2 1 30
Ring-tailed Cat (indivs.) 1 1
Fragments 4 4
Rock Squirrel (indivs.) 1 1 2
Fragments 1 3 4
Prairie Dog or Chipmunk (indivs.) 1 1 1 3

Fragments 1 1 2 4
Puma (indivs.) 1 1
Fragments 1 1
Bobcat (indivs.) 1 1 2
Fragments 1 3 4
Opossum (indivs.) 1 1 1 3
Fragments 1 1 1 3
Raccoon (indivs.) 1 1
Fragments 1 1
Gopher (indivs.) 1 1 2
Fragments 1 1 2
Dog (indivs.) 1 1 1 2 2 1 8
Fragments 1 1 1 5 3 1 12
Extinct Turtle (indivs.) 112 3 7
Fragments 16 1 33 52 102
Mud Turtle (indivs.) 1 1
Fragments 1 1
Snake (indivs.) 1 1
Fragments 1 1
Iguana (indivs.) 1 1 1 1 4
Fragments
Other Lizards (indivs.)
Fragments
2?
27 2
1
90
6? 3?
45
1
1
111
17 1
12
12 1
1
1
4
6
2
1 1
1
8
1
3
1
18
26?
187
Ouail (indivs.) 1 1 1 3
Fragments
Dove (indivs.)
Fragments
2 1
111
111
1
1 1
1 1
1
4
Song-birds (indivs.) 1 1
2-
Fragments 1 1 2
Crane (indivs.) 1 1
Fragments 1 1
Barn Owl (indivs.) 1 1
Fragments 2 2
Raven (indivs.) 1 1 2
Fragments 1 2 3
Hawk (indivs.) 1 1 2
Fragments 1 1 2
Turkey (indivs.) 1 1 1 3
Fragments 1 4 1 6
Other Birds (indivs.) 3 1 3 1 1 1 2 2 1 15
Fragments 4 1 4 1 1 1 2 2 1 17
Fish (indivs.) 1 1 1 1 4
Fragments 1 1 1 1 4
Coxcatkm Phase ing, equal to that of the Coxcatlan phase. The zones
The zones occupied during phase are interesting,
this
are primarily dry-season occupations, where hunting
in that most are dry-season camps. During at least one
far outweighed trapping. Further, the introduction of
of these dry-season occupations, deer appear to have the domestic dog took place early in this period. The
latter first appears in Zone X, dated about 3200 b.c.,
been intensively hunted. In fact, for the whole of the
phase in the Coxcatlan area, deer represent about 43 and stands therefore as the oldest evidence of domestic
percent of the individual animals recovered. Cotton- dog in Mesoamerica proper. Dog remains occur also
in the Abejas phase of Purron Cave, as will be noted
tails account for 30 percent, and other animals include
peccary, fox, and spotted skunk. During the Coxcat- below. In both cases the evidence involves teeth which
lan phase several species of birds, including the barn matched well with dog specimens in our collections and
owl and a very large bird tentatively identified as a differed drastically from those of the only other large
canid in the area, the coyote. Abejas dogs were much
sandhill crane, were included in the diet. Probablv the
smaller than coyotes, probably weighing only ten to
presence of crane in Zone XIII reinforces our evidence
for a dry-season occupation, for these birds tend to
fifteen pounds.
winter in southern Mexico. Both Allen Phillips and

Forty-five percent of the individual animals recov-
Robert Dickerman, who examined the specimen, stress ered from Abejas levels at Coxcatlan Cave and Terrace
caution in the matter of seasonality in view of a find of were deer, with cottontails representing only 23 percent
and the difference made up by peccary, fox, hog-nosed
sub-fossil crane eggs in the Valley of Mexico (Martin
skunk, and raccoon. The occupants of these levels also
del Campo 1944), which suggests these cranes may
ate doves and did some fishing. This represents the last
have stayed for the summer nesting season occasionally
in the past.
evidence of fish in the prehistoric sequence in the Te-
Once seem huacan Valley.

again, deer to have been brought whole to
the rock shelter for butchering, but during this phase
The entire carcasses of deer were still brought to the
there no evidence of the meticulous smashing of long
is
rock shelter, but there are larger numbers of intact, un-
bones seen in El Riego times. Some bones, in fact, seem

smashed long bones in this phase than in any of the
preceding ones. Less frugal wringing of nourishment
to have been discarded while still articulated with their
neighbors and held together with ligaments —espe- from deer was noted in the bone debris, probablv be-
cause agricultural products were beginning to repre-
cially bones of the lower leg like the metatarsal, astrag-
sent a significant part of the food supply. For the first
alus, navieulocuboid, calcaneum, and others. The en-
tire foot of a Mexican was found discarded
cottontail
time, some bones appear to have been cut part way
while articulated in Square E15 of Zone XI, and an
still
through with stone blades and then snapped. This may
articulated elbow-joint of Audubon cottontail was be related to the fact that prismatic obsidian blades
found in Zone XII. Once the feet were trimmed off. make their first appearance during Abejas times (see
these smaller mammals appear to have been roasted Volume II).

over hot coals (which produced a rather spotty char- Santa Maria Phase
ring pattern) and the limbs then eaten as units with no
disjointing of their individual components.
A hiatus in the Subarea 3 sequence follows the aban-
donment of Zone VIII in Coxcatlan Cave. Then comes
A to predominantly dry-season occupations
shift
Zone VII, a heavy occupation during the later Forma-
after the end of the El Riego phase is suggested by the
tive Santa Maria phase. Three divisions of the zone,
lack of ameiva lizards and the numbers of hardened or
into a west half and upper and lower east sections, give
“winter-season” antlers, coupled with an absence of
differing evidences of seasonality. The west half is a
the deer fetuses so abundant in El Riego times. That
rainv-season camp during which the occupants trapped
intensive deer hunting was one phase of this drv-sea-
cottontailsand spotted skunk, hunted deer and peccary,
son activity is suggested not only by the nearly 200
and and ameiva lizards. The lower part of
ate iguana
identifiable fragments of white-tail in these three zones
the east half of Zone VII, containing bones of ameiva
of Coxcatlan Cave, but also by large numbers of projec-
lizards, was a rainv-season camp with little sign of
tile points, including a number of types new to this
hunting, while the upper part had many deer bones,
phase (see Volume II: The Nonceramic Artifacts).
including antlers from bucks taken during the dry sea-
son. Trapping of rabbits would seem to have been more
Abejas Phase
intensive during the occupation of the west-end rainv-
This phase is characterized by intensive deer hunt- season camp; deer hunting was more intensive during
162
the occupation in the east end (probably because a dry- The Sequence Arroyo Leneho Diego,
in the
season camp was included here, if we are to judge by Subarea 4
Abejas and Coxcatlan phase customs). The deer bones ( Purron Cave and Abejas Cave
show charring, cutting with stone (obsidian?) blades, The Arroyo Leneho Diego is typical of the dry can-
and the occasional discarding of articulated hocks and yons in Subarea 4, whose natural resources have
lower limb fragments. The lack of dogs in this zone is already been discussed. Two rock shelters located near
view of the
interesting, in fact that in village sites of
the head of the canyon offered a long sequence of fauna
the Santa Maria phase as — will be pointed out below from the area. The two caves are located close together,
a larger proportion of dogs were eaten than during any but in significantly different situations. Purron Cave
other phase in the Tehuacan sequence.
(Tc 272) lies in a blind canyon 50 to 60 meters deep,
with no view except that of the opposite wall of the
Palo Blanco Phase arroyo. Large game could not have been sighted from
the cave, although small mammals could have been
The archaeological levels attributable to this phase
trapped in the area. Abejas Cave (Tc 307), located just
reflect mostly rainy-season camps of some duration.
around the corner from Purron, has an excellent view
Trapping of small game was practiced, cottontails rep-
of a small dissected alluvial apron on the slopes of the
resenting 21 percent of the animals recovered and
hills overlooking the main part of the Tilapa plain.
skunks about 11 percent; foxes and gophers are also
From here, herds of deer or peccary could easily be
present. Some hunting of deer and peccary was done,
seen as they moved through the area. The nearest sur-
with remains of the former constituting 18 percent of
face water is a tiny trickle a kilometer away, on the
the total animals. Iguana and smaller lizards seem to
other side of the ridge of hills which separates Abejas
have been collected, and a variety of wild birds in-
Cave from the main Tilapa plain. The topography is
cluded doves, songbirds, ravens, and others. At least
one of high rock cliffs, dry round-top hills, narrow can-
one domestic turkey and four domestic dogs were
yons, and thorn-covered alluvial fans, at an elevation of
eaten; this is the first evidence of the domestic turkey
925 meters above sea level.Bone remains show that the
in the Tehuacan sequence. Remains of ameiva lizard
fauna collected in the area by prehistoric man was
and fetal deer place all of the cave occupations some-
limited to deer, cottontails, skunks, gophers, foxes,
where in the period between fune and September.
iguanas and other lizards, and quail and other small
birds.
Venta Salada Phase

Remains from these levels are similar to those from O Phase
El Rievo
the Palo Blanco phase, with deer constituting only 14 There are nine occupations in the arroyo from this
percent of the total. Cottontails are only 14 percent period, but only four types of animals are involved.
of the individual animals represented, and fox, skunk, The Table 17) suggest that deer and
totals (given in
bobcat, opossum, and gopher are included in the cottontails each constituted about 40 percent of the
smaller game. At least three dogs and two turkeys were animals taken. Also present are birds and ameiva liz-
eaten,and a number of wild birds were used for food. ards, the latter indicating that at least two (and prob-
Iguana and ameiva lizards are also among the remains. ably most) of the occupations were during the rainy
An problem of seasonality is posed by
interesting season. As might be expected in this arid subarea, with
Zone which had two sections. One was an occupa-
II, little ground cover available, all but one of the rabbits
tion floor, and the other was tightly packed debris on are Audubon cottontails.
which the floor rested. The sub-floor debris appears to Numerous fragments of ribs, vertebrae, hooves, and
have accumulated during the middle of the rainy sea- such delicate cranial elements as the hyoid suggest
son, since Ameiva and Iguana were present, but no re- that deer carcasses were brought intact to the cave and
mains of deer. The occupation floor is possibly attrib- butchered there. The finding of calcanea, astragali, and
utable to the late rainv season or early fall; it con- distal ends of tibiae in articulation indicates that the
tained remains of one deer and a fawn, as well as a hocks were discarded as a unit after being separated
gray fox whose first molar was just erupting. Foxes are from the rest of the hind limb. Many of the long bones
born from March to May in southern Mexico (Alvarez have been split open longitudinally so that marrow
del Toro 1952: 218), and the first molar would certainly could be extracted. Countless bone splinters from these
have come through fully by late summer or early fall. levels show the marks of heavy stone choppers. The
163
Table 17. Food Animals from the Arroyo Lencho Diego
(Subarea 4) by Site and Phase
(Minimum Numbers of Individuals Derived from Identified Fragments
Cox- Ajal- Santa

El Riego catlan Abejas Purron pan Maria Pale Blanco
3: c/3 or o u_ uj o o Q_ O O CD ^ —
fbi CD CD 3 CD CD
White-tailed Deer (indivs.) 1 1 11111 2 111 1 1 20

antler 1
teeth 1 1 1 9
ramus, mandible 1 1
scapula 1 2
humerus 1 4
radius 1 1 5
innominate 1 1
femur 4
tibia 3 1 9
astragalus 4
calcaneum 4
metapod ia I 2 1 1 14
carpus/tarsus 1 2
phalanx 1 2
vertebra 4
rib 1 1 1 12
other 1 4
Cottontails (indivs.) 1 1 12 11 1111 2 6 12 2 111 36

maxilla 1 1 5
ramus, mandible 1 1 4 13
1111
1
scapula 2 1 2 11
humerus 3 4 2 1 1 12
radius 1 1 1 2 6
ulna 2 3
innominate 1 1 1 1 3 2 13
femur 1 3 3 2 1 10
tibia 1 1 1 1 3 1 11
metapodial 1 2 6
other 1 2 1 7
Dog (indivs.) 1 1 4
teeth 1 2 3
ramus, mandible 1 2
radius 1 1
vertebra 1 5
rib 2 3
Skunk (indivs.) 2
fragments 3
Gopher (indivs.) 1
fragments 1
Fox (indivs.) 2
fragments 12
Iguana (indivs.) 2
fragments 5
Other Lizards (indivs.) 4
fragments 8
Quail (indivs.) 1
fragments 2
Other Birds (indivs.) 7
fragments 9
164
deer involved are of allage groups, from fawns and Ajalpan Phase
yearlings to adult does and four-point bucks, showing Zone Purron Cave contained the remains of a
J of
a random sample of the local population of white-tails. deer, a domestic dog, and one Mexican cottontail.
Numerous as fragments of deer are, in only one case
is there evidence that more than a single individual was
Santa Maria Phase
present in a given zone. In all but two cases, this is true
This is the first phase in the subarea to produce evi-
of the cottontails as well. Our impression, therefore,
dence of occupations of some duration. For the first
is that occupations in the Arroyo Lencho Diego during
time there is evidence of significant amounts of trap-
the El Riego phase were mostly rainy season camps of
ping of the locally available small game, including
brief duration, during which little effort was made to
twelve cottontails, as well as gopher, black iguana,
exploit the small game of the area beyond trapping a
quail,and an unidentified small bird. Also present were
few cottontails. The band probably moved on as soon
as they killed whatever deer or other large game hap-
the remains of dog and white-tailed deer. Both the dog
and cottontail remains show signs of roasting, and the
pened to be in the vicinity.
deer bones are split lengthwise for extracting marrow.
Coxcatlan Phase The dog remains from Zone II of Purron Cave in-
Hunters camped only once in the Arroyo Lencho cluded a complete right mandible, 103 mm. long, with
six cheek teeth intact. The specimen has no first pre-
Diego during this phase; they killed a single deer and
then moved on. Remains of ameiva lizard suggest that molar (lacking even the alveolus), but this may be a
the camp was made during the rainy season. genetic defect of this particular individual. Dogs from
Santa Maria levels in other subareas do not lack the
Abejas Phase first premolar.
Six zones in Purron and Abejas caves date to this

period, and all show significant contrasts with the El Palo Blanco Phase
Riego and Coxcatlan phase levels in the same caves. Six occupations make this one of the best-represented
First of all, there is a complete lack of ameiva lizards periods in the prehistory of the Arroyo Lencho Diego.
or other rainy-season indicators, and a virtual absence The pattern of hunting has more similarities to the
of small game except The five deer killed
for cottontails. Santa Maria phase than to any of the preceramic pe-
during this phase are represented almost exclusively by riods. Cottontails represent about 45 percent of the
long-bone fragments; skull elements, hooves, and all individual animals present, and there is evidence for
bones below the metapodial are completely lacking, as trapping of fox, striped skunk, black iguana and other
are vertebrae. Thus, these animals may have been lizards, a few small birds. Also eaten were white-
and
butchered at the site of the kill and selected parts tailed deer and domestic dogs. Remains of ameiva liz-
brought to the cave from some distance away. Long ard indicate that Zone C of Purron Cave was a rainy-
bones were split for marrow, and hocks were discarded season occupation. The other zones from this period
with all three components still articulated. Our sus- lack Ameiva remains, but they are so similar in other
picion, although based on admittedly slim evidence, is characteristics that we suspect they too may be rainy-
that these six occupations are brief dry-season camps season occupations of the same type. Both of the deer
by wide-ranging microbands, during the time of year involved are immature.
when wild resources were at their lowest ebb. Occupations during the Palo Blanco phase, like those
In Zone L of Purron Cave (about 2500 b.c.) appears during the Santa Maria period, do not seem to have
the first evidence for domestic dog in Subarea 4. The been particularly brief, but there is no evidence of in-
evidence (a single tooth) is painfully slim, but is sup- tensive hunting.Whatever their reasons for being in
ported by the finding of dog remains from Abejas levels the Arroyo Lencho Diego, obtaining meat was not the
in Coxcatlan Cave, as pointed out above. primary purpose of these campers, and bone remains
reflect mainly their trapping of small game, probably
Purron Phase
incidental to their other rainy-season activities. The
Two brief occupations, Zones K and K
1
in Purron latter mav have included the harvesting of crops irri-
Cave, produced a single Audubon cottontail, one spot- gated by water held by the large dam that was built in
ted skunk, and a small bird. the area.
165
This is the last period for which we have evidence of
cave occupation in the arroyo.
Table Food Animals from Tecorral Canyon

18. The Sequence in Tecorral Canyon, Subarea 2
(Subarea 2) by Site and Phase (San Marcos Cave and Tecorral Cave)
(Minimum Numbers of Individuals Derived from
The ecology Subarea 2 and the specialized can-
of
Identified Fragments)
yon niches within it has already been discussed. Both
Tecorral Cave (Tc 255) and San Marcos Cave (Tc 254)
El Riego
and Palo Venta are shallow rock shelters overlooking the alluvial bot-
Ajuereado Coxcatlan Abejas Ajalpan Blanco Salada
tom deep travertine canyon at an elevation of 1550
of a
meters above sea level and have access to the fauna
C F E D C' C B B A
255, 254, 254, 254, 254, 254, 254, 255, 255, of both the brushy canyon floor and the barren rock
|
Totals,
Tc Tc Tc Tc Tc Tc Tc Tc Tc cliffs and slopes above it. Archaeological remains show
that during the nine or ten thousand years when these

Antelope (indivs.) 1 i
fragments 2 2
rock shelters were sporadically occupied, the animals
White-tailed Deer (indivs.) 1 1 1 3 most commonly eaten in the area were deer, peccary,
fragments 1 6 1 8
Peccary (indivs.)
cottontails, coyotes, foxes, gophers, striped and spotted
1 1 1 3
fragments 1 3 1 5 skunks, turtles, lizards, snakes, and several types of
bird, including dove and quail. Actually, Tecorral
Cottontails (indivs.) 2 1 3 3 1 1 4 2 2 19
maxilla 1 1 2 4 Canyon is a fairly poor hunting locality, and the bone
ramus 2 1 2 1 1 3 1 11
remains reflect this (see Table 18). Small game which
scapula 3 1 2 6
humerus 2 3 3 2 2 12 could be obtained by trapping is quite varied, but

larger mammals like deer and peccary which would
radius 1 1
ulna 1 1 2 1 5
innominate 1 1 3 1 3 9 have been obtained by hunting are poorly represented.

femur 1 2 4 7
In fact the whole aspect of Tecorral Canyon is one of
tibia 1 1 1 2 3 8
metapodial 2 2 an area where man may have camped temporarily to

vertebra 4 4 1 10 1 20 plant or harvest corn or collect wild plants, and only
other 1 8 6 15
incidentally did a little trapping.
Coyote (indivs.) 1 1 2
fragments 1 1 2 Z one C of Tecorral Cave

Dog (indivs.) 1 1
fragments 3 3 This level is unfortunately mixed, making it im-

Fox (indivs.) 1 1 2
possible to assign the bones to a given culture. Many
fragments 2 1 3
Gopher (indivs.) 1 1 1 2 1 6
of the diagnostic artifacts from Zone C are character-
fragments 1 1 4 2 3 11 isticof the El Riego phase (6500 to 4900 b.c.). A num-
Skunk (indivs.) 1 2 3
fragments 1 2 3
ber of the bone fragments, however, belong to an-
imals which were presumably already extinct in the
start of this phase. The
Extinct Turtle (indivs.) 1 1
fragments 2 2
Tehuacan Valley before the
Other Turtles (indivs.) 1 1 1 1 4 matrix of this shallow level was yellow rock dust in
fragments 3 3 3 5 14
which no sublevels could be discerned, and it was only
Lizard (indivs.) 1 1 1 1 4
fragments 1 7 2 2 12 during analvsis in the laboratory that the mixed nature

Snake (indivs.) 1 1
of the deposits became apparent. Perhaps the situation
fragments 2 2
was one which a very shallow Early Ajuereado
in
Quail (indivs.) 1 1
occupation with very few artifacts was closely overlain
fragments 1 1
Dove (indivs.) 1 1
by a brief El Riego occupation and possibly even fur-
fragments 1 1
ther intermixed by intrusive pits. We are left, thus,
with the following situation: Squares W2 and S1W2 of
Song-birds (indivs.) 2 1 3
fragments 2 5 7
Other Birds (indivs.) 1 1 2 zone contain two scutes of large “extinct turtle
this
fragments
and the cheek tooth of an antelope; Squares S2E1 and
3 1 4
S1E2 contained the remains of an ameiva lizard and

a Mexican cottontail. Faunally, Zone C is not very
helpful except to indicate that the rock shelters in
Tecorral Canyon were visited by man before (000 b.c.,
166
and that during the El Riego phase (when the climate with itsWest Niche facing east over the Tehuacan
had become roughly what it is today) they were occa- plain. The slight elevation of the cave is just sufficient
sionally used as rainy-season campsites. to afford it an unobstructed view of game moving

through the area for half a kilometer in three direc-
Coxcatlan Phase
tions. There are many springs along the base of the
Zone F of San Marcos Cave produced cottontail, and more surface water than
cliff, in most of the Te-
gopher, and quail, as well as fragments of coyote and huacan Valley.
mud turtle. The presence of the latter suggests that The ecology of the area has already been discussed.
there was water flowing in Teeorral Canyon at the The cave is in a specialized habitat located at the
time; whether this was seasonal or perennial is un- boundary between the long, level alluvial plain near the
known. Zone E, branded as a rainy-season occupation city of Tehuacan and the barren travertine cliffs behind
through the presence of bones of ameiva lizard, shows it, which are part of Subarea 2. Campers in the cave
the remains of deer and peccary, cottontails and could draw on the fauna of both subareas, as well as
gophers. Once again, fragments of turtle are present. animals that came to the springs for water. Animals
The seasonal information correlates well with the pres- eaten in the area included deer, peccary, cottontails
ence of primitive “improved” corn in this level, which and jack rabbits, gophers, opossums, foxes, skunks,
indicates that Teeorral Canyon was being occupied rock squirrels, snakes and lizards, turtles, doves, hawks,
briefly during the late rainy season when corn was and other birds.
ready to harvest.
Late Ajuereado Phase
Abejas Phase
A very brief occupation in Zone 6, West Niche of El
Zone D of San Marcos Cave, a rainy season occupa- Riego Cave, shows only the remains of Mexican and
tion, was probably also a temporary camp during the Audubon cottontails.
period of the corn harvest. Present are the remains
of three cottontails, a spotted skunk, turtle, ameiva El Riego Phase
lizard, and white-winged dove. Zone 5 of the West Niche, also a very brief occupa-
tion, contained only the bones of one white-tailed deer
Ajalpan Phase
and one cottontail rabbit.
The only bone present in Zone C of San Marcos Cave
is the mandible of an Audubon cottontail. Coxcatlan Phase
Zone 4 of the West Niche suggests that a brief camp

Palo Blanco Phase
was made during this phase. Two cottontails were
This includes three brief occupations, two in San trapped and one deer eaten, along with a jack rabbit, a
Marcos Cave and one in Teeorral Cave, at least one of gopher and one dove. Some long bones of deer appear
which (San Marcos, Zone B) can be traced to the rainy to have been cut on both ends with a flint blade, into
season. The totals for all three zones show seven cot- sections which could have been used as tubes or beads.
tontails, two deer, a peccary, a fox, three gophers,
turtle, lizard,snake, one domestic dog, and three birds. Palo Blanco Phase
The and hoof fragments present suggest that
skull
Two camps made in the East Niche of the cave dur-
carcasses of both deer and peccary were being brought
ing this phase support the contrast (seen in earlier
intact to the cave and butchered there.
levels at other caves) between rainy-season and dry-
Venta Salada Phase season activities. Zone E, a dry-season camp which in-
volved the butchering of a deer with fully hardened
Zone A of Teeorral Cave showed the remains of two antlers, had little small game. Zone D, a rainy-season
cottontails, a peccary, gopher, hooded skunk, spotted camp, was rich in small fauna, including six cotton-
skunk, and two birds.
tails, three opossums, and other forms like gopher, jack
rabbit, bird, and snake. The suggestion is that rainy-

The Sequence at El Riego Springs,
season camps were mostly for wild-plant collecting
Junction Between Subareas 1 and 2
and small-game trapping, with proportionally more
( El Riego Cave, East and West Niches
hunting of deer in the dry season.
El Riego Cave is an excellent location for a hunting Signs of rodent gnawing on bones from Zone D (East
site, since it lies almost at valley level (1650 meters) Niche) show that the bones lay on the surface for some
167
time after the deposition of this layer. Sometime be- the food counts from Zone 3 shown in Table 16.
tween the abandonment of Zone D and the deposition Whether they represent an offering or simply a prosaic
of Zone C, a barn owl inhabited El Riego Cave and left burial of two dead puppies is a problem beyond the
a large clutch of owl pellets, which were later un- scope of this paper. In Zone A of the East Niche we
covered by the excavators, all in a single one-meter found the completely mummified corpse of a striped
square. The abundant small rodent remains from these skunk Mephitis macroura) which had obviously bur-
(
pellets have not been included in the tables of animals rowed into the cave and died. Everything but the odor
eaten by occupants of the prehistoric sites, but are had been preserved.
listed above in the discussion of owl diet.
The Sequence from the Ajalpan Plain, Subarea 1
Venta Salada Phase (Ajalpan, Coatepec, Las Canoas, and Quachilco)
Six zones in El Riego Cave date to this period; Our picture of prehistoric hunting patterns in this
half of these showed evidence of camps made at inter-
part of Subarea 1 comes from the four village sites of
vals throughout all seasons of the year. Cottontails con-
Ajalpan (Ts 204), Coatepec (Ts 368), Las Canoas (Ts
stituted 33 percent of all the individual animals eaten
367),and Quachilco (Tr 218). The first three sites are
at the cave, and domestic animals ( dog and turkey)
located within a kilometer or so of each other near the
about 17 percent. Deer and peccary together consti-
east bank of the Rio Salado, just south of the present
tuted only 15 percent of the individual animals, with
village of Ajalpan. Quachilco lies five kilometers to the
the remainder made up of small species like gopher, southwest. Approximate altitude of the level plain is
opossum, jack rabbit, fox, skunk, rock squirrel, turtle, 1230 meters. The resources both of the alluvial plain
and lizard.
surrounding the sites and of the arrovo of the Rio
In these levels there is considerable evidence of the Salado, a specialized niche within Subarea are avail-
1,
butchering technique practiced during the Venta Sa- able to all four locations.
lada phase. Deer are of all age groups from yearling to The natural resources of Subarea 1 have already been
three-year-old, and the variety of bones suggests that discussed. Archaeological remains tell us that the wild
the carcasses were brought to the cave intact. There
animals selected as staple foods by the villagers in
the head was separated from the trunk by a cleaver
this subarea were white-tailed deer, peccary, cotton-
blow, dividing the atlas from the axis. The antlers were
tails, gophers, and mud turtles. Domestic dogs also
torn from the skull, and some were cut into sections
constituted a large percentage of their food supply.
for use as antler hammers. The deer were roasted,
Also hunted, but in lesser frequency, were jack rabbits,
and some of the bones must later have been kicked into coyotes, raccoons, striped skunks, lizards, quail, hawks,
the fire, for they are calcined almost white. Dogs were possibly ravens, and even an occasional mountain lion.
also roasted and the skull broken open for removal of With the exception of the latter species, most of these
brains. Long strings of vertebrae from these animals are animals would have been found the vicinity of the
in
completely charred, as if the dogs had been roasted village or in the riverbank thickets.
over hot coals. Cottontails, gophers, and opossums also
show charring of this type. Clusters of preserved tur- Ajalpan Phase
key feathers show that these animals were plucked in The Ajalpan phase is distinguished from later pe-
the cave, and bones from their limbs and trunk show riods in the area by more intensive dry-season deer
that they, too, were roasted. Charred skulls of two hunting, greater use of turtles from the Rio Salado,
ameiva lizards and the partial skull of another suggest and relatively lower proportions of small animals such
that these animals were roasted and eaten, all but the as cottontail. Totals from the phase (see Table 20)
snout. The skulls are all clipped off right at the occipital show that deer represented 50 percent of the animals
region, and with one exception the limb bones are com- eaten, domestic dogs about 17 percent, turtles about
pletely gone, suggesting that the whole animal was 11 percent, and cottontails only about 8 percent, with
crunched up in the mouth of the person eating it. peccary, coyote, lizards, and birds making up the re-
Not all the animals found in Venta Salada levels were mainder. Deer bones from middens dating to this phase
eaten, however. In Zone 3 of the West Niche were are frequently roasted or charred and sometimes show
found the complete, articulated skeletons of two buried Many of the bones look as
the marks of stone blades.
dogs, puppies which on the basis of the teeth should though they had been made into blanks for bone tools.
have been two or three months old. Since these animals In Zone G of the Ajalpan site we have evidence that
were obviously not eaten, they were not included in the common deer-metapodial awls of this period were
168
Table 19. Food Animals from East and West Niches of El Riego
Cave by Zone and Phase
Late Late
Ajuer- El Cox- Palo Ajuer- El Cox- Palo
eado Riego cat Ian Blanco Venta Salada eado Riego catlan Blanco Venta Salada
CD LO LU O O CD CO CNJ — . «£
co LO LU CD O CD co OsJ . 1
<£
5 5 sT of q> CD CD ? 5 $ CD
5
LO LO
5
LO
CD
LO LO
CD CD CD
LO LO LO LO LO LO
5 5 ? CD
LO LO LO LO LO LO LO LO CO CO CO CO CO CO CO CO CO CO CO
CO CO CO CO CO CO CO
o o
u o O 1—
o — O O — — —O 1—O o
CD CD CD CD CD
CJ CD cd CD "o 1— 1— f—
1— f— f- h- 1- 1-
1 h- 1— 1 1 1 h- t— 1
White-tailed Deer (indivs.) 1 1 1 2 2 2 1 1 1 12 Domestic Dog (cont.)

antler 1 2 2 1 6 ulna 1 2 1 4
teeth 1 3 1 5 innominate 1 1
ramus 1 1 1 3 femur 2 2
scapula 1 1 tibia 1 1 1 3
humerus 1 1 2 calcaneum 1 1 2
radius 1 2 1 4 astragalus 1 1
ulna 2 1 1 4 metapodial 8 6 1 15
innominate 1 1 phalanx 1 1
femur 1 1 2 vertebra 2 2 10 8 22
tibia 1 1 2 other 1 2 2 5
astragalus 2 1 3
calcaneum 1 1
Turkey (indivs.) 3 1 1 1 2 8
metapodia 1 1 2 1 1 5
beak 2
phalanx 1 1 2 7 3 2 4 20
humerus 1 1
vertebra 1 1 2 1 1 6
coracoid 1 1 2
rib 2 L 1 4
radius 1 1
other 1 1 1 1 1 2 7
carpometacarpus 1 2 3
femur 1 1
Peccary (indivs.) 1 1 2 1 5
tibia 3 1 2 6
maxilla 1 1 2
tarsometatarsus 1 1 2
mandible 1 1
other 10 3 2 15
teeth 1 1
femur 1 1
tibia 1 1 Jackrabbit (indivs.) 1 1 1 1 4
metapodial 3 3 fragments 2 1 1 1 5
phalanx 1 1 Gopher (indivs.) 1 1 3 3 1 2 11

other 6 2 4 1 13 fragments 1 3 5 6 1 4 20
Opossum (indivs.) 1 3 1 1 1 7
Cottontails (indivs.) 2 1 2 2 6 8 3 2 2 1 9 38 fragments 3 15 4 3 6 31
maxilla 1 3 1 5 Fox (indivs.) 1 1
ramus 2 7 7 3 3 4 26 fragments 1 1
scapula 1 3 2 1 1 1 9 Skunk (indivs.) 1 1
humerus 2 1 1 2 2 2 10 fragments 1 1
radius 1 1 1 3 Rock Squirrel (indivs.) 1 1
ulna 1 1 2 fragments 4 4
innominate 2 1 1 3 5 3 1 2 3 21
femur 2 3 1 4 2 1 2 15 Lizard (indivs.) 3 1 1 1 6
tibia 1 1 2 5 1 1 5 16 fragments 6 1 3 5 15
metapodial 2 1 5 8 Snake (indivs.) 1 1
phalanx 4 4 fragments 1 1
vertebra 1 3 1 5 Turtle (indivs.) 1 L 2
other 1 2 1 3 7 fragments 1 1 2
Dove (indivs.)
Domestic Dog (indivs.) 1 2 2 1 1 1 8 fragments 1 1
ramus 3 1 4 Hawk (indivs.) 1 1
scapula 1 1 fragments 4 4
humerus 2 2 Other Birds (indivs.) 1 1 2
radius 2 2 fragments 1 2 3
A
Table 20. Food Animals from the Ajalpan Plain (Subarea 1) by Site and Phase (Minimum Numbers of Individuals Derived from Identified Fragments)
Palo Venta
Santa Maria Blanco Salada Palo Venta
Ajalpan Ajalpan Santa Maria Blanco Salada
L
o O CD CD C‘
F C' C B‘ B 0 C B A A
XOOLlLluJ^^ Q Q O O O O O CD O CD C CD < < xoou-U-Luiii-, 1 H G F E D* O' C D 1C C* C B C-D
B A B A
S TT ~T -T T -T CO CO CO CO CO OD -rr 3
- -
368, 368, 368, 368, 368, 367, 367, 367, 368, 368,
218-10,
218-6,
218-10, 218-10, 218-10, 218-10,
218-6, 218-10,
368,
(MNNCMMMcocorOfO 368, 204, 204, 218-6,

CD CD CD CD CD co CO CO CO CO C5 <_? CD CD CD CD CD CD CD CO CO CO <_0
368, 368, 204, 218-6,
218-6,
367, 367,
^ !!^
( ( !!^ (
!! !!^
|
Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Ts Tr Tr Tr Tr Tr Tr Tr Tr Tr Tr Tr Ts Ts Ts
Totals
While-tailed Deer (indivs.) 3 5 4 2 2 3 1 1 3 112 11112 1 1 1 1111 1111 [Table 20 cont.]
antler 1 4 1 2 11111 1 1 1 1
Cottontails (indivs.)
maxilla
1 3
3
1 1 1 2 1 1 2 1 1 2 1 1 2 2 2 1 1 1 2 30
3
teeth
111111
2
ramus 11 till
11
2
112
8
ramus, mandible 1
scapula
scapula 15 5 5 2 1 1
humerus
1 1
12
1
humerus 3 3 2 5 1 1 2
radius
radius 3 2 2 1 1
13 12
1111
1 1
11
1
11
innominate
ulna
innominate 3
5
2 3
1 3
2
2 1
1 1 1
2
1 1 1
femur
tibia 1
1
11
11
1
1 1 2
12
10
1 1 9
femur 2 6
5111
2 3
other 13 1 2 1 1 9
tibia
calcaneum 3 111 1
Jackrabbit (indivs.) 1 1112 111 1 1 11
astragclus 3 2 12 1
ramus 1 1
mctapodial 14 2 2 2 2
scapula 2 1 3
1111
1 1 1 1 1
humerus 1 1
2
carpus/tarsus 1 1 1
innominate 1 1 6
phalanx 4 5 2 4 4 1 1 1 1 1 1 2 1
femur 1 1 1 3
vertebra 19 11 12 8 2 2 2 3 2 2 3 1 1 2 1
tibia 1 1 2
rib 1 3 7 2 2 1 1 1 1 1 3
other 3
1111
1 4
other 6 3 2 3 1 1
Peccary (indivs.) till 1 1

Gopher (indivs.) 1 112 112 13
toeth 1 1
ramus 1 2 12 6
ramus, mandible 1 2 1 tooth 1 1 1 3
maxilla 1 humerus 1 1
scapula 1 1 femur 1 1 2
humerus tibia 1 1
ladius other 2 2
ulna Coyote (indivs.) 1 1
femur fragments 3 3
tibia Puma (indivs.) 1 1
metapodial fragments 2 2
phalanx Raccoon (indivs.) 1 1 1 3
other 3
fragments 1 1 1
Domestic Dog (indivs.) 2 2 1 1 1 1 1 1 1 1 2 2 2 2 2 1 1 1 2 1 1 1 1 1 3 1 1 1 1 Skunk (indivs.) 1 1
teeth 1 4 11 5 2 13 3 1 1 fragments 2 2
ramus, mandible 2 1 2 12 1112 1 1 1 1

Turtle (indivs.) till 11 1 7
cervicals 1 1 1
fragments 1 1 13 5 12 1 24
humerus 1 1 1 1
Lizard (indivs.) 1 1
radius 1 1 2
fragments 2
ulna
2
innominate
Raven? (indivs.) 1 1
1 1 2
fragments 2 1 3
lemur
tibia
Ouail (indivs.) 111 1 4
calcaneum
3
1
1
1
fragments 12 1 1 5
1 1
astragalus Hawk (indivs.)

1 1
metapodial
fragments
3 4 4 1 1 1
1 1
phalanx Turkey (indivs.)

2 1 1 2
1 1
other fragments
1
manufactured while the bone was still fresh from the the custom to remove the head of the femur from the
butchering process; one metapodial awl from a very acetabulum of the pelvis in this manner, and cut marks
young deer showed the shaft and unfused articular end are present on several acetabula. Villagers made awls
still connected, as they would have been while the from deer metapodials, and numerous rejected tool
sinew was still holding them together. attempts or “blanks” were found in the debris. The
The lowest five zones of the Ajalpan site (Ts 204) canine teeth of the mountain lion killed at Las Canoas
show an interesting pattern of seasonal occupations. had been removed and polished as ornaments.
Zones F G and H, all of which are living floors, all

1
,
1
,
Palo Blanco Phase

have evidence of dry-season occupation. Lying on each
of these floors were completely hardened antlers, in The two components of this phase showed only the
the state in which they are normally found between remains of dog and cottontail. One unidentifiable frag-
October and February, broken off the skull in such a ment of long bone from Zone A, Test 10, at Quachilco
way that fragments of frontal bone are still clinging to had been slotted, as if by a burin or the edge of a blade.
them. None of the ameiva lizards so common in the Also included in the rubbish were blanks for bone
rainy season are present in these levels. Zone G, which beads. These were long shaft fragments separated into
is a layer of refuse under floor G 1
,
has evidence of a series of shorter tubes by means of deep grooves cut
year-round occupation. The stump of an antler which part-way through with a flint or obsidian blade, but
appears to have been in velvet at the time the deer never finished.
was killed, plus fragments of ameiva lizard, indicate
Venta S ala da Phase
a summer occupation; other fragments of hardened
antler suggest deer which were killed during the dry Venta Salada deposits exposed in Subarea 1 are few,
season. Occupation at the village of Ajalpan was cer- and the scarcity of bone fragments does not allow us to
tainly year-round, though we are not yet sure why liv- draw many conclusions. Most commonly eaten were
ing floors all appear to date to the dry season. dogs (30 percent), cottontails (30 percent), and deer
(20 percent). Traces of raccoon and turkey were also
Santa Maria Phase present.
Twenty-five components of the Santa Maria phase in
Animals Present but not Eaten
Subarea 1 yielded identifiable animal bones; a mini-
mum of 113 individual animals were recovered. Dog, The large numbers of small rodents found in the
deer, and cottontail represented nearly three-quarters Tehuacan caves have already been discussed. For a
of the total, and the rest of the animals were mostly number of reasons, we do not believe that many of
large pocket gophers and jack rabbits. those small rodents —wood rats, cotton rats, kangaroo
When compared with the Ajalpan phase levels in the rats, and others —represent animals eaten by prehistor-
same subarea, the Santa Maria phase shows an increase ic residents of the valley. First of most of them had
all,
in consumption of dog; that animal constituted about clearly been deposited in the caves owl pellets. Still
in
24 percent of the individual animals. Deer had dropped others, especially in the upper levels, were mummified
to about 23 percent, and cottontails had risen to 19 individuals who had obviously crawled inside and
percent. Mud turtles seem to have been less popular, died. None were burned. Most skulls were still intact,
with only one fragment recovered. Rare animals in- a condition which could never have obtained had the
cluded peccary, striped skunk, raccoon, quail, raven, small rodent crania been lying on the surface at a
and hawk. A single mountain lion appeared in Zone D 1
time when men were walking back and forth in the
at Las Canoas. cave. Hence the small rodents have been used mostly
With increased use of domestic animal food, and a as climate indicators (Table 15) and are not included
decrease in hunting, the Santa Maria phase also showed in the food remains.
more wasteful butchering techniques than the pre- This does not mean that small rodents were never
vious phases. Far fewer bones were split longitudinally part of the human diet Tehuacan; in fact, Eric Cal-
at
for marrow, and numbers of deer limb bones were dis- len’s coprolite analysis (Chapter 13) suggests that
carded while still in articulation. Repeated cutting kangaroo rats and other small wild mice occasionally
with flint or obsidian blades is obvious on a number of were eaten by the prehistoric occupants of the Tehua-
deer bones, in areas where tendons had to be cut in can caves. Probably, as indicated by bone and hair
order to free the muscle from the bone. During the fragments in the coprolites, these small rodents were
occupation of Zone G at the site of Coatepec it was eaten whole; hence almost nothing would have re-
169
1500 A D.
Fig. 95. Exploitation of the fauna by inhabitants of the Tehuacan Valley.
mained for the osteologist to identify. The rodents I (Tc 50, II, and Tc 35e, owl pellets between Zones C
examined had either been eaten by owls, or seem to and D); two juvenile cottontails, a juvenile gopher, and
have died a natural death in their burrows in the cave. a small bird from owl pellets (Tc 35e, C-D); one black
Other animals which obviously did not serve as iguana (Tc 272, A); one mummified striped skunk (Tc
sources of food for the human inhabitants include the 35e, A); and the skeletons of two puppies (Tc 35w, 3).
remains of at least four bats of unidentified species Intrusive fragments of modern sheep and goats were
(Tc 50, Zones XXVIII, XV, III, II); one fruit bat (Tc 50, found in the uppermost layers of Purron and San Marcos
VI) and one brown bat (Tc 50, II); two murine opossums caves.
170
Pre-Columbian Hunting Patterns Early Ajuereado Phase ( before 7000 b.c.)
What follows summary of the data from all

is a brief This phase took place when some of the Pleistocene
way hunting patterns in the
subareas, relating to the fauna was still Tehuacan Valley, but
lingering on in the
Tehuacan Valley changed through time. The recon- bone remains clearly show that man was not the “big-
structions are based on the concrete data given above, game hunter” he is often made out to be during this
but we have been quite free in the interpretation of that period. The hunting and/or trapping of rabbits pro-
data. Let us first state our basic premises with regard duced 55 percent of the individual animals found in
to prehistoric hunting technique: this phase, and other small game, such as fox, skunk,
1. We from the large numbers of spear, atlatl,

infer coyote, ground squirrel, turtle, and lizard, were also
and arrow points recovered by the Tehuacan Project important. The numbers of extinct horses and antelope
that large mammals like deer and peccary were ob- are relatively small. Remains of mammoth have been
tained by hunting. found underneath the city of Tehuacan and examined
2. We infer from large numbers of trap fragments by Arturo Romano of the Instituto Nacional de Antro-
(whittled sticks, some with slip knots of maguey fiber pologfa e Historia (MacNeish, personal communica-
attached, etc.) recovered in the same levels that small tion), and it is possible that even they were hunted
animals, especially the primarily nocturnal ones like during this period, though not, of course, carried back
skunk and cottontail, were trapped or snared. This to the caves and rock shelters.
may also apply to birds like quail. Gophers are widely In this respect, our Early Ajuereado faunal remains
trapped in Mexico today by means of deadfalls, and correlated well with Alexander’s data from earlv levels
J
in the Valley of Mexico, Villa (1953: 379) states that at the Levi Rock shelter in Texas (Alexander 1963:
some professional gopher-catchers or “tuzeros” average Table 1), where Lundelius has identified, beside the
ten animals a day. usual big game (bison, deer, antelope, horse, tapir, and
3. We suspect that animals with a sharply delimited an extinct peccary), numerous smaller animals such as
habitat and fairly predictable habits, like the mud tur- skunk, wolf, coyote, bobcat, raccoon, cottontail, jack
tles in the Rio Salado, were virtually “collected.” rabbit, gopher, mole, wood
rat, cotton rat, and others,
4. Small, abundant animals of little nutritive value, as well as fish, snails,and large quantities of mussels
like ameiva lizards, were probably casually collected (ibid. :513). We suspect that big-game “kill sites” dur-
around campsites with no more expenditure of effort ing the Early Ajuereado phase (and such must cer-
than is necessary to throw a rock. tainly exist somewhere in the Tehuacan Valley) prob-
Fig. 95 shows frequency polygons of the percentage ably looked no different from Santa Isabel Iztapan
of the recovered fauna which certain categories of (Aveleyra 1955). The campsites and rock shelters, how-
animals represented during various parts of the Tehua- ever, give a fuller and very different picture of the
can sequence. These frequencies are based on the num- economy of the time.
ber of individual animals present, and not on the num- If Julian Steward’s views on “cultural regularities”
ber of pounds of meat represented by each species. among primitive people in ecologically similar areas
We approve of White’s system (1953), but had we are valid (Steward 1955), we should be able to discern
done our graph on the basis of pounds of usable meat, some parallels between the Early Ajuereado hunters
the deer column in every period would have shoved all and other nonagricultural Indian groups sharing an
others to one side of the page, obscuring some of the antelope-jack rabbit desert environment. Perhaps the
subtler changes in small game trapping we wished to most applicable example given by Steward himself is
bring out. Our chart is not of meat, but of human the Great Rasin Shoshone, whom he knew particularly
activities, since we feel that the trapping of ten or fif- well. They, too, lived in a region of intermontane steppes
teen rabbits and
skunk or two is an expenditure of
a and desert with low xerophytic scrub where “the two
energy worth bringing to light, even if the number of species of greatest importance to the Indians were
pounds of meat involved does not total as much as antelope and rabbits the Shoshoneans probably
. . .
the meat from a single deer. Fig. 95, therefore, should ate more rats, mice, gophers snakes, and lizards
. . .
be taken as an indication of the different ways in which than large game” (ibid.: 104). This correlates well with
pre-Columbian man was spending his time in pursuit our impressions of Early Ajuereado subsistence, and
of meat during a given phase in the prehistory of the suggests that even the wood rats (
Neotoma and deer
)
Tehuacan Valley. mice (

Peromyscus ) found in Zones XXV-XXVIII of
171
Coxcatlan Cave may have been eaten (see Table 15). fox (U. cinereoargenteus ); when Pleistocene quail dis-
When game was sufficiently abundant, the Shoshone appeared, they began eating bobwhites; when the large
held collective hunts, of which the most common was Early Ajuereado turtles vanished, they began eating
the rabbit drive. Men, women, and children beat the mud turtles; rodents like Cratogeomijs probably re-
brush over a wide area, either driving the rabbits placed the extinct small ground squirrel; and the stream
ahead of them into nets or enclosures or else clubbing of Audubon cottontails continued uninterrupted. The
and shooting them along the way (ibid.: 109). Frankly, vanishing of the “megafauna” in the Tehuacan Valley
without attempting to claim that similar specific tech- led to a shift from larger to smaller hunting bands,
niques were used, I find it hard to explain the dispro- but in no sense triggered a change from a so-called
portionately large numbers of rabbits in Early Ajue- “paleo-Indian” to an “archaic” way of life.
reado levels — forty in Zone XXVI alone without —

recourse to communal drives of this kind. Communal El Riego Phase (6500-4900 b.c.)
antelope drives were also held by the Shoshone, but
During the rainy season, groups of macrobands came
less frequently; they depended to a certain extent on
together in Subarea 3 to exploit the abundant resources
the fact that antelope, as we have noted, are not par-
of that particular time of year. Their activities included
ticularly adept at staying away from man and, in fact,
the hunting of deer and peccary and the trapping of
“are drawn by toward strange objects”
curiosity
cottontail, fox, skunk, bobcat, and other small game.
(ibid. 111).
: We are probably not being too incautious
They also ate lizard and quail. They made short-term
in seeing parallels to both these activities in the Early
hunting forays into Subarea 4, where they killed a
Ajuereado phase.
few deer and brought them to Purron and Abejas
7000-6500 eaves; they trapped a few rabbits, ate a few birds and
Late Ajuereado Phase (
b.c.)
lizards, and then probably moved on, following the
During this period, trips were made throughout both deer. They made brief trips into Tecorral Canyon
the rainy and dry seasons to the Coxcatlan thorn forest
where they picked up rabbits, gophers, and mud tur-
(Subarea 3) for deer hunting, trapping of cottontail,
tles. They did some fishing in the Rio Salado, and
skunk, and probably other small animals, and the snar- probably collected a few more mud turtles there.
ing of quail, other small birds, and lizards. Deer hunt-
With the coming of the dry season, the pattern
ing and rabbit trapping also took place on the edge
changed. Most of the deciduous vegetation had turned
of the valley floor near the El Riego springs.
brown or lost its leaves, and the male deer had fallen
In spite of the presence during this period of a com-
in with the females, making intensive deer-hunting
pletely modern fauna, stone artifacts are identical to
profitable. Small hunting bands camped repeatedly
those of the Early Ajuereado phase (MacNeish, per-
in Subarea 3, stalking deer through the fall and winter
sonal communication). Material culture, therefore,
and occasionally snaring quail, doves, and cottontails.
seems to have lagged behind any “readaptation to the
On the valley floor, probably throughout the year, they
modern fauna. The changes we see are mostly be-
hunted deer and cottontail. The general characteristics
havorial: with the vanishing of the antelope and Pleis-
of the El Riego phase include an increase in deer hunt-
tocene jack rabbit, there is no longer even a suggestion ing over previous periods and a bone-smashing tech-
of communal game drives. Smaller groups seem to have
nique which suggests extensive use of bone marrow.
done the hunting. Steward had already observed that,
Both these activities suggest a fuller and more efficient
in arid North America at least, deer were not usually
adaptation to the modern fauna of the valley than we
hunted by the same techniques that worked best on
see in Late Ajuereado times.
antelope. The latter were most efficiently hunted by
surrounds and large cooperative drives, while deer
Coxcatlan Phase (4900—3500 b.c.)
(like nonmigratory game occurring in small and scat-
tered groups) are better hunted by “small groups of During the rainy season, groups of people camped
men who know their territory well’’ (Steward 1955: in Tecorral Canyon, where they trapped cottontails
38-41). The pattern in Late Ajuereado times would and gophers and collected turtles and lizards. They
seem to substantiate this observation. did little or no hunting, perhaps since the primary
Aside from this, we doubt that much “readaptation” purpose of these camps was the planting and harvest-
was necessary game.
as regards the trapping of small ing of the corn found in San Marcos Cave in levels
As Fig. 95 shows, when Pleistocene fox vanished, the dating to this phase. Late in the rainy season, they
prehistoric residents of the valley began trapping gray made short trips to Subarea 3 to hunt deer and trap
172
cottontails and other small game. They collected and of the phase include an intensification of deer hunting
ate a few lizards at this time. in the dry season.
During the dry season, hunting bands stayed for
Purron Phase ( 2300-1500 b.c.)
long periods of time in the Coxcatlan thorn-forest area.
These bands practiced intensive hunting of deer, oc- So little evidence is available from this phase that
casional hunting of peccary, and some minor trapping no reconstruction of the hunting pattern is possible.
of birds, cottontails, and other small game. They made We know only that there were brief camps in Subarea
wide-ranging forays into Subarea 4, primarily to hunt 4 where rabbits, skunks, and birds were eaten.
deer, which often seem to have been brought from a
Ajalpan Phase ( 1500-900 b.c.)
great distance to Purron and Abejas caves. While
briefly camped in this area, the group did some trap- By the Ajalpan phase, year-round residence in per-
ping of cottontails and birds. They fished occasionally manent villages on the valley floor was the rule, al-
in the Rio Salado. On the valley floor (probably though bands of hunters and collectors appear to have
throughout the year), they trapped many cottontails been sent out seasonally to gather wild fruits and an-
and jack rabbits, did some deer hunting, and ate doves imals as a supplement to agriculture. In contrast to the
and other birds. The general characteristics of the Cox- earlier phases, foraging bands took some of their food
catlan phase included a shift away from long rainy- along with them on these trips, including, judging by
season camps in the mountains as agriculture began to the cave remains, dogs and probably also corn and
affect food production, and an abandonment of the squash.
typical El Riego butchering technique which involved During the rainy season, deer and peccary were
the meticulous smashing of deer long bones for mar- hunted in the vicinity of the village, where cottontails
row. Intensive deer hunting in the dry season con- were trapped and turtles and lizards were collected.
tinued. Many village dogs were eaten. In the dry season, hunt-
ing of deer and peccary was intensified, many turtles
Abejas Phase ( 3500-2300 b.c.) were collected, and more dogs were eaten. A few birds
Many continuations of the Coxcatlan hunting pattern and small mammals also seem to have been trapped.
are seen in the Abejas phase. During the rainy season, In addition, small foraging parties passed through
camps were made in Tecorral Canyon (and probably Tecorral Canyon, picking up an occasional cottontail
along most wet arroyos where planting could be done) on the way. Brief camps were made in Subarea 4 by
and meat was procured by some trapping of cottontail Ajalpan villagers, who ate the dogs they had brought
and skunk, collecting of turtle and lizard, and snaring with them, while hunting deer and trapping a few cot-
of doves and other birds. It may be that man was tontails. General characteristics of the Ajalpan period
spending a greater part of the rainy season on the include considerable seasonal deer hunting, increased
valley floor and adjoining wet canyons now that agri- use of one domestic meat source, and in villages near
culture was more fully developed. Fishing was done the Rio Salado, a steady collecting of mud turtles.
in the Rio Salado, and a few brief camps (probably

Santa Maria Phase ( 900-200 b.c.)
late in the rainy season) were made in Subarea 3 for
the hunting of deer and trapping of cottontails, foxes, At the beginning of the Santa Maria phase consider-
and probably other small game. able deer hunting was being done in the vicinity
still
Once again, during the dry season, long-term of the village; peccary were also taken. This activity
macroband camps were made in the Coxcatlan area, dwindled toward the end of the phase, and by 200 b.c.
with intensive hunting of deer and some peccary. The hunting was a relatively minor activity which was
trapping of cottontail, raccoon, skunk, and birds was never to regain the importance of preceramic and early
carried on as a minor may be that the brief
activity. It Formative times. Dogs were eaten on the largest scale
trips evidenced Arroyo Lencho Diego (Subarea
in the ever in the villages of this phase. There was renewed
4), during which deer and cottontail were obtained, interest in the trapping of cottontail, jack rabbit,
can be traced to the dry season also; but for this we gopher, and quail in the vicinity of the village, sup-
have no concrete evidence. The significant develop- plemented by turtle-collecting and the occasional tak-
ment in the Abejas phase is the first appearance in the ing of small game like raccoon, skunk, raven, hawk,
Tehuacan Valley of domestic dog, now probably added and other forms.
to the food supply and certainly recorded from two During all seasons of the year, large parties traveled
cave sites in different subareas. General characteristics to the thorn forest around Coxcatlan. Remains in the
173
caves suggest that these were probably mainly excur- bers of people, assembled for all or most of a season,
sions to collect wild seasonal fruits, but the groups in- with stores of domestic meats and vegetables to help
volved also trapped cottontail, hunted deer and pec- support them during this period.
cary, and captured skunks, iguanas, smaller lizards,
and birds.During the dry season, other groups camped Venta Salada Phase (a.d. 700 to the Conquest)
in the vicinity of the El Riego springs, trapped opossum Only three small levels from Postclassic villages on
and cottontail, hunted deer, and ate the dogs they had the valley floor were excavated by the Tehuacan
brought with them. During the construction of the dam Project. They show that trapping of cottontails and
near Purron Cave, large groups seem to have camped other small game, as well as some hunting of deer,
in Subarea 4, where they carried on extensive trapping occurred in the vicinity of the village, while the
of cottontails, hunted deer, and also picked up gophers, eating of domestic turkeys and dogs continued. Con-
iguanas, quail, and other birds. sumption of turkey, in fact, seems to have increased
during this phase, probably reaching its peak just at
Palo Blanco Phase ( 200 b.c. to a.d. 700)
the time of the Conquest.
No villages or towns of this phase have been exca- Many campsites from this period represent groups
vated, so we have no evidence of the limiting pattern sent out from the villages, towns, or cities. As in the
in the vicinity of the towns themselves. There are many Palo Blanco phase, few of these can be definitely traced
campsites, however, reflecting the greatly increased to the dry season by means of animal remains, but
population of the valley at this time. Another domestic rainy-season collecting-and-trapping camps were very
meat source, the turkey, was introduced into the Te- big indeed. Foragers in large camps on the edge of the
liuacan Valley during the middle of the Palo Blanco valley floor, near El Riego springs, intensively trapped
phase ( ca . a.d. 180). cottontails and hunted some deer and peccary through-
During both rainy and dry seasons, there seem to out the year. They also ate turkeys and dogs in quan-
have been very large camps at El Riego, near the edge tity. During the course of their stay, it appears that
of the valley floor, where many cottontails were their trap lines also picked up opossums, skunks, and
trapped, as well as gophers, jack rabbits, opossums, jack rabbits; and one way or another they obtained
foxes, lizards, snakes, turtles,and other small game. gophers, rock squirrels, turtles, birds, and small lizards.
These groups, perhaps hunting with the bow and Late in the rainy season, sizable camps were made in
arrow now, took many deer and peccary, and ate num- Subarea 3, where white-tailed deer were hunted and
bers of dogs they had brought with them. Sizable cottontails, foxes, skunks, opossums, bobcats, and
camps were made during the rainy season in the wet gophers were trapped. The campers ate many of the
canyons in Subarea 2 by foragers who ate cottontail, dogs and turkeys brought with them, and rounded out
fox, gopher, lizard, snakes, and birds; hunted a few their meat diet with iguanas and other lizards, and many
deer and peccary; and cooked some of their dogs. different kinds of birds.
Campers during the rainy season in Subarea 4 trapped Brief camps, for which we have no seasonal evidence,
cottontail and hunted deer, ate a few dogs, and picked were made in Tecorral Canyon, where cottontails,
up an occasional fox, skunk, bird, iguana, or smaller gophers, and skunks were trapped, peccary hunted, and
lizard. Finally, late in the rainy season, big groups who small birds eaten. The general characteristics of the
may have been mainly plant collectors traveled to the phase include a tendency for big rainy-season collect-
Coxcatlan thorn where they trapped cottontail,
forest, ing-and-trapping camps and a slight increase in the
hunted deer and peccary, and ate many of the dogs eating of turkey and wild birds when compared with
and turkeys brought with them. Perhaps as a sideline the earlier phases.
to their other collecting activities, they caught a few
foxes, skunks, gophers, birds, iguanas, and smaller Conclusions
lizards. 1. Man was present in the Tehuacan Valley 10,000
Our general impression of the Palo Blanco phase is years ago as a contemporary of horse, antelope, jack
that it was a time of plenty, with a rich and varied rabbit, fox, small ground squirrel, gopher tortoise, and
diet composed both of domestic and of wild foods. other forms suggesting a climate like that of the arid-
An increased use of domestic meat sources made big- temperate plateau of northern Mexico today. During
game hunting of even less importance than it had this era he lived primarily on small game, though re-
been in the Formative phases. The rainy-season col- mains of large fauna are also present in his campsites.
lecting and trapping camps seem to involve large num- There is evidence for communal jack rabbit drives.
174
2. The change to a modem climate and fauna was lected: June 21, 1962, near Calipan. Contained: (1) Plant
essentially complete by 7000 b.c. Since that time, as material — great many corn seeds; mesquite ( Prosopis juli-
a
small rodent re- flora) seeds and pods; a number of tree-legume leaves, al-
suggested by the lack of variation in
most certainly those of mesquite; a single large solanaceous
mains, the basic ecology of the valley has not altered.
seed, possibly Capsicum other small solanaceous seeds;
Today the northern end of the valley is classified as
the base of a dicotyledon leaf, not mesquite. (2) Animal
semi-arid temperate, and the southern end of the val-
ley as arid tropical.
material —
among various fragmentary insect remains, parts
of a beetle wing cover and abdomen could be recognized.
3. During the period from 6500 to 1000 b.c., sub-
(3) Mineral material — nil.
sistence involved considerable hunting of white-tailed

deer, especially in the dry season, and cottontails of Order Lagomorpha
two species were trapped in great numbers. These two
Family Leporidae
animals, deer and cottontail, have remained the most
Sylvilagus aadubonii (Audubon cottontail), Specimen
important wild meat sources in the valley.
T-23. Collected: June 27, 1962, near Calipan. Contained:
4. The domestic dog was introduced into the Te-
huacan Valley at about 3200 b.c., in the agricultural

( 1 ) Plant material —
fruits of Castela tortuosa and the epi-
dermis of what is probably either the fruit or leaf thereof;

late preceramic Abejas phase. To the best of our knowl- several tvpes of trichomes, including compound stellate
edge, this is the oldest reliably dated evidence for dog ones (see Mexican cottontail, Specimen T-22); chewed
in Mesoamerica proper. From at least 1500 b.c. on, grass and probablv corn leaves. (2) Animal material nil. —
there is abundant evidence that this animal was eaten. (3) Mineral material nil. —
5. The domestic turkey was introduced into the Te- Sylvilagus cunicularius (Mexican cottontail), Specimens
huacan Valley at about a.d. 180, in the Classic, or Palo T-14 and T-15. Collected: June 21, 1962, near Calipan.
Blanco, period; from this time on, it was used as food. —

Contained: (1) Plant materia] the remains of a pod and
To the best of our knowledge, this a number of seeds of a tree legume, not mesquite; the rest
is the oldest reliably
was a mixture of finelv chewed grass and legume pod.
dated evidence for domestic turkey Mesoamerica
proper.
in
(2) Animal material —
nil. (3) Mineral material — nil.
Sylvilagus cunicularius (Mexican cottontail, juv. ), Speci-

6. Hunting continued in the later periods (900 b.c.
men T-22. Collected: June 27, 1962, near Calipan. Con-
but never on the scale seen in prepottery
to a.d. 1500),
tained: (1) Plant material —
well-chewed grass and corn
times.Trapping of such forms as cottontail and gopher leaves, compound stellate trichomes, and a fragment of
continued, but at the time of the Conquest domestic dicotyledon leaf. (2) Animal material —mammal hair. (3)
animals represented at least 20 percent of the individ- Mineral material— nil.
ual animals eaten in the valley.
Order Rodentia
A Note onthe Stomach Contents of Animals Family Sciuridae

Collected in the Tehuacan Valley, 1962 Citellus ( Spermophilus ) rnexicanus (gray rock squirrel),
By Eric O. Callen Specimen T-20. Collected: June 26, 1962, from El Riego
The material analyzed below was obtained from small

cliffs. Contained: (1) Plant material —
three large seeds and
some fruit epidermis, neither of which have vet been iden-
animals collected by Kent V. Flannery between June and
tified; Opuntia seeds and tissue; what appear to be mes-
August 1962 in and around the caves and archaeological
quite seeds; and seed coat and part of the pod of a tree
Tehuacan Stomach contents
sites of the Valiev.
animals collected were not available for analysis.
of the larger
The ma-
—
legume. (2) Animal material nil. (3) Mineral material
nil.
terial was first washed in water and then preserved in
methylated spirit, stopping further digestion or decay. Any Family Heteromyidae
material requiring microscopic examination for identifica-
Liomys irroratus (spiny mouse), Specimens T-31 and
tion purposes was mounted in glycerine jellv on a standard- T-32. Collected: July 8, 1962, from the stomach of a bob-
sized microscope slide.
as a preservative. It is
The jelly
hoped that more

contained a little phenol
definite identifica-
cat, freshlv eaten. Contained: (1) Plant material one seed —
of mesquite, two seeds of Vallesia glabra, and one seed of
tion of much of the material can be made at a later date.
—
Opuntia. (2) Animal material nil. (3) Mineral material
nil.
Mammals
Family Cricetidae
Order Marsupialia
Peromyscus melanophrys (deer mouse), Specimen T-41.
Family Didelphiidae Collected: August 2, 1962, near Coxcatlan Cave. Contained:
Didelphis marsupialis (opossum), Specimen T-16. Col- (1) Plant material —apparently nil. (2) Animal remains
175
well-chewed remains of many arthropods, showing dark Ctenosaura pectinata (black iguana), T-7. Collected:
brown and magenta color in or under the chitin. (3) Min- June 20, 1962, from Rio Salado area near Calipan. Con-
eral material —
some hardened matter that might have been tained: (1) Plant material —
the remains of a number of
wax comb. Vallesia glabra fruits, seeds, flowers and buds, as well as
inflorescence stalks. There were also three large, fat, almost
Order Carnivora beanlike seeds, one of which had apparently come from a
large, tough drupaceous fruit, densely covered with stellate
Family Procyonidae
hairs. These seeds and the fruit contained a curiously lumi-
Bassariscus astutus (ring-tailed
Collected: June 30, 1962, near Calipan. Contained:
cat). Specimen T-26.
(1)
nescent orange material. (2) Animal material three soil —
nematodes, a robber fly, probably of the genus Bombomina,
Plant material —exclusively seeds of mesquite and fruits
and 30 small spiders. (3) Mineral material nil. —
and seeds (2) Animal material
of Vallesia glabra. three —
crickets (one male, two female), several small beetles, one
Birds
very large beetle, and one arachnid (
probably a scorpion )
(3) Mineral material nil. — Family Piiasianidae
Procyon lot or (raccoon), Specimen T-27. Collected: July
Colinus virginianus (bobwhite quail), Specimen T-2.
2, 1962, near Calipan. Stomach empty; intestine contained:
Collected: June 18, 1962, near Ajalpan. Contained: (1)
Opuntia seeds, epidermis, trichomes,
(1) Plant material
and tissue; a piece of plant epidermis with closely packed,
Plant materia] —
all one kind of seed, with the seeds grouped
together into groups of three or four, reminiscent of sugar

sunken glands; a few stomata; and a single three-cornered
beet; but the presence of flower parts rather suggests the
seed (Family Polygonaceae) (2) Animal material mam-
.
— mint family. (2) Animal material —a single small spider.
mal hair and a single insect abdomen, badly damaged. (3)
—
mineral material a few small stones, i.e., gravel.
(3) Mineral material nil. —
Family Mustelidae Family Columbidae
Mephitis macroura (striped skunk), Specimen T-17. Col- Zenaida asiatica (white-winged dove), Specimens T-9
lected: June 22, 1962, near Calipan. Contained: (1) Plant and T— 10. Collected: June 20, 1962, near Venta Salada.
material — (2) Animal material

nil. —
wings and parts of a
—
Contained: (1) Plant material fruits and seeds of Vallesia
large beetle, parts of a pseudoscorpion, and many larvae,
glabra; five seeds of what at first looked like corn, but may
probably belonging to a beetle of the Family Scarabaeidae. be some type of bean; one seed of Riccinus, the castor oil
These larvae are probablv those of the white grub ( Phi/llo plant; five seeds of some large drupeacous fruit; pieces of
thopa ), one of the June beetles. Among the grubs were also another fruit as yet unidentified. (2) Animal material nil. —
numerous nematodes of the genus Physaloptera, sensuo (3) Mineral material —
nil.
lato. In addition, the wings of several small beetles were Zenaida asiatica (white-winged dove), Specimen T-24.
found. (3) Mineral material — nil.
Collected: June 30, 1962, from the Coatepec site near
Ajalpan. Contained: (1) Plant material —
what appear to be
Reptiles the remains of “tuna” (cactus fruit), plus fragments of a
hard, brown seed coat. The seeds are both the young and
Order Chelonia mature seeds of a single kind of cactus, probably either
Family Kinosternidae Pilocereus or Cephalocereus. (2) Animal material — nil.
Kinosternon integrum (mud turtle), Specimen T—21.

(3) Mineral material — nil.
Zenaida asiatica (white-winged dove), Specimen T-36.

Collected: June 27, 1962, from Rio Salado. Contained: (1)
Collected: July 25, 1962, near Coxcatlan Cave. The crop
Plant material nil. —
(2) Animal material a single large — contained: (1) Plant material —
a single seed of Opuntia.
water beetle, probably of the genus Belostoma, as well as
the wings and body of an ichneumon fly. Among some mud
(2) Animal material —
some 20 cutworms, half of them
pale all over and the rest with dark dorsal and pale ventral
were the remains of medium and small-sized water-snail
shells. Two termite wings were also present. (3) Mineral
surfaces. Mineral material — nil.
Zenaida (white-winged dove), Specimen T-38.

asiastica
material —mud. Collected: July 28, 1962, near Coxcatlan Cave. The crop
Order Squamatq, —
contained: (1) Plant material- nine largish fruits, not yet
identified, and twelve seeds of Opuntia. (2) Animal material
Family Iguanidae — nil. (3) Mineral material — nil.
Ctenosaura pectinata (black iguana) Specimen T-5. Col-

,
lected: June 18, 1962, near Ajalpan. Contained:

Family Caprimulgidae
(1) Plant
material —four large, reddish-brown fruits, and the tough Caprimulgus ridgwayi (buff-collared nightjar), Specimen
exterior of a drupaceous fruit, neither of which has been T-40. Collected: July 30, 1962, near Coxcatlan Cave. Con-
identified. —
(2) Animal material- nil. (3) Mineral material tained: (1) Plant material nil. —
(2) Animal material
— nil. remains of insects, many of them heads, of which a num-
176
ber were undoubtedly those of ants; also large insect abdo- (1) Plant material — many corn seeds and a good
a great
men which cannot be identified and the headless body of a deal of corn silk. (2) —
Animal material the remains of one
long, thin-waisted insect. (3) Mineral material — nil. cricket, one grasshopper, an ichneumon fly, numerous ants,
and a carnivorous beetle. There were also bones of a small
Family Corvidae
lizard, probably a race-runner (Family Teiidae). (3)
Corvus corax (raven), Specimen T— 12. Collected: June Mineral material — nil.
20, 1962, from the Coatepec site near Ajalpan. Contained:
REFERENCES
Alexander, 1963. Miranda, 1948.
Alvarez del Toro, 1952. Schmidt and Inger, 1957.
Aveleyra, 1955. Schmidt-Nielsen, 1964.
Ditmars, 1940. Smith, C. E„ 1965 a, 1965b.
Hall and Kelson, 1959. Smith, H. M., 1946.
Hester, 1960. Steward, J. H., 1955.
Leopold, 1959. Taylor, 1956.
MacNeish, 1962. Villa R., 1953, 1954.
Martin del Campo, 1944. White, 1953.
Miller and Kellogg, 1955.
177
CHAPTER 9
Prehistoric Wild and Cultivated Maize
Paul C. Mangelsdorf, Richard S. MacNeish,

Walton C. Galinat
rehistoric remains of maize were found in five of ited this valley a number of times beginning in 1941,
P
Riego.
the caves excavated in
catlan,
From
Tehuacan Valley: Cox-
Purron, San Marcos, Tecorral, and El
several standpoints the prehistoric maize
and at no time had it occurred to him that wild corn
might once have grown there. Closer examination of
the environmental factors, however, suggests that the
from these eaves is the most interesting and significant habitat furnished by this arid valley rnay, in fact, have
so far discovered. (1) It includes the oldest well-pre- been almost ideal for wild corn. The average annual
served cobs available for botanical analysis. (2) The rainfall in the valley is low (approximately 500 mm.
oldest cobs are almost certainly those of wild maize. per year at the valley’s center). More important than
(3) This maize appears to be the progenitor of two of the total rainfall, however, is its seasonal distribution.
the previously recognized ancient indigenous races of About 90 percent of the annual rain falls during the
Mexico, Chapalote and Nal-Tel, of which prehistoric growing season from April through October. It reaches
prototypes had previously been found in Swallow and its peak in midsummer, during corn’s most critical
La Perra caves respectively (Mangelsdorf et al. 1956; period when it would normally be silking, shedding
Mangelsdorf and Lister 1956). (4) Specimens of all pollen, and the young kernels developing.
parts of the plant were preserved, and these show —
The remaining months are quite dry in midwinter
beyond a reasonable doubt that the ancestor of culti- —
the valley is virtually a desert and comprise a period
vated maize was maize. (5) The collection portrays during which the seeds of wild maize and other annual
a well-defined evolutionary sequence covering a pe- plants could have lain dormant ready to sprout with
riod of about 6500 years. the beginning of the summer rains —never in danger
The finding of prehistoric wild corn in the envi- of germinating prematurely only to succumb to the
ronmental setting of the Tehuacan Valley is in some vicissitudes of winter. Seeds of many wild plants have
respects quite surprising. Maize is not notable for its mechanisms for delaying germination, devices which
drought resistance, and in order to thrive it requires provide protection against this danger, but corn — at
a steady and adequate supply of water. For this rea- least modem com —has no such characteristic, perhaps
son, among others, Mangelsdorf and Reeves (1939) because it evolved in an environment where none was
postulated that wild maize, if still extant, should be needed. Such is the environment of Tehuacan. The
found in the humid regions of the tropics and sub- perennial vegetation of this valley, which year after
tropics. At first glance the Tehuacan Valley, with its year must survive the dry winter months, is necessarily
arid climate and
predominantly xerophytic, drought-
its xerophytic. The annual vegetation, which grows only
resisting vegetation comprising cacti and thorny during the rainy season, need not be especially drought
leguminous shrubs, did not appear to be a suitable resistant. Nor need the wild corn have been subjected
habitat for wild corn. Actually Mangelsdorf had vis- to severe competition with the drought-resistant peren-
178
PREHISTORIC WILD AND CULTIVATED MAIZE
the collections from the other caves adding little more

than corroboration. (An additional 750 specimens came
from Tc 272, Tc 255, and Tc 35e.) Consequently we
propose to describe the specimens from San Marcos
Cave in some detail and those from the remaining caves
much more briefly.
San Marcos Cave, Zones F and E

From Zone F, representing the early part of the Cox-
catlan phaseand dating to nearly 5000 b.c., there is one
small cob. Since this specimen is essentially like the
Fig. 96. Cobs of wild corn (Coxcatlan phase, San Marcos cobs from Zone E, it probably represents only a very
Cave, actual size) are characterized by uniformity of intact
slightly earlier part of the Coxcatlan phase than the
cobs, relatively long glumes, and fragile rachises.
latter zone. The following description of the cobs from
Zone E applies also to this single specimen.
nial vegetation. The sites which are most suitable for The maize from Zone E consists of 26 cobs or frag-
corn, the alluvial terraces and fans of the canyons ments of cobs. A number of these are illustrated in Fig.
and barrancas, seem not to be well adapted to the 96. These cobs are remarkably uniform not only in size
growth of cacti and shrubs. Indeed the deeper the allu- but also in their botanical characteristics. Of eight
vial soil the less likely are the cacti and shrubs to be cobs which are apparently intact with respect to length,
found growing on it. Instead there are grasses and other the shortest is 19 mm. and the longest 25 mm. The
annual plants among which wild corn might have been number of rows of kernels is usually eight, but one
quite at home. cob is distichous with four rows and another has eight
In all, 24,186 specimens of maize were found in the rows in the basal and four in the apical region. The
five caves; 12,860 of these, or more than
half, are whole number of functional spikelets per row varies from
or almost intact cobs. There are, in addition to the six to nine, and the total number of spikelets (esti-
intact cobs, 3,941 identified cob fragments and 3,878 mated by multiplving row number by number of spike-
unidentified cob fragments. Among the remaining lets) varies from 36 to 72. The average number is 55.
specimens are all parts of the corn plant: 46 roots, 506 The glumes of the spikelets are rather long in rela-
pieces of stalks, 442 leaf sheaths, 282 leaves, 245 inner tion to other structures and are soft and herbaceous
husks, 706 outer husks, 12 prophylls, 127 shanks, 384 rather than horny and indurated. In some cobs the
tassel fragments, 47 husk systems, 5 midribs, and 797 glumes are folded back, probably as the result of the
kernels. There are also numerous quids, representing removal of the kernels. The general aspect of these
83 chewed stalks or leaves and 140 chewed husks. early cobs is that of a weak form of pod corn similar
Botanists studying a collection of prehistoric speci- to that of a genotype of modern pod corn involving
mens of maize are usually confronted with the problem the allele at the tunicate locus on chromosome 4
till’
1
of where best to begin. In undertaking our study of combined with the tunicate inhibiting factor, Ti, on
the prehistoric maize from Tehuacan, we decided to chromosome 6. When examined under a dissecting
concentrate first on the relatively limited collection scope these cobs produce an impression of pod com
from San
Marcos Cave, totaling 1,248 specimens. from which the kernels have been removed. A photo-
Since the number of specimens was not large, it was
possible to spread out the entire collection and view
it as a whole. These specimens revealed a well-defined
sequence showing how maize had evolved over a

—
period of some 6500 years how a wild grass with
tiny ears had developed into a highly productive culti-
vated food plant. We hoped, too, that the study of this
relatively small collection would provide a pattern
for studying the much larger collection of over 15,000
specimens from Coxcatlan Cave. This not only proved Fig. 97. An intact cob from Fig. 96 magnified to show the
to be the case, but it also turned out that the collection long soft glumes and the general impression of pod corn
from San Marcos Cave tells virtually the whole story, (x 3.0).
179
graph of one of these cobs taken at low magnification

is reproduced in Fig. 97.
The spikelets, which are uniformly paired as in

modern corn, are attached to cupules which are as long it
as they are broad or longer, fleshly rather than indur-
ated, and almost glabrous, bearing only sparse short
hairs.
The rachis consists of individual cupules joined to-
gether at their sides and ends. The cross section of the
. /
rachis of an eight-rowed ear thus has the appearance of

a square with flaring comers. The rims of the cupules 5
form the structures which in modern maize are some-
times referred to as “rachis flaps.” In the earliest San
Marcos maize these are quite prominent. The adjoin-
ing cupules are not tightly bound to one another so
that the rachis disarticulates rather easily and is by
no means as rigid as the rachises of modern maize.
The majority of the ears represented bv these cobs
were originally bisexual, bearing pistillate spikelets
below and staminate spikelets above. Of the fifteen
apparently intact cobs, ten had recognizable stumps at
their tips where a staminate spike had presumably
been broken off. Others may also have had staminate Fig. 98. Cobs of earlv cultivated corn (Abejas phase, San
spikes, since it is not always possible to distinguish Marcos Cave, actual size) are larger and more variable than
between the stump of a staminate spike and the broken cobs of wild corn but are similar in having long glumes and
fragile rachises.
end of a cob which has lost one or more of its terminal
pistillate spikelets.
In bearing both male and female spikelets in the

same inflorescence, the prehistoric wild corn resembles firm evidence from other plant species that agriculture
our genetically reconstructed ancestral form (Mangels- had vet become well established in this valley, at least
dorf 1958a) as well as certain primitive races of mod- in the earlier part of the Coxcatlan phase. (6) The
ern corn, Nal-Tel and Chapalote of Mexico (Well- predominating maize from the following phase, Abejas,
hausen et al. 1952), Polio of Colombia (Roberts et al. in which agriculture definitely was well established,
1957), Confite Morocho of Peru (Grobman et al. 1962), is and more variable than the earliest corn.
larger
and corn’s wild relative, Tripsacum which regularly
,
This combination of circumstances leads to the con-
bears pistillate spikelets below and staminate spike- clusion — an almost inescapable one — that the earliest
lets above on its lateral spikes. prehistoric corn from the Tehuacan caves is wild corn.
The earliest cobs —27 from San Marcos Cave and We have assumed here that it is.
44 from Coxcatlan Cave, some of which date to nearly These cobs represent the first ears of wild maize
5000 b.c. are regarded as being those of wild maize so far discovered. However, still earlier evidence of
for six reasons. (1) They are remarkably uniform in the existence of maize in the form of fossil pollen has
size and other characteristics and in this respect resem- been found in the Valley of Mexico (see Barghoorn et
ble most wild species. (2) The cobs have fragile rachises al. 1954, and Irwin and Barghoorn 1965).
as do many wild grasses; these provide a means of The separation of the staminate and pistillate spike-
dispersal which modern corn lacks. (3) The glumes are lets, the regular pairing of the pistillate spikelets, and
relatively long in relation to other structures and must the relatively soft tissue of the rachis and the glumes
have partially enclosed the kernels as they do in other all provide convincing evidence that the wild an-
wild grasses. (4) There are sites in the valley, such as cestor of cultivated corn was com and not one of its
the alluvial terraces below San Marcos Cave, which relatives, teosinte or Tripsacum. Together with the
are well adapted to the growth of annual grasses, in- much earlier fossil pollen from the Valley of Mexico,
cluding corn, and which the competing cacti and these prehistoric cobs from the Tehuacan Valley pro-
leguminous shrubs appear to shun. (5) There is no vide virtually conclusive evidence that com is an Amer-
180
Fig. 99. Cob of early cultivated corn (Abejas phase, San

Marcos Cave; x 1.6). The slender rachis and relatively long
glumes are characteristics of weak forms of pod corn.
ican plant and not of Asian origin as some writers have

suggested (Stonor and Anderson 1949).
San Marcos Cave, Zone D Fig. 100. Cobs of early cultivated corn (Abejas phase, San
Marcos Cave; x 1.8). Left: a rachis of an eight-rowed cob
Cobs with floral bracts removed showing the long relatively shal-
From zone representing the earlier part of the
this low cupules and lack of compression. Right: intact cob of
Abejas phase, which extended from about 3500 to 2300 a similar ear.
b.c., 102 cobs or fragments of cobs were recovered. All

but one of these, described below, were similar in their
general characteristics to the 27 specimens from Zones In the structure and texture of the glumes and cu-
E and F, but the majority were larger. The character- pules, these cobs, altogether 145 in number, are quite
istics of ten totally intact cobs from this zone are set similar to those in the earlier Zones F and E. Their
forth in the accompanying tabulation. A number of the characteristics are illustrated in Figs. 99 and 100. Their
cobs are illustrated in Fig. 98. larger size indicates, however, that they were grown in
a better environment. Since these larger cobs came
Length No. of Spikelets Total from levels in which were found remains of two species
in mm. rows per row spikelets of squashes ( Cucurbita moschata and C. mixta), tepary
28 8 8 64 and common beans ( Phaseolus acutifolius and P.
30 8 8 64 vulgaris), bottle gourds (Lagenaria siceraria), chili
32 8-10 10 90 peppers, avocados, and amaranths, it is assumed that
40 8 11 88 they represent the product of an improved environ-
44
44
8
8
16
17
120
136
ment resulting from the practice of agriculture — that
maize had been domesticated and become
at this stage
45 8 12 106
a cultigen. Consequently we are designating this maize
51 8 17 136
55 as “early cultivated.”
8 18 144
61 10 19 190 If this maize from Zone D is indeed cultivated maize,
it is noteworthy that domestication had effected little
43.0 8.3 13.6 113.8
change except in size. In its botanical characteristics
Not originally assigned to Zone D but obviously be- the early cultivated maize is virtually identical with
longing to it is a collection of 43 cobs from Zone C 1
wild maize from which it stemmed. And if the longer
which are quite similar to those described above. One cobs do indeed represent cultivated maize then it seems
of these cobs which is intact with respect to length is all the more likely that the smaller cobs from the lower
49 mm. long, has eight and ten rows, 15 spikelets per zones represent maize grown in the wild.
row, and a total of about 135 spikelets. The one exceptional cob, a fragment mentioned
181
Table 21. Maize Remains from

by Zone and Phase
All Sites
A
| ! j 1
stalks)
races
or
races
all
(husks) (leaves
systems
sheaths
husks
husks
all
Prophylls
Kernels, Tassels
Leaves
Shanks
Midribs Quids Quids Totals
Roots Stalks Outer
Cobs, Leaf Inner Husk
Venta Salada
Tc 35e, A 895 158 i 26 13 9 33 1 1136
Tc 50, 1380 16 18 40 39 8 17 44 9 4 17 5 11 1608
Tc 50, II 2273 21 2 2 93 65 82 68 64 40 5 4 18 2737
Tc 35e, B 841 215 2 35 18 4 13 4 14 5 1151
Tc 50, III 4115 3 4 125 103 54 119 227 29 54 16 13 17 4879
Tc 255, A 56 54 24 2 136
Tc 35e, C 2613 2 2 8 13 35 9 2 16 22 4 8 2734
Palo Blanco
Tc 35e, D 959 323 2 26 3 5 24 1 5 1348
Tc 35e, E 140 1 4 2 1 148
Tc 50, IV 1793 1 1 48 10 48 1 98 122 1 3 1 2127
Tc 254, B 447 4 43 79 17 9 36 3 6 33 26 32 735
Tc 254, C‘ 134 134
Tc 50, V 639 7 4 5 4 13 7 5 1 685
Tc 272, A 200 4 6 2 212
Tc 272, B 97 3 1 4 105
Tc C 160
272, 1 5 4 3 1 1 1 176
Fig. 101. Left to right: fragment of a stem, a husk system, a
Tc 272, D 67 4 6 6 8 91
Tc 272, E 170 3 4
cob, and a staminate spike from earlv zones of San Marcos
4 1 4 186
Tc 50, VI 2302 4 3 54 15 10 78 1 25 4 1 2497 Cave.
Tc 272, F 29 3 1 3 3 39
Santa Maria
Tc 50, VII 592 32 14 24 12 36 7 1 1 719
tripsacoid —
as similar cobs were found in contem-
Tc 272, G 30 2 3 2 37
poraneous levels of Coxcatlan Cave (see Table 22).
Tc 272, H 12 13 25
Tc 11 21
272, 1 10
Husks
Ajalpan
Associated with this collection was an almost intact
Tc 254, C 174 5 14 20 4 11 1 229
Tc 272, J 12 1 1 6 20 husk system which yielded a surprising amount of
Purron
information. Its shank or peduncle was broken imme-
Tc 272, K‘ 1 3 4
diately below the lower leaf sheath and has a diameter
Tc 272, K 2 2
of 4 mm. The shank has two nodes 4 mm. apart. Since
Abe) as
a fragment of an ear bearing the floral bracts of two
Tc 50, VIII 15 1 8 24
Tc 50, IX 27 4 2 33 pistillate spikelets is attached immediately above the

Tc 254 D 114
,
102 1 1 1? 9
upper node, it is certain that the upper leaf sheath of
Tc 50, X 2 2
this specimen must have been the uppermost leaf
Coxcatlan
Tc 50, XI 25 1 5 1 1 33 sheath of the original husk system. That the lower leaf
Tc 254, E 26 9? 35 sheath of the specimen was the lowermost husk and
Tc 254, F 1 1
Tc
therefore the outer husk of the original system is less
50, XII 5 5
Tc 50, XIII 18 18 certain, although it does have the aspect of an outer

Totals 20,364 797 46 506 442 282 245 706 12 127 384 47 5 140 83 24,186 rather than the inner husk, as the description below
suggests. Both husks appear to be intact with respect
to length, although they are somewhat frayed at their
tips. The lower one is 90 mm. long and the upper 50
above, must now be accounted for. It is larger in mm. Both have conspicuous venation, the lower one
diameter than the cobs of early cultivated maize and having more prominent veins than the upper. The
has strongly indurated tissues of and glumes. its rachis ridges of the lower husk bear short hairs, and the sur-
It resembles a type which is quite common in Zone C face and margins of theupper husk have scattered
and will be described below. This cob, however, seems long hairs. Neither husk is terminated by a ligule and
to represent the beginning of a new race — early leaf blade, as are the husks of many modern varieties
182
two ranks, indicating that the spike was a lateral branch

from a branched tassel. The spikelets occurred in pairs,
one sessile and the other pedicellate; the glumes, which
are 7.8 mm. in length, are thin, lacking in distinct keels,
weakly veined, and completely glabrous. The branch

is illustrated in Fig. 101. After this photograph was
taken, we attempted to clean the specimen by dipping

it in hot water. This caused most of the spikelets to
separate from the rachis. Floating on the surface of

the water, they provided a very interesting picture
(Fig. 102), showing that the glumes are thin and trans-
lucent, quite different from the much thicker glumes of
Tripsacum or teosinte. Thus in the staminate spikelets
of the tassels as well as in the pistillate spikelets of
the cob, the early corn from Tehuacan has exactly the
same basic botanical characteristics as modern maize
Fig. 102. Staminate spikelets of a tassel fragment (Cox- and furnishes additional evidence that the ancestor of
catlan phase, San Marcos Cave; x 1.6). The glumes are cultivated corn was corn.
soft, membranous, and glabrous.
San Marcos Cave, Zone C

of corn. The husks are illustrated in Fig. 101, which
also shows a cob and a section of stalk.
Cobs
Because these two husks differ in their venation and The cobs of this zone, representing the Ajalpan phase
hairiness, we have assumed that one is the outer and which dates from 1500 to 900 b.c., included a new
the other the inner of a two-husk system. Ears with type similar to the single specimen described imme-
few husks are usually borne in the upper part of the diately Because of the indurated tissue of
above.
stalk, sometimes immediately below the tassel. The rachises and glumes, we have designated this type as
husks protect the young ear before pollination and in “early tripsacoid.” The term “tripsacoid” is one pro-
the early stages of development but flare open at ma- posed by Anderson and Erickson (1941) to describe
allowing the ear to disperse its seeds. Ears with
turity, any combination of characteristics which might have
few husks may also occur in lateral inflorescences in been introduced into corn by hybridizing with its rela-
which ears are borne at secondary positions at the tives, teosinte or Tripsacum. In both of these species
nodes of the peduncle. We are unable to determine the tissues of the rachis and the lower glumes are highly
whether this specimen represents the leaf sheaths of an indurated and the lower glumes are thickened and
ear borne high on the stalk or whether it was part of curved. Archaeological cobs showing these character-
a branched lateral inflorescence. istics are thus suspected of being the product of the
Since the average length of the longer ears from this hybridization of maize with one of two relatives.
its
zone about 50 nun., it appears that the inner sheath

is How this highly tripsacoid maize came into existence
would have barely covered the pistillate portion of is still an unanswered question. There seems little
an average ear. If there was a terminal staminate spike, doubt that it is the product of the hybridization of
as the stumps at the apices of some cobs indicate that maize with teosinte or Tripsacum. Where did this oc-
there may have been, the young spike would have cur? Neither teosinte nor Tripsacum is known in the
been enclosed or partially enclosed by the outer husk, Tehuacan Valley today, nor are there remains of either
which probably opened before anthesis allowing the among the archaeological specimens. This does not,
staminate spikelets to shed pollen. We have observed however, rule out completely the possibility that one or
many ears of this type in extensive cultures of pod- both species once grew in the valley. Another possibil-
popcorn representing our genetically reconstructed ity is corn of Tehuacan was
that early cultivated
ancestral form. carried into other regions where it hybridized with
teosinte or Tripsacum, and some of the hybrid progeny
Tassel Fragment was later returned to the Tehuacan Valley. Both
Found also in Zone D was a small fragment bearing Tripsacum and teosinte occur widely today in the ad-
staminate spikelets. The spikelets were arranged in joining state of Guerrero. A third possibility is that wild
183
Fig. 103. Three types of cobs (Abejas phase, San Marcos

Cave; x 1.6) early cultivated, with slender rachis and
:
relatively long, soft glumes; early tripsacoid, with stiff Fig. 104. Various hvbrid combinations from the crossing of
indurated glumes; a possible hvbrid of the two preceding Tehuacan early cultivated corn with earlv tripsacoid corn
types. and backcrossing to both parents. Actual size.
maize grew in other parts of Mexico, and once domesti- glumes, of the early tripsacoid putative parents. Speci-
cation began, hybridized with Tripsacum or teosinte to mens of these are shown in actual size in Fig. 106. A
produce the new race “early tripsacoid.” The fossil substantial proportion of the cobs are two-ranked like
pollen from the Valley of Mexico described by Barg- the spikes of Tripsacum and teosinte. These disarticu-
hoorn et al. (1954) is evidence that wild maize once late readily, and in this respect some of the wild-type
existed there. There may be still other possibilities. It segregates may have been better adapted for survival
is at least clear that as early as the Abejas phase a in the wild than the original wild corn. In Table 22 the
new element had been introduced into the maize com- wild-tvpe segregates are listed together with the early
plex in the Tehuacan Valley. cultivated specimens, not because of a close genetic
Whatever the origin of the early tripsacoid corn, it relationship, but because the two types would yield
evidently hybridized with both the wild and early about the same amount of food.
cultivated corn of the Tehuacan Valley to produce
hybrids which in their cob characteristics were inter-
H usks
mediate between those of the parents. First generation Eleven pieces of outer husks intact in length have
hybrids in turn crossed back to both parents to pro- lengths of 90, 90, 90, 93, 100, 110, 114, 120, 120, 135,
duce great variation in both the wild and cultivated and 140 mm. These specimens are quite thick and con-
populations (see Figs. 103-106). spicuously ridged compared to the inner husks. All are
One of the most conspicuous products of this pre- lacking in ligules and leaf blades. Most are glabrous,
sumed hybridization was a type which we have called but several have sparse long hairs on the ridges.
“wild-type segregates.” Cobs are about the size of cobs Four pieces of inner husks have lengths of 90, 90, 120,
of the original wild corn but possess some of the char- and 120 mm. The two shorter pieces have profuse long
acteristics, especially the indurated raehises and hairs; the longer pieces are almost glabrous.
184
Fig. 106. Wild-type segregates (Palo Blanco phase, Cox-

catlan Cave, actual size). At lower left is a tassel branch
with basal pistillate spikelets. Another specimen is disti-

chous with several spikelets.
sheaths as completely glabrous as the leaf sheaths of

the early Tehuacan corn. Several of the sheaths are
depicted in Fig. 107.
Stalks
In addition to the stalk pieces with attached roots

Fig. 105. Cobs from a single cache in San Marcos Cave,
showing the great variation which followed hybridization of to be described below, there are 14 other short pieces
Tehuacan corn with earlv tripsacoid corn. The two cobs at of stalks from this zone. These vary in diameter from
upper right are wild-type segregates. Actual size. 4 to 15 mm., with an average of 9 mm. Since the posi-
tion of the pieces on the original plants is not known,
there is no way of distinguishing the thicker basal
pieces from the more slender apical pieces. However,
Leaf Sheaths even the thickest of the pieces, which may represent
Of 20 specimens of leaf sheaths in this zone, 14 are one of the lower internodes, is still quite slender com-
completely glabrous. All of these are narrow, indicating pared to the basal internodes of modern corn.
that they had come from slender stems. Two of these
Roots
are still attached to slender pieces ofThree of the
stalk.
sheaths have spatse long hairs on the ridges, two have Five specimens of basal-stalk fragments with roots
short hairs between the ridges, and one, the broadest attached provide some interesting bits of information.
of the lot, has profuse long hairs on the ridges and Four of these were intact (Fig. 108) and have diameters
profuse short hairs between. The glabrous condition of 12, 11, 17, and 10 mm. respectively. These dimen-
of the narrow sheaths is probably a characteristic of sions are considerably smaller than those of modern
the wild type. No modern corn, known to us, has leaf corn grown under agricultural conditions. The primi-
185
Fig. 107. Leaf sheaths (Palo Blanco phase, San Marcos Cave, actual size).
tive races of Colombia, Polio and Pira, described by until the permanent root system, which develops ad-
Roberts et ah, 1957, had minimum and maximum stem ventitiously from the lower nodes of the stalk, begins
diameters slightly above ground level of 21.8-24.2 and to function.However, in certain drought-resisting
21.6-24.1 mm. respectively. The mesocotyls on two of varietiesgrown by the Indians in the southwestern
our specimens were intact and were quite short, meas- United States, the primary root system makes an early
uring 11 and 7 mm. Since the mesocotyl elongates and rapid penetration into the deeper, moister soils
during germination until the emerging seedling reaches and continues to function throughout the life of the
the surface of the soil, this indicates that the kernels plant (see Collins 1914). Persistence of a primary root
were barely covered with soil. One of the roots, il- system one of the archaeological specimens indicates
in
lustrated separately in Fig. 109, had virtually no meso- that the Tehuacan maize had a similar adaptation to
cotyl, indicating that the kernel from which it grew subhumid conditions. This would have been a useful
germinated on the surface of the soil. trait in a wild maize growing in this region. A seedling
One of the specimens had the main root of the

still resulting from germination induced by the first rain
primary or seminal root system. Although broken, this does not immediately develop its permanent root sys-
was 95 mm. long and had the stumps of numerous tem if the surface soil has become dry by the time the
broken secondary roots. The seminal root system, which seedling emerges. It is in a precarious state with respect
arises from the radicle of the embryo, is in most modern to its water supply unless its seminal root system can
maize a temporary one, maintaining the seedling only quickly penetrate the deeper moister soils and so main-
186
Fig. 108. These roots (Palo Blanco phase, San Marcos Cave, actual size) show that the plants had no tillers.
tain the seedling until additional rains moisten the sur- show the scars indicating that the plants once had
face of the soil and the permanent root system de- tillers.
velops. One more bit of information can be gleaned from

Another fact revealed by these five roots is that the these five root specimens. The upper regions of the
plants had single stalks and no tillers, otherwise there permanent have no secondary roots or scars where
roots
would have been visible scars where the tillers had secondary roots had once been attached. This shows
been removed or lost. This discovery is somewhat sur- that they probably were not covered with soil. The
prising. Virtually all grasses, wild or cultivated, have practice of hilling —pulling soilaround the base of the
the capability of producing tillers. Most grasses are plant —which is common in Mexico today, had appar-
single stemmed only when they are grown under condi- ently not yet been invented.
tions of great stress and are depauperate. Some of the
plants of our genetically reconstructed ancestral form
San Marcos Cave, Zones C and B
1
tiller profusely. It has generally been assumed that These zones, representing the Palo Blanco phase of
wild corn was a freely tillering plant (Weatherwax about 200 b.c.-a.d. 800, contained an abundance of re-
1954; Mangelsdorf 1958a). Yet none of the basal stalks mains of all parts of the maize plants except the ker-
from San Marcos Cave or from any of the other caves nels, of which there were only two. The numbers of
187
9
•V I
Fig. 110. Kernels of the races Chapalote, left, and Nal-Tel,
right, from the Venta Salada phase of El Riego Cave.
Actual size.
In other characteristics the two races are quite sim-

ilarand it is not generally possible to distinguish the
two by their cobs. For this reason we hoped to find
well-preserved kernels to show us with which race we
were dealing. In the last of the boxes containing the
hundreds of maize fragments from Zone B there were
two kernels wrapped in separate small packages. Open-
Fig. 109. Root (Palo Blanco phase, San Marcos Cave, ac-
ing the first, we found a kernel with orange pericarp like
tual size). Extreme shortness of the mesocotyl indicates
Nal-Tel. Opening the second we found a kernel with
that this plant grew from a kernel germinating on the
brown pericarp like Chapalote. We have since en-
surface of the soil.
countered both colors in collections from the other
caves in Tehuacan Valley. Kernels of the two races
from El Riego Cave are illustrated in Fig. 110.
specimens of each part identified are shown
Because we were unable to distinguish the cobs of
in Table
these two races, we combined them and described
21. A brief description of these follows.
them in an earlier article (Mangelsdorf et al. 1964) as
Cobs the Nal-Tel-Chapalote complex. We shall continue to
The introduction of the tripsacoid maize into the use this term. Further study, however, makes it now
Tehuacan Valley and its subsequent hybridization with seem probable that the original wild corn of the Te-
the wild and early cultivated maize resulted in the huacan Valley was Chapalote rather than Nal-Tel.
production of several new types. The most conspicuous One reason for reaching this conclusion is that the
of these is the new race which appears to be the an- majority of early cobs are brownish in color. Earlier
cestral form of two of the ancient indigenous races of we considered this to be the product of aging —old
Mexico, Chapalote and Nal-Tel, described by Well- cobs often do turn brown. Later became evident that
it
hausen brown pericarp color, Cha-

et al. 1952. In its even among the well-preserved cobs the lemmas and
palote one of the most distinctive races of Mexico.
is paleas of some are light and of others are brown. Of
It is found today in northwestern Mexico, principally the various alleles at the A locus on chromosome 3,
in the states of Sinaloa and Sonora. Nal-Tel, which which is responsible for plant color, two, A h and ap ,
is closely related to Chapalote (Wellhausen et al.), dif- in combination with the P locus for pericarp color on
fers from it primarily in having an orange-colored chromosome 1, produce brown pericarp and cob colors
pericarp. Nal-Tel also tends to have ears somewhat which are dominant to other colors of these tissues
shorter and with slightly fewer rows than those of (Emerson et al. 1935). Dominant genes are often more
Chapalote. common in wild or primitive populations than in highly
188
evolved ones, and in this respect Chapalote may be

regarded as more primitive than Nal-Tel. Also the fact
that both T ripsacum and teosinte have brown pericarps
suggests that this may be the “wild” color. The orange
color of Nal-Tel may be a mutation which occurred
after maize had hybridized with Tripsacum or teosinte.
W We know that the introgression of Tripsacum or

teosinte into maize had mutagenic effects (Mangelsdorf
1958 6 ).
Of 581 cobs in these zones, 133 or 23 percent could
be assigned to the Chapalote-Nal-Tel complex (Fig.
ji
*
* j i-'i 111). Eight cobs from Zone C (Ajalpan phase) can now
•aim also be included in this complex, as can two cobs from

Zone IX (Abejas phase) of Coxcatlan Cave. All of these
y*\*< are very primitive.
*« v The remaining cobs from Zones B and C represent

for the most part types previously encountered. The
introduced tripsacoid had increased in frequency and
now comprised 30 percent of the total cobs. Strongly
tripsacoid segregates accounted for another 18 percent.
These two types together, now called “early tripsacoid,”
were the predominating type, accounting for 48 per-
i* cent of all identified cobs in Zones B and C 1 .
Second importance numerically, accounting for

in
38 percent of the cobs in Zone C is the type that we

1
,
call “wild-type segregates” (included in Table 22 un-
der “early cultivated”). These cobs resemble the original

v- wild maize of Zones F, E, and D in size, but the major-
have glumes and rachises which are more indurated
A* js ity
U ,*r M:, than those of wild corn. A number are two-ranked and
* -
T v
•*
kV
A i \
!
some have single spikelets. Representative specimens
of this type are illustrated above in Fig. 106.
i* * ?*/.
One explanation to account for these wild-type
:
'T
'
e
*.» *
>5 segregates is that the introduced tripsacoid com hy-
*v > bridized not only with cultivated corn but also with
J:; wild corn still growing in the valley. The first generation
** > S‘
/ *
I v54 hybrids arising in the wild habitat would then have
backcrossed to the wild maize, producing segregating
it <3? j*‘
populations containing some individuals which were
*
*#*4
* •“', quite capable of surviving in the wild. Indeed the wild
). -* M •
,Vv #
*
maize may actually have been improved in some re-
#I 1
*.
-<-v; spects by the genes for toughness and hardness con-
ferred on it by the tripsacoid maize.
*% * 5 Some of the remaining cobs were classified as inter-
* * v ’
mediate between the introduced tripsacoid and the
V*^ * iA wild maize. In the frequency polygons in Fig. 123
•l>'| all of the tripsacoid types have been combined in the
category designated as “early tripsacoid.”

Beside these more primitive tripsacoid cobs there
were 108 cobs that showed tripsacoid introgression
Fig. 111. Cobs of the Nal-Tel— Chapalote complex (Palo with the more modern races (Nal-Tel-Chapalote).
Blanco phase, San Marcos Cave, actual size). These are termed “late tripsacoid” and are very difficult
189
Husks
There were many fragments of husks in this zone,
but only 45 were intact in length. Thirty-six of these
were classed as outer husks and nine as inner. The
length of the former varied from 70 to 210 mm., with
an average of 145 mm. The majority were fairly close
to the average. The husks classified as inner varied from
100 to 160 mm., with an average of 135 mm. All of the
husks except two were lacking in flag leaves. One of the
exceptions had a rudimentary leaf, the other a well-
developed leaf that is 100 mm. in length. As mentioned
below in the description of leaf sheaths, some of the
husks, which actually are modified leaf sheaths, were
colored. Husks are shown in Fig. 112 and 113.
Prophylls
“In the branch system of grasses, each lateral shoot

has a bikeeled first leaf (prophyll) facing the axillant
leaf and addorsed to the main axis” (Arber 1934). This
structure, the prophyll, is especially prominent in mod-
ern maize where it subtends the ear-bearing branch.
By the laymanit may be regarded as one of the husks.
It from the husk, however, in being distinctly

differs
bikeeled. Three prophvlls were found in Zone B, and
as was to be expected, all were bikeeled.
Shanks
The stem subtending the ear and from which the
husks arise, in botanical terms a peduncle, is commonly
called the shank. On
complete husk enclosures, the
six
diameter of the shanks at the base of the lowermost

husk average 8.3 mm., and the distance between the
lowermost and uppermost husk varied from 23 to 74
mm., with an average of 36 mm.
Fig. 112. Husks (Venta Salada phase, Tecorral Cave,
Stalks
actual size).
The 43 pieces of stalks were sufficiently well
total of
preserved to be measured with respect to diameter.
Thirteen of these had remnants of roots and therefore
to define exactly (see Fig. 120 below). Data on intact represent the basal internodes. Their average diameter
cobs from all zones of San Marcos Cave are presented was 11 mm., which is about half that of the Colombian
in the accompanying tabulation. The figures for each primitive races, Polio and and Pira, grown under agri-
zone are averages based on the total number of intact cultural conditions. The remaining fragments, probably
cobs from the zone. representing a more or less random sample with respect
to their original positions on the plant, varied in diam-
Tc 254 Length Spikelets Total
eter from 3 to 18 mm., with an average of 7.2 mm.
zones Cobs in mm. Rows per row spikelets
Pieces of stalks are illustrated in Fig. 114.
B 91 55 11.0 14.8 163
C 1
60 47 10.8 12.4 134
Leaf Sheaths
C 2 45 10.0 12.0 120
D 10 43 8.3 13.6 113 Of the 79 leaf sheaths examined, 67 proved to be
E-F 8 22 7.3 7.5 55 completely glabrous, and 11 had short hairs in the
190
grooves between the ridges. One had sparse hairs on

the ridges. Actually, to the naked eye as well as to the
touch, even the 11 short-haired sheaths appeared
glabrous. Thus with respect to hairs or bristles these 79
specimens represent the most uniform lot of sheaths
which we have encountered in either archaeological or
modern maize. It seems highly improbable that one of
the ancient indigenous races of maize of Mexico, Palo-
mero Toluqueno, which has strongly pilose leaf sheaths,
could have originated from the populations found in
the Tehuacan Valley. This is one of several reasons for
supposing that there may have been more than one kind
of wild corn.
Practically all of the prehistoric leaf sheaths are
brownish or reddish brown in color. We supposed
initially that this represented discoloration from aging.
Some specimens, however, had a distinct reddish cast
and when soaked in hot water imparted a reddish tinge
to the water. It now seems probable that the color of
these specimens is not mere discoloration from age but
represents the remains of natural plant color. As men-
tioned above, some of the husks were also colored.
Leaves
Numerous fragments of maize leaves were found.
The majority were too badly broken to yield much
information, but four were intact for both length and
width and ten additional ones for width. The intact
leaves had an average length of 47.6 cm., and an av-
erage width of 4.1 cm. The leaves intact for width
varied from 2.7 to 5.0 cm., with an average of 3.8 cm.
These dimensions are much smaller than those of any
modern races of maize in Mexico even the ancient —
indigenous races, which compared to some other races
have relatively narrow leaves. Leaves arising from the
ear-bearing nodes of Chapalote and Nal-Tel, for exam-
ple, have average lengths of 80.5 and 65.6 cm. respec-
tively and widths of 7.6 and 8.7 cm. (Wellhausen et al.

1952).
The venation index —a
measure of the number of
veins per unit (cm.) of width —
of 17 leaves which were
scored for this characteristic varies from 3.8 to 6.7, with
an average of 4.5. This is much higher than the vena-
tion indices of modern Mexican maize (Wellhausen et
al. 1952) and puts prehistoric maize in a class with
teosinte with respect to this particular characteristic.
Tassel Fragments
Thirty-three tassel branches or fragments of branches

were found in this zone. The length of the pedicel on Fig. 113. Branched lateral inflorescences (Palo Blanco
the pedicellate spikelet was measured on one spikelet phase, San Marcos Cave, actual size). Such plants probably
from each specimen. This varied from 1 to 6 mm., bore a cluster of small ears.
191
Fig. 114. Pieces of stalk (Palo Blanco phase, San Marcos Cave, actual size).
192
Fig. 116. Tassel fragments (Venta Salada phase, San

Marcos Cave, actual size).
these four specimens resemble the original wild or

early cultivated corn.
One of the unique characteristics of maize is the
central spike of the tassel which differs from the
branches in being polystichous rather than distichous.
There have been various theories about the origin of
Fig. 115. Tassel (Palo Blanco phase, Coxcatlan Cave) and this structure (see Mangelsdorf and Reeves 1939 for
leaf fragment (Venta Salada phase, Tecorral Cave). Both a review of this subject). It is interesting to note that
actual size. the polystichous central spike was characteristic of
corn in prehistoric times. The several fragments repre-
senting central spikes are illustrated in Fig. 116. An
averaging 3.7 mm. The lower glumes on these same intact tassel including a central spike from Coxcatlan
spikelets varied in length from 7 to 12 mm., averaging Cave is shown in Fig. 115.
9.2 mm. There was some variation in the shape of the Because our genetically reconstructed ancestral form
spikelets, the majority being somewhat flat dorsally, bore seeds in the branches of the tassel thus providing
giving the impression of a bikeeled structure. This is a means of dispersal which modern corn lacks, we
characteristic of the glumes of some species of Trip- expected some evidence of this characteristic
to find
sacum. Five of the 33 specimens, however, did not in Tehuacan corn. In this
the tassels of the prehistoric
exhibit the impression of distinct keels and all of these we were disappointed. Of the 384 specimens of tassels
were shorter than the average and resembled the spike- found in the five caves, only one from Coxcatlan Cave
lets in the single fragment found in Zone D and il- (Fig. 115) seemed to be a tassel branch bearing pistil-
lustrated in Fig. 101. Four of these five tassel fragments late spikelets at its base. Since depauperate corn plants
differed also from the majority in lacking hairs on the even of modern corn sometimes exhibit this condition,
glumes. In size, shape, and hairiness of the glumes, its rare occurence among the prehistoric specimens is
193
not especially significant. On the contrary, it is sur-

more
prising that there are not of these, since both
wild and cultivated corn must at times have been
exposed to environmental conditions which were con-
ducive to production of depauperate plants.
Quids
There were in this zone 58 remains of quids, of which
32 were chewed stalks and 26 chewed husks. The latter
probably originally contained young ears. We had
previously found from trials associated with explain-
ing the quids in La Perra Cave (Mangelsdorf et al.
1956) that young ears enclosed in husks are quite sweet.
Also it is well known that growing com stalks from
which the ears have been removed, or which are barren

for other reasons, often accumulate sugars and are about
as sweet as sugar cane —
although somewhat less pal-
atable when chewed. Since a “sweet tooth” is almost
universal in the human race, it is not surprising to find
that the Tehuacan people made use of the sweetness
derived by chewing young ears and stalks of com. Quids
in various stages of maceration are shown in Figs. 117
and 118.
Collections from Other Caves

The maize specimens from the remaining four caves
(Coxcatlan, Purron, Tecorral, and El Riego) did little
more than to corroborate the conclusions drawn from

the collection from San Marcos Cave. But they did so
very effectively, as perusal of Table 22 will show. Cox-
catlan Cave, which was occupied more often and over
a longer span than San Marcos, had a large sample
totaling over 15.000 specimens. The majority of the
earliest specimens from components of the Coxcatlan
phase (11 fragments and 6 whole cobs from Zone XIII,
5 whole cobs from Zone XII, and 6 fragments and 6
whole cobs from Zone XI) showed that Coxcatlan
Cave’s earliest maize has four- and eight-rowed cobs
with relatively long glumes and slender rachises. This
we assume to be wild com. However, one cob in
Zone XIII and three from Zone XI were somewhat
larger in size, and these are presumably the first prod-
ucts of cultivation. In the Abejas levels of Coxcatlan
Cave cobs of both wild and early cultivated types occur,
but the early cultivated specimens (one cob from Zone
X, 11 cobs and 7 fragments from Zone IX, and 7 cobs
from Zone VIII) far outnumber the wild ones (7 from
Zone IX and 2 from Zone VIII). Also, at this time the
Fig. 117. Quids (Palo Blanco phase, San Marcos Cave, first tripsacoid cobs (one cob from Zone X and 6 from
actual size). Most of these are chewed corn stalks. Lower Zone VIIII) make their appearance. Two tiny cobs
specimens show how the chewing proceeds along the stem. of the Nal-Tel-Chapalote complex were uncovered in
Actual size. Zone IX.
194
Fig. 118. Quids in various stages of maceration (Palo Blanco phase, San Marcos Cave, actual size).
The Purron phase was meagerly represented in Zones

K and K of Purron Cave by three burned cobs of early
1
tripsacoid com. Zone J of Purron Cave supplemented

our larger Ajalpan sample from Zone C of San Marcos.
Zone showed a dominance of
J also early tripsacoid
types (10 cobs) with a lesser number (2 fragments) of
early cultivated types.
In the Santa Maria zones of Coxcatlan Cave (Zone
VII) and Purron Cave (Zones G, H, and I) there is
further evidence of more use of the Nal-Tel-Chapalote
complex as well as hybridization of this maize with
both wild and cultivated corn to produce new varia-
tion on a grand scale (late tripsacoid, Fig. 120).
A New Type: Slender Pop
In the Palo Blanco phase the collections from Cox-

catlan, Purron,and El Riego caves do, however, reveal
a type of maize which, if it occurred in San Marcos
Cave, was not clearly recognized. We have designated
Fig. 119. Cobs of slender pop (Venta Salada phase, Cox-
this type as “slender pop.” The cobs (Fig. 119) are
catlan Cave, actual size).
slenderer than those of Chapalote or Nal-Tel and are
more nearly cylindrical in shape. Remains of occasional
kernels indicate that they were quite small, rounded in Naranja found in the Department of Nariho in southern
shape, and orange in color. This may be the prototype Colombia and described by Roberts et al. (1957). It
of a Mexican popcorn, Arrocillo Amarillo, one of the is difficult to see how this Colombian race could have
four ancient indigenous races described by Wellhausen had its antecedents in the Tehuacan Valley and have
et is now mixed with many
al (1952). This race, which left no trace of its former presence in any areas between
others, occurs in most nearly pure form in the Mesa

its these widely separated points in Mexico and Colombia.
Central of Puebla at elevations of 1600-2000 meters, Perhaps the resemblances are merely coincidental, but
not far from the Tehuacan Valley and at similar alti- perhaps future investigations will show that the pre-
tudes. Another modern race of maize which resembles pop of Tehuacan and the orange pop-
historic slender
the prehistoric slender pop even more closely is Pira corn of Colombia are in some way related. In any case,
195
Table 22. Maize Cobs and Kernels by Race
Nal-tel
Wild Early Early and Late Zapalote Slender Zapalote Tepe- Chal-
Maize Cultivated Tripsacoid Chapalote Tripsacoid Chico Pop Grande Conico Dent cintle queno
fragments fragments fragments fragments fragments fragments

cobs cobs cobs cobs cobs
Kernels
Whole Whole Whole Whole Whole

Cob Cob Cob Cob Cob Cob
Venta Salada
Tc 35e, A 67 126 2 232 137 22 16 205 1
Tc 50, I 22 16 632 28 208 6 121 4
Tc 50, II 1 46 589 16 349 152 444
Tc 35e, B 68 143 105 1 4 130 231 1 3
Tc 50, III 18 1395 1613 96 392 41 107 208
Tc 255, A 7 2 32
Tc 35e, C 16 184 420 134 445 37 409
Palo Blanco
Tc 35e, D 41 130 100 1
Tc 35e, E 3 2 12 26 1
Tc 50, IV 13 38 257 553 122 528 2 34 95
Tc 254, B 1 2 226 123 93 2
Tc C
254, 5 51 53 10 15
Tc 50, V 1 130 157 12 286 2
Tc 272, A 20 84 12 16 31
Tc 272, B 10 28 16 23
Tc 272, C 29 73 2 11 2 12
Tc 272, D 2 20 21 5 6 3 4
Tc 272, E 4 1 12 92 4 13 5 25
Tc 50, VI 163 1521 9 159 291 1 1
Tc 272, F 5 12
Santa Maria
Tc 50, VII 30 37 35 54 268 26 47 73 1 2
Tc 272, G 14 16
Tc 272, H 11 1
Tc 272, 1 2 6 2
Ajalpan
Tc 254, C* 9 28 7 36 86 8
Tc 272, J
Purron
l
Tc 272, K
Tc 272, K
Abejas
Tc 50, VIII 2
'
7
Tc 50, IX 7 7 11
Tc 254, D 20 38 15 28
Tc 50, X 1
Coxcatlan
Tc 50, XI 6 16
Tc 254, E 5 21
Tc 254, F 1
Tc 50, XII 5
Tc 50, XIII 11 6
196
ern races were found in the upper zones of Coxcatlan

and El Riego caves. These included several cobs and
kernels resembling the race, Conieo (Fig. 121), and one
cob each of Zapalote Chico, Tepecintle, and Chal-
queno, as well as kernels of an unidentified dent corn.
The surprising thing is not that specimens of these
modem races were present, but that there were so
few of them. These races were certainly in existence
in Mexico centuries ago. At least two modem races,
Tuxpeno and Olotillo, are recognized in the molds on
Zapotec funerary urns dated at about a.d. 600-700. If
they were common enough to furnish the motifs for

funerary urns in the region now represented by the
state ofOaxaca, it seems strange that they did not find
their way in any substantial numbers into the agri-
culture of a region which is now the adjoining state of
Fig. 120. Cobs of late tripsacoid corn (Venta Salada phase,

Coxcatlan Cave, actual size).
the slender pop,which first appeared at the very end

of the Santa Maria phase (2 cobs from Zone VIII,
Coxcatlan Cave), increased rapidly and steadily in
frequency, comprising 21 percent of all cobs of the
finalphase.
Judged by its cobs alone the slender pop might be
expected to be less productive than Chapalote or Nal-
Tel. The fact that it increased in prominence suggests
that although the ears are small the stalks may have been
prolific, bearing more than one ear. The present-day
race, Arrocillo Amarillo, to which the slender pop bears
some resemblance and which it may be related, is
to
prolific, usually producing two or three ears per stalk.
The maximum yields of corn in the United States are
made by prolific tvpes of corn grown under irrigation.
Perhaps this was also true in prehistoric Tehuacan. Fig. 121. Cob of the race Conieo, on the left (Venta Salada
phase, Coxcatlan Cave), and kernels and cob of the race
Cobs of Modern Races
Tepecintle (Venta Salada phase, El Riego Cave). All
A few cobs which could be assigned to various mod- actual size.
197
Fig. 122. Cobs (actual size) illustrating an evolutionary sequence from about 5000 b.c. to about a.d. 1500: wild corn,
Coxcatlan phase, San Marcos Cave; earlv cultivated, Abejas phase, San Marcos Cave; Chapalote, Palo Blanco phase, San
Marcos Cave; Chapalote, Venta Salada phase, Coxcatlan Cave; Conico, Venta Salada phase, Coxcatlan Cave.
Puebla. Here is one of the numerous minor questions from the latter only in the fact that the cobs are on
which the prehistoric remains of corn from Tehuacan the average more tripsacoid than those of Chapalote
Valley have raised while answering the much more and Nal-Tel. Also some of the cobs are similar to those
important major ones. Cobs illustrating an evolutionary' of the slender pop, but somewhat more tripsacoid. In
sequence are shown in Fig. 122. any case it is these three types, Nal-Tel-Chapalote,
The polygons in Fig. 123 show when each of the late tripsacoid, and slender pop which supported the
recognized types of corn made its first appearance, expanding populations of the Tehuacan Valley from
reached its highest frequency, and in the case of the about 900 b. c. to a.d. 1500.
wild and early cultivated corn, disappeared. The
boundaries between the categories represented in Fig. Wild Corn Reconstructed
123 are by no means hard and fast. The type called A well-preserved early cob and an intact husk sys-
“late tripsacoid,” for example, might well be included tem consisting of an inner and outer husk from the
as part of the Nal-Tel-Chapalote complex. It differs Abejas phase, together with a piece of staminate spike
198
Fig. 123. Changes in types of corn in the Tehuacan Valley, 5000 b.c. to a.d. 1500, in terms of percentages of number of
cobs identified. Specimens of corn are almost totally lacking from the Purron phase.
(actually found in the Ajalpan phase), all from San which they enclosed the kernels to a genotype which
Marcos Cave, provide the material for the reconstruc- we have produced in our experimental cultures by com-
tion of an ear of Tehuacan wild corn. This is illustrated bining one of the alleles at the tunicate locus, tu h on
,
in actual size in Fig. 124. An ear with only two husks chromosome 4 with a major tunicate modifying gene,
was probably borne either in high position on the main Ti, on chromosome 6. The terminal inflorescences, the
stalk or in a terminal or near terminal position on a tassels, were similar to those of modern maize in hav-
branched lateral inflorescence. In either case the husk ing a polystichous central spike and distichous
would have served primarily to protect the young ear branches. Apparently they were completely staminate
and would have flared open at maturity, permitting and in this respect differ from our genetically recon-
the ear to disperse its seeds.The ear, like ears of some structed ancestral form in which many of the tassels
modern primitive races, was terminated by a fragile bear both staminate and pistillate spikelets. The ker-
staminate spike bearing spikelets in pairs, one mem- nels, which were borne in pairs, were almost isodiamet-
ber of each pair pedicellate and the other sessile. The ric in their dimensions, were rounded dorsally, and
glumes of the staminate spikelets were membranous, were either brown or orange in color, probably the
glabrous, and lacking in keels. former. The specimens from both San Marcos and
The wild corn was a form of pod corn. Its seeds were Coxcatlan Caves show that the plants apparently had
partially but not completely enclosed in floral bracts. single stalks, although they may have had the capacity
These are similar in relative lengths and the extent to to produce tillers under especially favorable condi-
199
lateral branch had as its first leaf a bikeeled prophyll.

The root system, like that of modern maize, comprised
both seminal roots arising from the radicle of the em-
bryo and permanent roots arising from the first and
higher nodes of the stalk. The seminal root system, like
that of certain drought-resistant varieties grown by the
Indians of the American southwest, may have been
more extensive and persistent than in most modern
varieties.
In all of its essential botanical characteristics, the
wild and early cultivated corn of Tehuacan was iden-
tical with modern corn but was smaller in all of its
parts. Indeed, the wild corn with its tiny ears must
initially have been less useful as a food plant than
the wild squashes and scarcely more promising than
some of the weedy grasses of our gardens and lawns.
It was, however, responsive to the improved environ-
Fig. 124. Wild corn (actual size) reconstructed on basis of ment provided by agriculture, and cultivation imme-
a fragment bearing male spikelets and kernels from earlv diately produced a substantial increase in size. Subse-
levels of San Marcos Cave. The husks probablv enclosed quent hybridization with its relatives, Tripsacum or
the voung ears completely but opened at maturity, per- teosinte, initiated an explosive evolution which resulted
mitting dispersal of the seeds. The kernels were round,
in tremendous variability and a manyfold increase in
brown or orange, and partlv enclosed bv glumes.
size. Despite the spectacular increase in size and pro-
ductiveness under domestication, which helped make

corn the basic food plant of the pre-Columbian cultures
tions, as do virtually all other wild grasses. The leaf and civilizations of America, there has been no sub-
sheaths and husks were completely glabrous and some, stantial change in 7000 years in the fundamental
if not all, were reddish brown in color. The ear-bearing botanical characteristics of the com plant.
REFERENCES
Anderson, Edgar, and Erickson, 1941. Mangelsdorf and Lister, 1956.
Arber, 1934. Mangelsdorf and Reeves, 1939.
Barghoorn et ah, 1954. Mangelsdorf et ah, 1956, 1964.
Collins, 1914. Roberts et ah, 1957.
Emerson et ah, 1935. Stonor and Edgar Anderson, 1949.
Grobman et al., 1961. Wagner, 1964.
Irwin and Barghoorn, 1965. Weatherwax, 1954.
Mangelsdorf, 1958a, 1958/;, 1961. Wellhausen et ah, 1952.
200
CHAPTER 10
Archaeological Phaseolus from Tehuacan
Lawrence Kaplan
he genus Phaseolus of the legume family com- snap beans and dry or field beans is based on use rather
T prises about 160 species (Willis 1951). Because

of extensive synonymy of recorded species
names, particularly among the Old World species in
than on botanical characteristics.
This
tropical
is the most widely
indigenous
grown of all food beans. In
American agriculture common
this genus, the actual number is probably considerably beans are grown on neutral to slightly alkaline soils
smaller. On the basis of Piper’s monograph (1926) of from sea level to over 2000 meters elevation.
the American Phaseolinae, about 80 species are na- P. acutifolius Gray, var. latifolius Freeman — tepary,
tives of the New World. escomite. At the present time this bean is not cultivated
Many of the New World species are perennials, with on a commercial scale, although some field stations
white flowers or, more pigmented with
often, flowers have been investigating its potential as a dry-land seed
red or purple anthocyanins. Yellow flowers are uncom- or forage crop. The seeds of this species are smaller
mon. Pods of the cultivated species are cylindrical or than those of other cultivated American beans, but
broad. Many of the Old World species have yellow they overlap in size the smaller-seeded common bean
flowers and all have morphological features of the varieties. Collections made a few decades ago (de-
flowers that contrast with floral structure of the Ameri- posited in the Museum of Anthropology, University of
can beans. Six Asian-African species, including the California, Berkeley) show the tepary to have been
grams, urds, mung, and rice beans, are widely culti- common markets along the Pacific coast of northern
in
vated in Asia and have been introduced elsewhere to Mexico. In 1954 it was infrequent in markets in that
a limited extent. These Old World cultivated species region. The disjunct range of this bean from the —
are annuals and have small seeds and small cylindrical Sonoran Desert south through Jalisco and then a gap in
pods. Differential susceptibility to certain fungus its occurrence until the Tapachula-Guatemalan border
diseases further distinguishes the Old and New World region —suggests that the present distribution is a relic
species. distribution (Fig. 127).
Four cultivated species of New World origin are P. lunatus L. — sieva, small lima, Carolina bean, frijol
important as food crops: de haba Mexico this may also apply to the intro-
(in
P. vulgaris L. —
common bean, frijol (derived from a duced broad bean, Vicia faba), comba (Guerrero),
name for theOld World broad bean, Vicia faha); etl, patashete (Chiapas), guaracara (Venezuela), cubace
yeti (Nahuatl); buul, chenek (Mayan); poroto (Queehua (Costa Rica). The curved pod and smaller, often flat
and Aymara). This polymorphic, poorly understood seed distinguishes the sieva from the lima, P. lunatus
species includes hundreds of cultivated varieties; for var. macrocarpus Bentham (synonym, P. limensis
example, Navy, Red Kidney, and Pinto beans. P. vul- Macf., the big lima, or in Peru, pallar). Any lima bean
garis and the three species discussed subsequently in- encountered in Mexico, Central America, or in in-
clude strongly to weakly vining forms and bush or digenous agriculture in the southwestern United States
erect forms. The distinction between green, string, or is likely to be a sieva bean. In the southeastern United
201
States this type of bean lias been collected among to time, boiled, and eaten. The cooking water is said
Cherokees and Seminoles. The commercial limas of the to be discarded.
United States, such as Fordhook and Henderson, are Runner beans are cultivated to a more limited ex-
of the macrocarpus group and are widely grown in tent in the northern Mexican states of the Central
gardens and commercially in the humid coastal valleys Plateau, and according to Bukasov (1930), in Costa
of California. Rica, Panama, and Colombia.
The morphologically similar sievas and limas are
fully interfertile and produce viable offspring (Mackie Nutrition and Domestication
1943). On the basis of the distribution of the wild The principal economic significance of Phaseolus
species and the archaeological evidence, it is now clear beans has probably always been as sources of vegetable
that these two groups are not distinct species, but are protein in the form of dry seeds. Several species, par-
independent domesticates deriving from geograph- ticularly P. vulgaris, supply edible immature pods and
ically separate races or subspecies. Wild P. lunatus (R. seeds — that is, snap beans— but because of their high
W. Allard, personal communication) ranges from Mex- w'ater content at this stage (about 89 percent; Block
ico to northernSouth America and exhibits differences and Weiss 1956), they do not function as an important
separating the domesticated groups. Arehaeologically, dietary protein. On a dry-weight basis, the protein con-
the geographic separation of sievas and limas attests tent of the immature pods and beans together totals
to independent domestication. The lima is present in about 10 percent (Chatfield and Adams 1940), or not
preceramic, preagricultural times (5200 years before quite half that of dry mature seeds, whose protein con-
the present; collections from Chilca, Peru, by Dr. Fred- tent is 22 percent (Block and Weiss 1956).
eric Engel) and continue into later times on the north The close relationship between beans and corn in
coast of Peru.The Peruvian lima is absent from the the indigenous diet of the populous cultures of Meso-
archaeology of Mesoameriea. The sieva, with the ex- ameriea and the Andean region is, like other tradi-
ception of a few pods of problematic identity in later tional dietary combinations, no accident. Quantitative
agricultural (Cupisnique) times, is absent from Peru- chemical analysis (Jones et al. 1938) of corn and beans
vian sites. of contemporary Yucatan Indians have shown com-
P. coccineus L. — runner beans, ayecote (Nahuatl plementation in the amino acids of zein, the principal
derivation); botil (Chiapas); patol, pato (northern Mex- corn protein, and a and b globulins of the black beans
ico). The most common variety in the United States is (P. vulgaris) that with the corn form the mainstay of the
the large-seeded Scarlet Runner, sometimes listed in diet. Later studies cited by Block and Weiss (1956) and
seed catalogues as an ornamental with edible seeds. Albanese (1959) indicate that where zein is the pro-
Seeds of most runner varieties are larger than those of tein in the diet of a monogastric animal, lysine is the
sievas and common beans and are usually purple or amino acid that is limiting and must be made up from
variegated purple, although there are white-seeded another source. Beans, with their relatively high lysine
varieties. The flowers of the purple-seeded varieties are content, supply this amino acid that is deficient in the
strikingly red; those of the white-seeded kinds are main food crop, corn, with the result that a dietary
white. protein of high biological value is achieved. It may
Cultivation of indigenous runner beans is most be noted, how'ever, that while this protein appears to
highly developed in the cool humid uplands of Chiapas be adequate for some members of the population, such
and Guatemala, in oak- and pine-forested regions as working males, it is inadequate for the more pro-
above 1800 meters. In these areas the cultivars and tein-sensitive members such as lactating mothers and
their abundant wild relatives are perennials. Sprouts recently wêaned children (Altschul 1962).
from the tuberous roots often take over maize fields for Beans of different varieties and even of different
the first year or two of the fallowing period. The Phaseolus species are high in lvsine and tryptophane.
strongly vining runner beans are interplanted with Corn varies considerably in total protein but is always
maize in some fields and in plots adjacent to houses. deficient in lysine and tryptophane (Block and Weiss
In home gardens they are treated as herbaceous peren- 1956). Because amino acid complementation exists in
nials with a life-span of two to several years. In Ma- combinations of many corn varieties and bean species
deira (Lowe 1868) the roots persist for seven years and and varieties, a kind of universal flexibility in adapta-
are regarded as poisonous. Tubers of the branching tion of these dietary components to human require-
root system in Chiapas may be taken up from time ments resulted. It is now clear that corn and beans in
202
ARCHAEOLOGICAL PHASEOLUS FROM TEHUACAN
—
some areas for example, Tamaulipas (Kaplan and of the extent of their use, ethnohistorical data from the
MacNeish 1960) and coastal Peru (Towle 1961) were — Valley of Mexico is presented here.
not domesticated simultaneously nor did they neces- A major source of information on pre-Hispanic trib-
sarily diffuse together. Wherever these two food crops utes is the Codex Mendocino or Codex Mendoza
met, an immediate adaptive combination favored by (Cooper Clark 1938), a manuscript based upon older
human selection was formed. sources and prepared at the direction of the Viceroy
The evolution of human diet is marked by synergistic Antonio de Mendoza, who served in New Spain from
combinations perhaps to a greater extent than by small 1535 to 1550. The second part of the codex is made up
increments in the food value of individual nutrient of tribute lists from settlements subject to Mocte-
sources.Corn could presumably be selected for incre- zuma II, the last Aztec emperor. A total of 371 towns
ments in total protein. However, because the nutri- were required to make regular contributions of corn,
tional insufficiency of plant-seed proteins is the result beans, and other foods, raw materials, and manufac-
of limiting amino acids rather than low total protein tured items. The symbols for maize and beans in the
content, raising the total protein content would not codex are “troxes” or bins shown filled with corn ker-
make it much more effective in the diet. The answer to nels, beans, and other small seeds, or mixtures of these.
the problem of lysine as a limiting amino acid in corn James Cooper Clark, editor and translator of the codex,
in indigenous America has been the supplementation says that “the troxe was a large 10 foot high chest of
of corn with beans, rather than changes in the total woven oziers and plastered inside with mortar. The
proteins or in the array of amino acids of com. etl or bean can easily be recognized as the black variety
How did such a useful combination come about? with the white eye commonly cultivated in Central
Sampling of the available flora is probably the only America and Mexico” (1938: p. 56). A total annual
answer to this question. Seasonality, yield, compatabil- tribute of twenty-eight bins of corn and twenty-one
ity with other crops and ways of human life were fac- bins of beans was exacted. Each bin is estimated by
tors, as well as the nutrients, in determining the adop- Clark to represent 8,000 bushels.
tion ofsupplementary crops. Sampling of the flora over Beans were evidently being grown in huge quanti-
a period of many thousands of years prior to the estab- ties in centralMexico at this time. Yet in Tehuacan
lishment of corn-and-bean agriculture roughly 6000 during the late pre-Hispanic Venta Salada period,
—
years ago in Middle America and possibly in the Te- beans are present only in the same two caves, El Riego
huacan Valley itself, it beans were domesticated there and Coxcatlan, in which they had been found in earlier
—was evidently sufficient to establish this combina- levels.
tion. Given sufficient time, a comparable flora, and

similar human practices, the same plants would prob- Structural Changes
ably be domesticated. Although the domesticated beans differ in several
In some regions having different floras — in, say, the ways from the wild species, the fact that generally only
Great Basin of the western United States or the humid pods and seeds appear in the archaeological remains
eastern United States —
indigenous plants do not limits the possible record of change through time to
appear have been brought under cultivation to the
to certain anatomical features. One of these is seed size.
extent that com and beans were in Mesoamerica. Why Seeds of the cultivated taxa vary greatly in size, but
does domestication take place in some areas but not they are always larger than the seeds of the most closely
in others? Further archaeological and botanical evi- allied wild species. Increased seed size can be deter-
dence of the emerging from Tehuacan should en-
sort mined from the seeds or indirectly from re-
directly
able groups of scientists to weigh the roles of time mains of pods. It constitutes the clearest evidence that
and the sequence of human culture, climates, and floras a given seed is from a domesticated rather than from
of the past in approaching this question. a wild plant. All of the archaeological beans that have
been studied are domesticates. Wild bean seeds have
Ethnohistory not yet been reported in archaeological remains nor
There seem to be no direct ethnohistorical sources have transitional phases in size increase been detected.
relating to the use of beans in the Tehuacan Valley. The early pod remains of P. vulgaris at Ocampo in
Be cause the paucity of beans in relation to corn re- Tamaulipas and in the Tehuacan caves contained
mains in some pre-Columbian sites, including some of seeds within the size range of moderately large com-
the Tehuacan caves, tends to give an erroneous notion mercial beans. The earliest teparies at Tehuacan and
203
the earliest limas from Huaea Prieta in Peru are not Table 23. Bean Remains from All Sites
outside the size range of later prehistoric or contem- by Zone and Phase
porary cultivars of the same species. Thus, from times
of incipient agriculture when remains of P. vulgaris, P.
Maria
acutifolius, and Peruvian P. lunatus first appear, no <q
To
Palo Blanco Venta Salada
u
X
significant size changes have occurred. The archae- Abejas
Oo
Santa
ological evidence clearly demonstrates that transition

in seed size took place in times earlier than the earliest XI
VIII
VII
VI
V IV E D C III B A
archaeological records of beans — that is, during early
50,
Tc
50,
Tc
50,
Tc
50,
Tc
50,
Tc
50,
Tc
35e,
Tc
35e,
Tc
35e,
Tc
50,
Tc
35e,
Tc
35e,
Tc Totals
incipient agriculture or even during pre-agricultural
P. vulgaris
times.
Pods
Any conclusions regarding the cultural context of this
1 1 1 1 1 3 3 3 14
Seed, var. 1 1 1 2 108 2 3 183 38 338

transition or the selective conditions under which it Seed, var. 11 149 10 159
occurred must be speculative and tentative at best. Seed, var. 13 2 2
The association of corn and beans at Tehuacan some Canavalia sp.
Seed 1 1
6000 years ago suggests that the transition in seed size
P. acutifolius
may have come about among beans that were inter- Seed, var. 7 118 485 603
planted with corn. Large seeds, by providing an abun- P. coccineus
dant energy reserve for young bean vines, enabled Pod 1 1
Seed, var. 9 6 7
them to survive better in competition for light with 1
Seed, var. 12 1 24 25
the young rapidly growing corn. P. lunatus
Another important feature of seeds that may be Pods 1 1 2
affected by domestication is permeability to water. Seed, var. 8 1 1 2
Totals 1 120 1 487 2 1 2 109 163 4 176 67

“Hard-seed,” an agronomically important and unde- 1,153
sirable trait, denotes the inability of the seeds to im-

bibe water. When a substantial proportion of seeds
designated for planting are impermeable to water, the
result is uneven germination, with the soft or permeable lost since heating of relatively dry protein and carbohy-
seeds imbibing water and germinating promptly and drate mixtures causes the formation of indigestible
the hard seeds exhibiting a delay in the start of growth. protein-carbohydrate complexes (Ellis 1959). Prepara-
This determined by characteristics of the seed
trait is tion of dry seeds by roasting and grinding, as in the
coat which are genetically and environmentally production of pinoles made from wild grasses, maize,
affected. Different varieties of beans may differ greatly or beans (Kaplan 1956), is probably even more det-
in the permeability of their seeds to water. Some culti- rimental nutritionally.
vated varieties have very low frequencies of hard seeds; Seed dissemination is another plant characteristic
others, such as runner beans from Cholula, Mexico, that may change under cultivation. Economic annual
produce seeds of which 50 percent fail to absorb water plants grown legumes
for their seeds or grains, chiefly
after 48 hours of immersion at room temperature. Seeds and have usually undergone a reduction or loss
cereals,
derived from wild species of the genus Phaseolus are of the ability to release their seeds. Wild beans have
invariably impermeable. The contrasting permeability pods that open along both sutures and forcibly expel
of seeds derived from domesticated and nondomesti- their seeds by a twisting of the pod valves that results
cated species suggest that seed permeability has in- from desiccation of the inner parchment tissue of the
creased under domestication. pod. This process of expulsion takes place violently
The widespread practice in Mexico of boiling beans among wild species but less so or not at all among
without preliminary soaking may reflect the high inci- cultivated varieties, owing to a reduction of the parch-
dence of impermeability in the local culivated varieties. ment layer among domesticates.
Imbibing of a seed lot of varying permeability followed The oldest P. vulgaris materials, the pods from Ta-
by boiling produces an unevenly cooked pot of beans, maulipas and Tehuacan, have reduced parchment lay-
whereas boiling of dry seed reduces the differences in ers and for this reason, as well as that of their size, are
cooking time between permeable and impermeable considered to be derived from domesticates. However,
seeds. The latter method may produce a uniformly the oldest P. coccineus pods from Tamaulipas (Kaplan
palatable pot of beans, but some protein value may be and MacNeish 1960) are tightly curling wild types.
204
Perennialism is more prevalent in Phaseolus than is were present 2000 years ago, but they did not become
animalism and is based on drought resistance of the abundant until 800 years ago. In the Southwest beans
tuberous root or root and basal portion of the stem. are abundant in the remains of cultures in which pot-
Probably about 90 percent of the American species are tery was utilized and agriculture was the main food
perennials. P. vulgaris and P. acutifolius are known source.
only as annuals with fibrous root systems. Some culti-
P. coccineus
vated varieties of P. lunatus and P. coccineus are tuber-
forming perennials in the tropics. Although annualism The runner bean specimen from Tehuaean is
earliest
is a feature characteristic of the cultivated species, a single seed from Zone VII of Coxcatlan Cave, a level
there is insufficient evidence to indicate whether this which represents the Santa Maria phase of over 2000
trait arose under domestication or long prior to the years ago. A single runner bean pod from a Palo Blanco
advent of human selection. component of the same cave was the only specimen
Impermeability of seeds functions, ecologically, as a A number of seeds were found
representing that phase.
drv-season dormancy mechanism just as does the inVenta Salada levels of El Riego Cave, East Niche,
tuberous root. The tuberous root and wild-type seed and one seed came from a Venta Salada level of Cox-
can probably be considered as ecological equivalents. catlan Cave.
Because tuber-formation and seed impermeability are The runner bean seeds found in Coxcatlan and El
reduced among cultivated varieties, an ecological dis- Riego caves were probably not grown in the immediate
tinction may be added to the morphological distinc- vicinity of either cave. Conditions in both places are
tions between the wild and domesticated groups. These too arid and hot for this cultigen of cool, pine-and-oak-
reductions in dormancy mechanisms geared to the wet- forested uplands. Cultivation of this species was not
season, dry-season cycle have resulted in a crop plant noted in a brief reconnoiter of the Sierra de Zongolica
that is flexible in adapting to new climatic regimes, above Coxcatlan Cave, but a probably related adven-
home-garden cultivation, or to irrigation. tive species is present in road cuts and other places
where the soil has been disturbed. The vegetation indi-
The Tehuaean Remains
cates that good crops of runner beans could be raised
P. vulgaris in these mountains.
Although the occurrence of single pods of a culti- The late occurrence of runner beans in Tehuaean
vated variety of this species in the Coxcatlan and Abe- suggests that the highlands above Tehuaean are not
jas phases of the Tehuaean sequence demonstrates a region of early domestication of this species. Because
that common beans were known at periods when agri- runner beans thrive under humid conditions that are
culture was still subordinate to plant-collecting, not favorable to the preservation of vegetal remains in
common beans do not become abundant until Palo archaeologicalsites, a search for early runner beans
Blanco times. The earliest pod of was re-

P. vulgaris might well be conducted in arid sites adjacent to a
covered from Zone XI of Coxcatlan Cave, which dates humid upland in central or southern Mexico or Guate-
to almost 6000 years ago. The earliest date (4300-6000 mala.
b.p.) for the presence of pods of P. vulgaris cultivars Runner beans today are a common product of the
in Tamaulipas (Kaplan and MacNeish 1960) is about Indian highlands and are traded to towns in the cen-
the same as that for the Tehuaean Valley. In Tamau- tral depression of Chiapas (field notes, 1957). Evi-
lipas as in Tehuaean, common beans, although present dently they were an item of prehistoric highland-low-
in the excavated remains, did not become abundant land transport in the Tehuaean Valley as well.
until agriculture replaced plant-collecting as the chief
P. lunatus
food source. In Tamaulipas this transition took place
in the Palmillas phase 1100-1800 years ago. The his- The appearance of the few sieva beans recovered
tory of common beans in Tamaulipas resembles the from Venta Salada levels of the Tehuaean sites co-
history of common beans in Tehuaean both in absolute incides with the growth in importance of common
time and in cultural sequence. beans. Their appearance in levels dating to about 1000
A similar sequential pattern, if absolute dates are years ago corresponds with the earliest occurrence of
moved up and time spans compressed, can be discerned the scanty sieva remains of the Ocampo, Tamaulipas,
in the Southwest of the United States (Kaplan 1956). caves, where a few pods are present in the Palmillas
At Tularosa Cave in western New Mexico and in the phase (1100-1S00 b.p.) and subsequent phases. Dr. G.
San Juan and northern Arizona regions, common beans Willys Andrews, excavating at Dzibilchaltun, near
205
Fig. 125. Occurrences of four kinds of bean at archaeological sites in the New World, dated in years b.p. Sources:
Basketmaker II, Kaplan 1956; Chilca, correspondence of F. Engel; Durango, Carlson 1963; Dzibilchaltun, correspondence
of G. Willys Andrews; Fort Ancient, Yarnell 1964; Huaca Prieta, Towle 1961; Nazca, Towle 1961; Ocampo, Kaplan and
MacNeish I960; Oneota and Owasco, Yarnell 1964; Rio Zape, Brooks et al. 1962; Snaketown, Tularosa Cave, Verde Valley,
Kaplan 1956.
Fig. 126. Beans from the Tehuacan Valley. Phaseolus vulgaris: a, Tehuacan 1, Tc 35e, Zone B; b, Tehuacan 11, Tc 50,
Zone VI; c, Tehuacan 13, Tc 35e, Zone C; d, pod, Tc 50, Zone VIII. P. acutifolius var. latifolius: e, Tehuacan 7, Tc 50,
Zone VIII. P. lunatus: f, Tehuacan 8, Tc 50, Zone VII; g, pod, Tc 50, Zone III. P. coccineus: h, Tehuacan 12, Tc 50, Zone
III; i, Tehuacan 9, Tc 35e, Zone C.
207
scribed in a manuscript in preparation) of tepary beans

among the Tehuacan remains casts a new light on the
problem of the origin of this cultigen. Since Freeman’s
recognition (1912) of the botanical distinction between
teparies and common beans there has been consider-
able discussion of the origin of the tepary.
The earliest tepary beans of reliable date in the
Southwest were found in a Basketmaker III site, a.d.
742, near Durango, Colorado (Carlson 1963). There
were no tepary remains in earlier caves in the Anasazi
area, dating from Basketmaker II times (Kidder and
Guernsey 1919), nor were teparies found in late
Cochise sites of more than 2000 years ago (Martin et al.
1952). Thus it appears that the early inhabitants of the
Southwest did not domesticate tepary beans themselves
but probably received the domesticated plants from
another area.
In the sense used here, domestication means a
greater increase in seed size than could be expected
from selection without genetic change. The earliest
Fig. 127. Distribution of tepary beans.
teparies from Coxcatlan Cave have already passed
through this stage of domestication. The seeds are
Merida in Yucatan, recently found abundant charred more than twice the size of those of wild teparies. Other
seeds that I have identified as sieva beans. These beans micro-evolutionary changes are involved in the transi-
date to about 1200-1400 years ago. Earlier beans from tion from wild to cultivated beans, but most of these
this site (2270 ± 80 b.p., correspondence from Dr. cannot be discerned in the seeds and pods of archae-
Andrews) are Canavalia ensiformis. The earliest sievas ological specimens. There is no direct evidence that
from the southwestern United States grew in the Verde increase in seed size was an adaptive change, but be-
Valley about 1000 years ago (Kaplan 1956). Carter cause the formation of large seeds requires a greater
(1951) has noted a protohistoric record probably refer- energy input than does the formation of small seeds,
ring to sieva beans in the eastern United States. it is likely that some advantage accrues where relative
The evidence provided by the widely distributed seed size increases. Because tepary beans first appear
Mesoamerican sites, each excavated with due regard during Abejas times, selective pressures toward in-
for the preservation of vegetal remains, suggests the creased seed size must be referrable to conditions that
entry of domesticated sieva beans into Mexico some- prevailed then —
that is, conditions arising from sub-
time between 1000 and 1400 years ago. Taken with sistence based on food-collecting and incipient agri-
this evidence, the absence of sievas in the Rio Zape culture (MacNeish 1962).
cave in Durango suggests the possibility of an introduc- The use of wild green pods for food may be very old;
tion to Mexico by way of the Gulf coast. The lack of however, no mechanism by which this mode of use
archaeological vegetal materials from the dry coastal would select for large seed size is apparent. The selec-
and lowland regions of Guerrero and Oaxaca, how- tive propagation of large-seeded variants required
ever, precludes the formation of anything more than an selection for such variants. A comparison of seed di-
impression in this regard. mensions of cultivated and noncultivated Phaseolus

reveals that the increase in seed size has been dispro-
P. acutifolius var. latifolius
portionate in favor of cotyledon size. The reserve food
The earliest tepary beans from Tehuacan were a supply of the seeds has increased more rapidly than
batch of 118 recovered from Zone VIII of Coxcatlan has the growing part of the embryo. Stored food in the
Cave. This level dates from the Abejas phase of roughly seeds of vining plants such as beans enables the young
5000 years ago. Another, much larger quantity of plants to grow through a shadowing screen of covering
teparies came from a Palo Blanco level of Coxcatlan plants before their leaves reach full sunlight.
Cave. Small increments in seed size are not likely to be
The identification (based on anatomic characters de- favored or differentially propagated by pre-agricultural
208
Table 24 Characteristics
. of Remains of Phaseolus
No. Median dimensions (cm.)

of seeds length width thickness Color* Comments
P. vulgaris
Variety 1 337 1.2 0.7 0.66 Moderate reddish brown, varies from 10R 3/4 to Resembles bolitas varieties; widely distributed in contemporary
2.5 YR 3/3. Some seeds with dark longitudinal indigenous agriculture in Mexico and southwestern U.S.; present
stripes. in Mesa Verde, Colo, in Pueblo III times. (Variety 1 hilum aver-
ages 0.2 long, 0.15 wide.)
Variety 11 159 1.1-1. 0.6-1.0 0.4-0. Mixed. Black; strong brown 5YR 4/5; moderate Components of this mixture are sufficiently distinct that each
reddish brown 10R 3/4; brownish orange 5YR 5/8. might be ranked as a cultivar. However, the same mixture is
still grown and can be purchased in the Ajalpan market.
Variety 13 2 0.8 0.6 0.5 Pale orange-yellow 7.5YR 9/4. Similar to Southwest type C9 found at Tularosa Cave (Kaplan
1956).
P. acutifolius
var. latifolius
Variety 7 603 0.8 0.6 0.4 Very dark red-brown to brownish orange 5YR 5/8. Teparies having this form but usually white in color have been
Darkened by heating. common in markets of the Pacific slopes of the western moun-
tains of Mexico, among the Hopi, Zuni, and Colorado River
tribes, and, archaeologi cally, in the Verde Valley of Arizona.
P. coccineus
Variety 9 7 1.2 1.0 0.6 Very dark purple, appear black. Occurs in highlands of central and southern Mexico; smaller
than most indigenous runner beans. New to archaeology. (Var-
iety 9 hilum averages 0.4 long, 0.2 wide.)
Variety 12 25 1.5 1.0 0.7 Moderate red brown 7.5R 3/6 to dark red 2.5R 3/7. Resembles some of the contemporary botiles of Chiapas.
P. lunatus
Variety 8 2 1.2 0.9 0.4 Very dark mottle on moderate orange-yellow Resembles widely grown commercial cultivar, Florida
7.5YR 8/8. Speckled Butter. New to archaeology.
•Ref. Nickerson color fan, Anon. 1957.
seed collection. Furthermore, parching, which was M aize first appears in the Tehuacan sequence about
probably the predominant preceramic method of 6500 years ago (see Chapter 9). About 1500 years later,
preparation for eating (Kaplan 1956), may be more domesticated teparies appear. Over a succeeding pe-
effective when applied to small seeds than to large riod of nearly 3000 years in the archaeological record,
ones. The refuse heap, however much a garden spot of these teparies remained stable in seed size, which is
candidates for domestication, is not an automatic within the range of teparies cultivated at the present
domesticator. It is merely a reservoir of genetic mate- time. If domestication of teparies took place in the
rial for selection. In the disturbed, sunny, fertile con- Tehuacan Valley under the circumstances suggested
ditions of the refuse heap, small increments in seed here, the transition in seed size began with the initial
size would have no adaptive value. cultivation of maize some 6500 years ago and was
However, an environment that places a premium completed 1500 years later. This beginning date is at
on the ability to survive shaded conditions during the least 2000 years prior to the earliest record of maize in
early post-germination phase and that repeats its re- the Southwest.
quirements generation after generation is one that The principal difficulty in designating the Tehuacan
would favor an increase in seed size. Maize plantings Valley as the center, or even as one of many centers, of
during the period of incipient agriculture would have domestication of the tepary bean lies in the phytogeog-
provided such an environment. The fidelity of the raphy of the putative wild ancestor (see Fig. 127). The
maize-bean association in Mesoamerica attests to its nearest collection records for P. acutifolius are one
some regions. If being planted with maize
antiquity in thousand miles away in the Jaliscan vegetation area.
—or even adventive growth in maize plantings did — Records for P. acutifolius var. latifolius, believed by
supply the competitive environment in which beans Freeman to have been the immediate ancestor of the
developed large seeds, then the determinate growth cultivated tepary, are in the northern part of the species
form must have come later and not necessarily in the range (Piper 1926). Smith (Chapter 12) reports no wild
same region in which the transition in seed size took Phaseolus species from the Tehuacan Valley and indi-
place. cates that a comparison of the archaeological and
209
present xeric vegetation reveals no appreciable climatic them to be. The form and dimensions of the seed are
changes. similar, as are characteristics of the hilum of scar of
However, an obvious factor in bringing about attachment.
changes in vegetation during the historic period is the
enormous domestic goat population. Their numbers
Summary
in the Tehuacan Valley, and in other highland valleys A total of 1,136 bean seeds and 17 pod fragments
of central Mexico, are astonishing, and their pressure from Coxcatlan and El Riego caves have been identi-
on the vegetationis tremendous. To think of the drastic fied and analyzed. Coxcatlan Cave is the only archae-
reduction range of a plant that is without spines or

in ological site on record in which the four principal culti-
other deterrent to browsing flocks hardly challenges vated American bean species have been found. These
the imagination. species are listed here with the number of varieties
The suggestion have brought about the
that goats present and the earliest estimated date for the species:
extinction of the wild ancestor of the domesticated P. vulgaris (common beans), 3 cultivars, 6000 b.p. ±
tepary in the Tehuacan Valley, however, does not 200; P. coccineus (runner beans), 2 cultivars, ca. 2200
bear directly upon the question of the identity of the b.p.; P. acutifolius var. latifolius (teparies), 1 cultivar,
wild ancestor. Although the geographic origin of the 5000 b.p. ± 200; P. lunatus (sieva beans), 1 cultivar, .500
tepary has been much discussed (Carter 1945), Free- b.p. ± 100.
man’s proposal as to the botanical origin of the tepary Beans as a group are represented in the cave remains
has not been questioned. If teparies did, as the evidence over a span of about 6000 years, from the Coxcatlan
presented here suggests, pass through the domestica- phase to the surface levels which correspond approxi-
tive phase of increased seed size in the Tehuacan mately with the Spanish conquest. Although present
region, then the relationship of the tepary to the wild, in the Coxcatlan phase, beans did not become abun-
broad-leafed P. acutifolius var. latifolius of the South- dant in the remains until about 1000 years ago in Palo
west should be seriously examined. Blanco times. The appearance of common, runner, and
sieva beans is not earlier respectively than the appear-
Contemporary Beans ance of these species in northwestern and northeastern
Beans sold in the markets of Tehuacan and Ajalpan Mexico. Data are lacking for other areas in Mexico.
were collected in August 1962 to determine the extent The common bean pod from Zone XI of Coxcatlan
to which contemporary varieties resemble prehistoric Cave is about 4000 years earlier than common beans in
varieties. Sieva beans, runner beans, and teparies were the southwestern United States.
not available in the markets. In some regions sieva Common bean varieties represented by the remains
beans are grown for household consumption, but they from Tehuacan may have been introduced into the
seldom reach the market stalls. region, or they may have been domesticated there. The
Five varieties of locally grown common beans were beans themselves provide no strong evidence either
collected. These are all non-vining types grown with- way. Runner beans were probably not grown in the
out irrigation in the hills or surrounding mountains. valley, and sieva beans were undoubtedly introduced.
Beans grown in the valley itself are planted on the The finding of teparies at Tehuacan extends the
plowed mesa tops, like that above El Riego Cave. These range of these beans a thousand miles east of the
too are bush types and are not irrigated. The mesa previously known range of tepary cultivation. The
tops were probably used at one time for teparies, Chiapas-western Mexico-Sonoran distribution of culti-
which were replaced by newer introductions of com- vated teparies can now be thought of as a relic dis-
mon beans. tribution. The Tehuacan teparies are about 4000 years
One of the common bean varieties grown in the hills older than the earliest archaeological teparies of the
above Coxcatlan is identical in seed characters with Sonoran Desert region. They appear in the Tehuacan
Tehuacan variety 11 (see Fig. 126 and Table 24). Te- record about 1500 years after the earliest maize. The
huacan 11 and its contemporary counterpart are similar transition from small wild-type seeds to larger seeds
with respect to the ratio of dark-colored to light- characteristic of tepary domesticates could have taken
colored seeds. The actual colors are not identical, but place in the Tehuacan Valley in conjunction with the
because color changes with time, one would not expect domestication of maize.
210
REFERENCES
Albanese, 1959. Kaplan, 1956.
Altsehul, 1962. Kaplan and MacNeish, 1960,
Block and Weiss, 1956. Kidder and Guernsey, 1919.
Brooks, R. H., et al., 1962. Lowe, 1868.
Bukasov, 1930. Mackie, 1943.
Carlson, 1963. MacNeish, 1962.
Carter, 1945a, 1951. Martin et al., 1952.
Chatfield and Adams, 1940. Piper, 1926.

Cooper Clark, 1938. Smith, C. E., 1965b.
Cutler and Whitaker, 1961. Towle, 1961.
Freeman, 1912. Wheat, 1954.
Ellis, 1959. Willis, 1951.
Jones, D. B., et al., 1938. Yarnell, 1964.
CHAPTER 11
Cucurbits from the Tehuacan Caves
Hugh C. Cutler and Thomas W. Whitaker
he remains of both wild and cultivated species gourd family, Cucurbitaeeae (Cutler and Whitaker
T of cucurbits (gourds, squash,

some measure the history of agriculture
record in
and the development and movement of cultures in the

and pumpkins) 1961).
Up most valuable series of
to the present time, the
prehistoric cucurbits are collections from the Ocampo
New World. Although cucurbit material from archae- caves in the state of Tamaulipas in northeastern Mex-
ological excavations is not as abundant as remains of ico (Whitaker, Cutler, and MacNeish 1957). However,
maize, it has four distinct advantages for study over the large collection of almost 1400 specimens excavated
the latter: first, species, and often subspecific groups, from five caves in the Tehuacan Valley is at least as
can usually be identified reliably from seeds or pe- important. The collection includes the oldest known
duncles (fruit stems); second, the five cultivated species specimens of Cucurbita mixta Pang., a species most
of squashes and pumpkins do not hybridize readily variable in Oaxaca and southern Puebla, and also of
among themselves, as do practically all races of maize; C. moschata Poir. Also, there is new material of the
third, specimens of cucurbits from many dated sites third annual cultivated species grown in Mexico, C.
are available for comparison; and finally, pre-Colum- pepo L.
bian distribution of the cultivated species is known with Mexico is the center of greatest diversity of culti-
reasonable accuracy (Cutler and Whitaker 1961). vated squashes and pumpkins. All twenty-six wild and
The common terms “pumpkin,” “squash,” and cultivated species of Cucurbita, except the cultivated
“gourd” are of little value in distinguishing species of South American C. maxima Duch., the closely related
Cucurbit a. The common New England pumpkin, weedy C. andrcana Naud. of Argentina, and C. okee-
White Bush Scallop squash. Zucchini squash, Summer chobecnsis from south-central Florida grow in Mexico
Crookneck squash, and Acorn squash, for instance, are and were growing there in prehistoric times (Cutler and
all cultivars of the same species, C. pepo L., whereas Whitaker 1961). Cultivated and wild species are var-
Kentucky Field pumpkin and Butternut squash are iable, their ranges frequently overlap, and some of the
cultivars of C. moschata Poir. The word “gourd” usually wild species hybridize readily with the cultivated ones.
refers to the bottle gourd, Lagenaria siceraria (Mol.) Although we know a great deal about the taxonomy
Standi. Several tropical trees bear fruits which are often and many more collec-
distribution of the Cucurbita,
used for containers or musical instruments and are tionsmust be made and many plants grown and studied
called gourds. The most common of these is Crescentia before the inter-relationships of wild and cultivated
cujete L., usually called the calabash or gourd tree. species of the genus are reasonably understood.
It is, however, a member of the Bignoniaceae, the fam- The best reports of indigenous, cultivated Cucurbita
ily towhich the catalpa tree and trumpet creeper be- of Mexico and Central America were published by
long.These tropical tree fruits should be called tree Russian botanists after studv of plant collections ob-
gourds to distinguish them from the gourds of the tained from expeditions led by N. I. Vavilov (Bukasov
212
CUCURBITS FROM THE TEHUACAN CAVES
1930; Zhiteneva 1930a, 1930 b\ Vavilov 1931). K. I. Cucurbita mixta Pang.

Pangalo (1930) described C. mixta from Russian col-
lections made in Mexico. The Tehuacan excavations yielded more specimens
Collections of Cucurbita made by Edgar Anderson, of mixta than any other species. Practically all the seeds
Martin Cardenas, Hugh C. Cutler, Lawrence Kaplan, are of a single type, relatively long and narrow, with
Carl O. Sauer, Jonathan Sauer, and Thomas W. Whit- slightly enlarged margins. In some seeds the corky
aker have been grown and studied at the Missouri marginal tissue is silvery gray. The white thickening of
Botanical Garden, St. Louis, or at the United States the seed body is moderate in all seeds, and there is no
Horticultural Field Station, La Jolla, California. A furrowing or cracking as there is in the thickened body
from Guate-
significant collection of cultivated species of typical seeds of the cultivars Japanese Pie and Green
mala was grown in Massachusetts by II. E. Moore, Striped Cushaw. Most of the mixta in northern Mexico
and extensive plantings of seeds from many commer- and the southwestern United States resembles these
cial sources have been made by A. M. Rhodes of the two cultivars. On the other hand, most of the Tehua-
University of Illinois. Whitaker and G. N. Davis (1962) can mixta, archaeological and modern, is similar to the
published a summary of the available knowledge of silver-seeded group of Cutler and Whitaker (1956,
cultivated Cucurbita and included a concise report 1961). Mixta seeds with smooth and thickened brown
of crosses Whitaker made between many cultivated bodies belong to the Taos group (Cutler and Whitaker
and wild species of the genus. The late Dr. L. H. 1961) and were found at Tehuacan only in Coxcatlan
Bailey (1943, 1948) collected and described most of Cave. There were forty-six seeds of the Taos group
the wild species in a series of papers illustrated with in Zone I (Venta Salada phase) and two in Zone VII
photographs and excellent pen drawings. (Santa Maria phase). One seed from Zone XIV (El
Cutler and Whitaker (1961) described and illustrated Riego phase) may belong to this group.
the parts of cucurbit plants used to identify archae- All of the mixta peduncles found in the Tehuacan
ological specimens. Peduncles and seeds usually have caves are moderately corky (Fig. 128). Only a few frag-
reliable characteristics for identification of the culti- ments of rind were found with the corky growths com-
vated species of Cucurbita. The rinds, however, with mon on the upper portions of mixta fruits grown in
a few exceptions, are of little use. By means of their northwestern Mexico (Cutler 1960), the southwestern
cell structure, Cucurbita rinds can be distinguished United States, and in the cultivars Japanese Pie and
from those of the bottle gourd and the tree gourd: but Green Striped Cushaw.
it is presently impossible to identify the species of culti- The oldest known mixta seeds come from Coxcatlan
vated Cucurbita by rinds alone, except for some forms Cave. A poorly preserved seed from Zone XIV of the
of C. mixta that are marked by prominent corky ridges El Riego phase may be mixta. A small seed found in
on the upper portions of the fruit. feces in Zone XIII, dating roughly to about 5000 b.c.,
The remains of cucurbits found in five dry caves in may be mixta. A well-preserved peduncle and
also
the Tehuacan Valley — the East Niche of El Riego four seeds from Zone VIII of the Abejas phase of about
(Tc 35e), Coxcatlan (Tc 50), San Marcos (Tc 254), 3000 b.c. were readily identified as mixta. The latter
Tecorral (Te 255), and Purron (Tc 272) were well pre- — seeds are similar to those sold in Tehuacan today. The
served. It was therefore possible
compare this mate-
to oldest mixta previously known comes from the Ocampo
rial directly with other archaeological and recent caves in northeastern Mexico and dates from a.d. 150
specimens for size, shape, and structure of the to 1050 (Whitaker, Cutler, and MacNeish 1957; Cutler
peduncles, seeds, and rinds. Comparisons are more and Whitaker 1961).
difficult when specimens are carbonized, since partial We expected to find mixta specimens in the older
burning not only destroys some surface features but stratigraphic levels of theTehuacan excavations, since
causes distortion and shrinkage. the center of diversity and relativeabundance of this
Of the four cultivated Cucurbita species known from species is not far from the Tehuacan area in southern
the Tehuacan region today, only C. ficifolia Bouche Puebla and Oaxaca. The kinds of mixta recovered from
was not found in the Tehuacan caves. It is occasionally the Tehuacan caves are still found in the vicinity of
seen in the Tehuacan and Ajalpan markets, where it is Tehuacan and to the south and west; they are not
brought from farms at higher altitudes. As far as we found to the north. Mixta fruits were absent from the
know, it is not grown on the floor of the Tehuacan Tehuacan Valley markets in August 1961 and March
Valley. 1962. At both times, however, abundant supplies of
213
mixta seed were being sold in the markets for roasting,

and street-corner vendors were selling small lots of the
already roasted seed. Fruits of mixta are sold in the
Tehuacan market in late October and November.
A few large seeds of mixta with enlarged silver-gray
corky margins were found in debris on a modern dump
near the town of Tehuacan (Fig. 128), hut no seeds of
this type were found in stores and markets and none
were found in the caves. Mixta seeds from the Tehua-

can caves show less variability than one would antici-
pate, assuming the location is near the center of origin
of the species. The seeds of modern mixta from the
Tehuacan area, however, are remarkably uniform, even
though the fruits may vary greatly in size, shape, and
color.
Cucurbita moschata Poir.

Moschata seeds from the caves are relatively uni-
form and resemble the kinds grown today in and near
the Tehuacan Valley and over much of central and
southern Mexico below an elevation of 5000 feet. None
of the small, hairy-margined seeds common near Vera-
cruz and in most of southern Mexico and Guatemala
were found.
Two seeds found in Coxcatlan Cave at levels attrib-
uted to the Coxcatlan phase (about 4900-3500 b.c.)
were identified as moschata and are the oldest known
seeds of this species. The next oldest specimens,
which date from 3000 to 1000 b.c., are from Huaca
Prieta, Peru. Specimens from the Ocampo caves of Ta-
maulipas are dated about 1850 b.c. (Cutler and Whit-
aker 1961).
Afew roasted moschata seeds were sold by street
vendors in Tehuacan. Kelly and Palerm (1952) col-
lected eultivars of moschata which are grown only for
their seeds in the state of Veracruz, but both seeds
and flesh are used in the Tehuacan region. Women in
the market report that the common way to prepare

moschata is as a sweet, boiled in thick cane syrup.
Squash prepared in this way is sold in Tehuacan and
other markets and along the roadside. It would be in-
teresting to know whether this modern use of squash
is derived from prehistoric times in Mexico. Although
prehistoric Mexicans lacked sugar cane, they could

have obtained sugar by action of an enzyme on starch,
a method well known in Bolivia (Cutler and Cardenas
1947).
Fig. 128. Cucurbita mixta. Top row: seeds, Tc 50, Zone

Cucurbita pepo L.
VIII (70-8). Left center: seed, Taos form, Tc 50, Zone
VII (9-4). Right center: seed, extreme form of Silver- Although all but possibly the earliest pepo seed
seeded Gourd, Tehuacan dump. Bottom: peduncle, Tc 50, found in the Tehuacan Valley are much more recent
Zone III (33-3). All x 1.5. Photos G. A. Sanderson. than the earliest ones from the Ocampo caves, some
214
Fig. 130. Fruits bought in markets near Tehuacan, March

1962. Left front: a cultivar of C. moschata; others: C. pepo.
show close resemblances to the latter group (Fig. 129;

see also Fig. 130, center and right; Whitaker, Cutler
and MacNeish 1957). What may be the oldest seed of
—
pepo a single specimen from Coxcatlan Cave, Zone
XIV, dating from the El Riego phase of about 5000
b.c. is similar to the oldest pepo seeds known, those
from the Infiernillo culture of the Ocampo caves of

about 7000 b.c. It probably came from a wild camp-
follower plant growing near the site on disturbed soil.
The seed from Coxcatlan Cave is slightly smaller than
any of the Ocampo specimens. It approaches in size
and shape the larger seeds of the wild Cucurbita from
later levels of Coxcatlan Cave and those of the wild
species that grows in the region today, probably C.
pcdatifolia Bailey.
Other pepo seeds from the caves, like the mixta
seeds, are relatively longand narrow. The distribution
of long, slender pepo and mixta seeds is not well known,
but they have been collected from farms and markets
south of Guadalajara, in villages near Oaxaca, through
the Tehuacan Valley, and in the vicinity of Veracruz.
The pepo and mixta seeds collected from modern In-
dians or from archaeological sites in northern Mexico
and the southwestern United States are relatively
broad. The presence of seeds of similar shape in dif-
ferent species occupying approximately the same
region suggests the possibility of hybridization. Crosses
between mixta and pepo have been reported (see
Whitaker and Davis 1962: 106).
Pepo fruits are often stored and can be found at
Fig. 129. Cucurbita pepo. Roiv 1: seeds, Tc 35e, Zone A
(32-6). Roiv 2: seeds, Te 35e, Zone C (38-3). Row 3:
almost any time of year. Stored fruits are sold in the
seed, wild form, Te 50, Zone XIV (148-11). Row 4: seeds Tehuacan market in March and April. Large numbers
of fruit from Tehuacan market, March 1962. All x 1.5. of fruits, all with narrow seeds, were seen in Puebla,
Photos G. A. Sanderson. Cordoba, and Oaxaca in October. Lawrence Kaplan
215
(personal correspondence) reports that pepo was the C. pedatifolia Bailey. Other species may be present, but
only species of Cucurbita sold in the Tehuacan market it is usually impossible to identify wild cucurbits by
in August 1962. The same cultivar is grown along the fruits alone.
Puebla-Mexico highway and is stored on rooftops dur-
ing the winter. There is little variability in the fruits: Apodanthera buraeavii Cogn.
those seen in and near Tehuacan were similar to the Apodanthera, the wild coyote melon, is similar to
fruits shown in Fig. 130. wild species of Cucurbita in many respects, although it
Twenty-six staminate flowers, probably of pepo, is in a distant section of the family Cucurbitaceae.
were found in Coxcatlan Cave, Zone III (Venta Salada Most Apodanthera species in Mexico prefer dis-
of the
phase). Similar flowers are sold today in the Tehuacan turbed and are often found in the same habitat as
soils
market. They are usually covered with a meal-and-egg wild Cucurbita species. The seeds are roasted and eaten
batter and are fried or baked. by the Indians today; they have also been recovered
from several archaeological sites. Identification of
Wild Species of Cucurbita Apodanthera species from seeds alone is difficult be-
Fifteen fragments of rinds of a wild Cucurbita spe- cause seeds of a single fruit may vary greatly in size.
cies, probably C. pcdafolia Bailey, were identified from Also, there are few herbarium specimens with mature
Zones D and E to Zone A of the East Niche of El Riego fruits, and the nomenclature of the genus is confused.
Cave. A peduncle came from a late zone of Coxcatlan Specimens from Tehuacan probably are A. buraeavii,
Cave. A portion of a fruit and part of the vine were but A. galeottii Cogn. and A. aspera Cogn. are cited as
found on a plant growing in 1961 at the mouth of El present in the region.
Riego Cave. Green fruits of practically all species of Apodanthera seeds and a few fragments of fruit were
Cucurbita are used throughout Mexico as a laundering found onh/ in the East Niche of El Riego Cave: 11
detergent and for eating when cooked; the mature seeds from Zone A (Fig. 131), 125 from Zone B. and 23
fruits are used for the seeds, which are roasted and from Zone D. These seeds could have been introduced
eaten. Prehistoric Indians probablv used the fruits accidentally by rodents or from fruits which grew near
in identical fashion. There is the possibility, however, the mouth of the cave. Although a few of the seeds bore
that some of the specimens in archaeological sites were the teeth marks of rodents, these marks do not neces-
introduced without the help of man, perhaps even sarily mean that man did not bring the seeds to the
when a site was unoccupied. If so, the most likely cave. Modern food-gatherers often collect caches of
source would be weed plants growing on disturbed seeds assembled by rodents, and it is likely that the
soils of the site, similar to the plant found in front of occupants of El Riego Cave did the same.
El Riego Cave
in 1961. The fact that the wild Cucurbita
specimens from the Tehuacan caves all occur in either Lagenaria siceraria (Mol.) Standi.
late Palo Blanco or Venta Salada levels suggests, how- The oldest known specimens of the white-flowered
ever, that they were introduced by man. bottlegourd were found in the Ocampo caves of Ta-
There is no evidence that any of the cultivated Cu- maulipas in deposits of about 7000 b.c. (Cutler and
curbita of the Tehuacan caves originated from the wild Whitaker 1961). The bottle gourd is the only cultivated
species found in the vicinity. The distribution of the plant which has been found in pre-Columbian deposits
cultivated species and the evidence of their movement in both the Old and the New World. It probably oc-
into various areas suggests that there are distinct points curred wild in both hemispheres before the develop-
of origin for the several species. If any Cucurbita was ment of agriculture. The number of kinds found in the
domesticated in the Tehuacan area, it would most New World is limited, and the center of origin lies in
likely be mixta, because the center of greatest diversity Africa, where more forms and related species grow.
of this species apparently lies within a few hundred The fragments of bottle-gourd rinds excavated from
miles of Tehuacan. There is, however, little diversity in the Tehuacan caves are similar to specimens excavated
the mixta material from the Tehuacan caves and no from other sites in Mexico and the southwestern United
evidence of change in this species during the thousands States. There were no seeds to compare with other
of years covered by the Tehuacan deposits. archaeological specimens.
The wild Cucurbita fruits from the caves resemble a
mature from a collection made near Tehuacan by
fruit Crescentia cujete L.
J.
N. Rose, Joseph H. Painter, and J. S. Rose (their 8994, Fragments of the tree gourd were found in most of
in the United States National Herbarium) identified as the Tehuacan sites. The tree gourd comes from a wild
216
The cucurbit and tree-gourd material from El Riego

and the other caves is summarized in Table 25.
Coxcatlan Cave (Tc 50)

The only pepo specimens from Coxcatlan Cave which
we can pedun-
identify with certainty consist of three
cles from Zone I and one from Zone II (Venta Salada
Fig. 131. Apodanthera seeds from Te 35e, Zone A (14—3), phase). A single seed from the El Riego phase (Zone
x 1.5. Photo G. A. Sanderson. XIV) mav be an early form of pepo, or it could be a
seed from a well-developed fruit of a wild species.
There are scattered moschata specimens from levels
assigned to the Venta Salada phase. Two seeds from
tree, occasionally planted in the Tehuacan region, as in Zones XI and XII of the Coxcatlan phase are probably
other parts of the American tropics. The hard shell of moschata.
the fruit is used to make vessels for eating, drinking, is the dominant cultivated species of this site
Mixta
and storage. The small, hard cells of the rind make it and abundant through the Abejas phase. A mixta
is
easy to distinguish fragments of C. cujete from rinds of seed was identified in a coprolite from Zone XIII of the
the bottle gourd or the cultivated and wild species of Coxcatlan phase. Two seeds from Zone XIV (El Riego
Cucurbita, which have larger and weaker cells (Cutler phase, roughly 5000 b.c.) probably are mixta; if so, they
and Whitaker 1961). are the oldest known specimens of this species. The
oldest prior recorded collections of mixta, dating from
about a.d. 150 to 1050, are from the Ocampo caves.
El Riego Cave, East Niche (Tc 35e)
Far more bottle-gourd specimens in relation to the
Substantially more intact wild cucurbit specimens number of tree-gourd specimens appear in Coxcatlan
were excavated from the East Niche of El Riego Cave Cave than in El Riego Cave. This may indicate that the
than from the other Tehuacan caves. The only Apo- occupants of El Riego Cave relied more heavily on
danthera collections are also from this site. These two gathering food than those who lived in Coxcatlan. On
observations suggest that more wild Cucurbita and the other hand, since El Riego Cave is closer to water,
Apodanthera were growing adjacent to this site than its occupants may not have needed gourd vessels in
near the other sites excavated, or that El Riego Cave which to store it, whereas the people of Coxcatlan
was occupied during the season when these cucurbit Cave did.
fruits were mature, while Coxcatlan and the other
caves were not. San Marcos Cave (Tc 254)

Pepo remains are the most numerous of the cultiva- A pepo peduncle and five rinds of Cucurbita
single
ted squash specimens from El Riego Cave, indicating species from Zone B are among the late materials from
that it may have been occupied mainly during the this site. Five tree-gourd rinds were found in the same
spring and summer when stored pepo was almost the zone, which dates from the Palo Blanco phase. A single
only cucurbit available. Two kinds of pepo seeds are pepo peduncle from Zone C (Ajalpan phase, about
present: slender seeds similar to those common in the 1000 b.c.) is the earliest cultivated pepo specimen found
Tehuacan region today and broader seeds almost iden- in the Tehuacan Valley. The total sample is too small
tical with seed collected farther north in the Ocampo for valid comparisons.
caves of Tamaulipas and from modern fields and mar-
kets in the Ocampo area.
Tecorral Cave (Tc 255)
Some of the earliest narrow-seeded, cultivated pepo A and a fragment of Cu-
single slender mixta seed
from Tehuacan dates from the Palo Blanco period curbita species rind in Zone A (Venta Salada phase)
(about 200 b.c. to a.d. 500) of El Riego Cave. In Ta- are the only cucurbits found in this cave.
maulipas, however, pepo remains appeared in the ear-
liest levels of the Ocampo caves, dating from about
Purron Cave (Tc 272)
7000 A few specimens of the other cultivated Cu-
b.c. Zones C, E, and F (Palo Blanco phase) and Zone G
curbita,moschata and mixta, were found in El Riego (Santa Maria phase) contained the only preserved cu-
Cave. However, they do not appear in the two earliest curbit material at this site: three mixta seeds, a pepo
zones, although pepo remains are present. seed, and three Cucurbita species rind fragments. Five
217
Table 25. Cucurbit Remains from All Sites by Zone and Phase
Riego
Ajalpan
Santa
El Coxcatlan Abejas Maria Palo Blanco Venta Salada
XIV XIII
XII XI
X IX
VIII
C G VII F VI E C V B IV
EF D C A III B II 1 A
254, 272, 272, 272, 272, 254, 35e, 35e, 35e, 255, 35e, 35e,
50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
Totals
Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc Tc
Apodanthera
Seeds only 23 125 11 159
Crescentia cujete
Rinds only 1 5 5 2 3 3 1 3 6 29
Wild Cucurbita sp.
Rinds 2 3 1 5 2 2 15
Peduncles 1 1
C. pepo
Flowers 26 26
Seeds It 1 12 8 1 42 65
Peduncles 1? 1? 1 2 1 3 5 14
Cucurbita sp.
Rinds only 6 2 1 37 19 1 1 1 5 29 2 22 6 1 105 215 253 62 768
C. moschata
Seeds 1? 1 1 1? 1 3 1 26 2 37
Peduncles 1? 1 1 1 4
Lagenaria siceraria
Rinds only 1 2 2 2 2 1 3 1 1 2 9 15 11 19 8 79
C. mixta
Seeds 2? 1* 4 1 3 2 1 1 2 1 1 8 29 100 156
Peduncles 1 7 3 4 2 19 13 17 66
Totals 3 2 1 9 2 4 10 1 2 49 1 27 2 7 2 11 34 16 56 28 2 183 128 351 399 139 1419
‘Found in feces.
fWild form.
tree-gourd fragments were found at the same levels, in the southwestern United States and in adjacent
dating from about 600 b.c. to the beginning of the Mexico.
Christian era. The pattern of persistence of ancient cultivated types
is true to a considerable extent of maize. Some of the
Markets are a good index to the principal products ancient kinds of maize may seldom be found, but they
of a region. Most Mexican markets and those of parts may persist if the areas where they are grown have
of Guatemala and the highlands of South America have been continuously occupied. Even when the population
not changed greatly from pre-Hispanic times, even of an area is replaced by newcomers, the new inhab-
though many plants and products of European origin itants usually take over the cultivated plants of former
have been added to the native wares. In a brief study residents.For example, in the Pueblo Indian area of
of the Tehuacan market in 1962, we found eighty-five New Mexico the dominant maize of early periods is still
species of plants used for food (Whitaker and Cutler to be found, although now it is grown only as one of
1966). Forty-eight are native to the New World, several minor kinds.
thirty-five to the Old World, and two are common to
both hemispheres. In the market we found nearly every Summary
kind of cultivated cucurbit represented in the Tehua- Cucurbit a mixta is represented by more specimens
can caves. Judging from the few reports we have of from the Tehuacan caves than any other species. The
cucurbits grown today in northeastern Mexico, we earliest known mixta seeds are probably those from
could probably find in the markets and fields of this Zone XIV of Coxcatlan Cave dating from about 5000
area the same kinds of cucurbits recovered from the b.c. There is a strong possibility that mixta originated
Ocampo caves. The persistence of ancient cultivated near the Tehuacan region. Several small mixta fruits
types of plants into the present area is well documented collected near Tehuacan had rind and seed characters
218
which suggest that some hybridization with wild spe- than pepo. The greatest diversity of moschata is found
cies is taking place at the present time. Archaeological in the warm regions of coastal and southern Mexico and
and modern specimens of mixta from northern Mexico southward to northern South America. Specimens
and the southwestern United States are relatively uni- from Huaca Prieta, Peru, have been dated at about
form and can usually be classified as belonging to one 3000 b.c., but the Tehuacan specimens are older. The
of three cultivars: Green Striped Cushaw, Japanese Pie, Tehuacan Valley is considerably north of the center of
or Taos. The Tehuacan area is not far from the center diversity for present-day moschata, and it is likely that
of greatest diversity for mixta. earlier moschata will be found in sites farther to the
Specimens of C. pepo are less abundant in Coxcatlan south.

Cave, but pepo is the dominant species throughout the The bottle gourd, Lagenaria siceraria, was excavated
East Niche of El Riego Cave. This observation suggests from some of the oldest levels of the Tehuacan caves,
that El Riego Cave may have been occupied seasonally, along with the cultivated Cucurbita. This finding sug-
when pepo was abundant. There is little diversity in the gests that in the Tehuacan Valley, as in northeastern
pepo of southern Mexico. The species is far more vari- Mexico, L. siceraria was present as early as the oldest
able to the north and is the only squash which spread to cultivated squash or pumpkin and may even have been
the north and east of the southwestern United States present as a weed plant in pre-agricultural times.
in prehistoric times, reaching as far north as south- Specimens of the native wild species of Cucurbita,
eastern Canada. In the Ocampo caves of Tamaulipas Apodanthera (the wild coyote melon), and the bignoni-
pepo was found in the earliest levels, perhaps ante- aceous tree-gourd, Crescentia cujete, were also found in
dating true agriculture, but already showing an increase the caves.
in seed size over most wild species. This observation Nearly all of the different kinds of cultivated and
suggests that pepo may have been a tolerated weed wild cucurbits found in the caves are grown in the area
associate of man prior to the practice of agriculture. today, implying that there has been an essential con-
C. moschata is present in small amounts in both servatism in crop husbandry among the various native
caves, almost as early as mixta and probably earlier cultures that developed in Middle America.
REFERENCES
Bailey, 1943, 1948. Pangalo, 1930.
Bukasov, 1930. Peterson, 1959.
Cutler, 1960. Vavilov, 1931.
Cutler and Cardenas, 1947. Whitaker and Bird, 1949.
Cutler and Whitaker, 1959, 1961. Whitaker and Davis, 1962.
Kellv and Palerm, 1952. Whitaker et al., 1957.
MacNeish, 1961a, 1962. Zhiteneva, 1930a, 1930/?.
219
CHAPTER 12
Plant Remains
C. Earle Smith, Jr.
housands of plant fragments were recovered which there has been a stable climate for the isolation
T from excavations in the Tehuacan region and

cover a span of perhaps 9,000 years, beginning
before 6500 These materials were recovered from a
b.c.
of species particularly
know them in the
adapted
Tehuacan
to conditions as we now
region. According to the
wild plants excavated, there seems to have been no
total of sixty levelsfound in six rock shelters: the East major climatic change over the last 10,000 years, al-
Niche of El Riego Cave (Tc 35e), Coxcatlan Cave (Tc though there is geological evidence that the water
50), San Marcos Cave (Tc 2.54), Tecorral Cave (Tc 255), table has fallen continuously over this period. The
Purron Cave (Tc 272), and Abejas Cave (Tc 307). This archaeological plant remains from the Tehuacan caves
chapter will treat all but the well-known cultivated truly constitute an uninterrupted history of the use of
plants —
corn, squashes, and beans (Chapters 9, 10, and plants and their domestication and continued use over
11)— and plant specimens recovered from coprolites thousands of years.
(Chapter 14). During the earliest periods of occupancy, man was
An important facet of the interpretation of archaeo- largely a predator of the fauna and flora of the valley.
logical plant remains is the relationship between the His activities were so closely tied to natural patterns
plants and the developmental history of the area. Once that he made few alterations detrimental to the vege-
the lake that covered the valley was drained (see above, tation. From the time that he first conceived the idea
Chapter 5), vegetation doubtless rapidly invaded the of cultivating certain plants, man began to make im-
newly available land — an event accomplished long be- portant changes in the plant cover of the Tehuacan
fore man came into the valley. Avenues for invading area. Agriculture in any form implies favoring useful
plants lie along the west side of the Sierra Madre de plants over nonuseful plants. The primary aim of the
Oaxaca and up the Rio Balsas Basin from the Pacific first agriculturists, then, was to encourage certain de-
Coast of Mexico and are reflected in the current vegeta- sirable plants among the local flora. Cultivation can
tion of the valley. A reconnaissance of the flora of the assume many forms, however. In temperate zones, it
valley shows the following relationships: species widely popularly implies large clear areas devoted to a single
distributed in all directions from the area, 50 per cent; crop, or a few kinds of crops. In contrast, the earliest
species derived primarily from the flora of Northern forms of cultivation among the people of the Tehua-
Mexico and Southwestern United States, 14 per cent; can area probably involved only the removal of some
species derived from the vegetation of the Pacific Coast unwanted plants from a small patch of ground to make
of Mexico, 4.5 per cent; species derived from Southern room for a few desirable plants grown from planting
Mexico, 2.5 per cent; species endemic in the general either vegetative parts (cuttings or offshoots) or from
area, including the southeastward extension of the seeds.
valley around Tomellin, 29 per cent (Smith 1965b). Patterns of Indian agriculture in the Americas are
The last remarkably high for a continental
figure, still available for study.The local gardens of the pres-
area, probably reflects the extensive period during ent people of the Tehuacan Valley are good examples
220
PLANT REMAINS
of a kind of agriculture which must have evolved di- maintained in the same places until changes in water
rectly from the first attempts at cultivation and have level or some event such as salting or mineralization
become highly efficient means of food production. forced abandonment. As an irrigation system grew,
Similar gardens in Guatemala have been observed by larger and larger areas would be removed from natural
Edgar Anderson, who suspects that a time study would vegetation. The major deterrent to complete removal
show that such gardens produce more food per man of the vegetational mantle across the valley would be
hour per square foot than the typical, carefully laid abrupt changes in level too great to be overcome by
out and weeded gardens of this country (1952: 141). hand-grading.
Modern household gardens of Coxcatlan village are As Jean Brunet has pointed out in Chapter 5, lower-
examples of Indian cultivation. The areas involved are ing of ground water level followed headward erosion
generally small, not exceeding an acre. Scattered trees, of streams. In a way, irrigation builds into itself a
almost all useful to the owners, include American similar factor for lowering the water table. Where
species— Leucacno esculenta, mesquite, nanche, white formerly all streams in the valley flowed along their
sapote, black sapote, zapote de nino —intermixed with courses, on the surface or beneathit, diversion of water
Old World species — orange, lemon, lime, mango, coffee, from the upper reaches of streams into irrigation sys-
banana, and others. Planted irregularly in openings are tems drastically reduced the volume of water available
various herbaceous plants, including beans, corn, downstream. The ground water level was inexorably
squashes, and taro, the latter usually in a moist spot. reduced as the irrigation systems drew more and more
Herbs supplying flavorings, including chili peppers, water out of natural circulation into and through the
grow here and there. A few animals and poultry may
T
valley. This, in turn, just as drastically sharpened the
be kept, but destructive animals like pigs are fenced hydrostatic gradient in the water table, and upper
away from the plants. All household wastes go into the reaches of the hydrostatic system began to lose water
garden, adding to the organic matter in the soil. The to depleted lower areas, eventually reducing the sur-
garden has no obvious pattern nor is the ground kept face flow into the irrigation channels. If all factors
clear, but only the most necessary weeding is done. could be precisely measured, it would be possible to
In pre-Columbian times such gardens must have forecast the life of any irrigation system in an area of
been planted throughout the Tehuacan Valley. When deficit water balance, because the irrigation system
the Indians cleared space for planting, most useful will eventually destroy itself.
plants would not have been removed. Mesquite and The slow crawl of the irrigation systems downslope
other species of pod-bearing trees, chupandilla, prickly to accommodate thefalling water table was accom-
pear and other edible cacti, and many other species panied by a crawl of irrigation systems upslope to tap
would be left. Among them the farmer would plant streams ever higher up their courses. Fortunately for
such species as he wished to harvest and which would the survival of the native vegetation, all fields served
grow successfully in the chosen area. The success of the by an irrigation system could not be contiguous. Water
planted crop would ultimately depend upon the ar- was fed by gravity and therefore succeeding fields had
rival of rain. Thus, the natural vegetation was never to be at lower elevations. Frequently areas of natural
completely removed, and after a plot was abandoned, vegetation which harbored propagules to recolonize
would eventually reclaim it. abandoned land were left undisturbed.
At its highest development, dry farming during the Soil irrigated by water from mineral springs inevita-
rainy season would not expose very large areas of bare bly becomes mineralized. In years of abundant rainfall,
soil to erosion. Remains of retaining walls across the a certain amount of the deposited mineral may be
beds of barrancas indicate that the Indians recognized leached from the fields, but generally the process is
the higher fertility and higher moisture content of the irreversible. Once many salts have precipitated, par-
alluvial soils of the stream courses. Terracing created ticularly calcium salts, they are very difficult to redis-
in this manner served to spread the water from rain- solve. The gradual crusting of minerals near the surface
storms over wider areas and to prevent serious erosion. of the soil forces the abandonment of fields. Species
The development of irrigation by the Indians prob- which can invade such areas are highly tolerant of
ably has had a far greater effect on the vegetation of minerals, and other species are eliminated.
the valley than the clearing of garden plots. Once an Altogether, agricultural activities in the Tehuacan
irrigation network was established, with considerable Valley, particularly at the height of the Indian popula-
expenditure of energy to dig channels and level plots tion, has brought about the successive removal of native
so the water would spread evenly, gardens would be vegetation over most of the valley surface. Today, areas
221
of seemingly primeval vegetation are all second growth, rainy season and falls during the dry season. Any spe-
and many areas of fertile soil have probably been cies which can survive the normal dry season can
cleared and abandoned several times. Because of the adapt to the gradual changes in ground water supply
mixed nature of Indian farming, however, very few caused by headward stream erosion, deflection of water
species were eliminated from the vegetation prior to for irrigation, or any other cause. Conversely, no plant
the Spanish conquest. species can persist in the Tehuacan Valley unless it
The coming of the Spanish to Mexico marked a dis- can adapt to the seasonal variation in available effective
tinct change in methods of farming, though probably water.

not in systems of controlling water, already highly Any prolonged departure from the current pattern
developed by aboriginal engineers. However, the new- of a five- or six-month dry seasonwould have permitted
comers introduced two immensely destructive forces. a number which are not now present, to in-
of species,
The first was the European system of agriculture based vade the Tehuacan region. Because a certain percent-
on growing a single crop in a cleared field plowed with age of species of any flora are useful to a gathering
the aid of draft animals. The second was the introduc- people, at least one or two fragments of species sugges-
tion of sheep and goats. Herds of these animals swept tive of a climatic difference should have been found
the vegetation from the surface of the land more com- among the plant remains. Actually, because all species
pletely than any other force. In a short time, hillsides have a certain amount of tolerance for adverse condi-
were cleared of their vegetation, which had always tions and because climatic changes are gradual over
tended to absorb and break up the sheets of water from many years, fragments of species reflecting changes
occasional torrential storms. The resulting open soil would have been present over a marked time span.
became susceptible to the destructive forces of sharp, No indication was found in the species composition of
heavy rains. the plant remains to suggest any climatic change.
More damage has probably been done to the native Furthermore, the species composition of the plant re-
and to the hydrostatic
vegetation, to valley land forms, mains from any one site do not suggest any local
system since the Spanish conquest than was done in all changes near the site which can be attributed to
of man’s pre-Hispanic occupation of the valley. Byers changes in ground water level over the entire 9000 year
has pointed out the deficit water balance existing in the period.
area (see above, Chapter 4). Under completely undis- Inview of the above categorical statements, a dis-
turbed conditions, the plants and the animals can reach cussion of vegetational changes may seem anomalous.
an equilibrium which conserves as much of the avail- In point of fact, it is not. Vegetational changes occurred
able water as possible. The Indians disturbed this in the Tehuacan area and are still occurring. The be-
equilibrium during their development of agriculture, ginnings of agriculture can never be known with exact-
but the balance was toward the negative
tilted gently ness. Probably the first areas to be planted and culti-
side. by the Spanish tilted the
Practices introduced vated were plots of alluvial soil newly deposited by a
balance sharply toward the negative side and acceler- heavy storm or those partially cleared of vegetation by
ated the process of destruction. the scouring action of running water. Today and in the
While it is true that the pattern of vegetation on the past the most likelv such areas are in the barrancas
Tehuacan
surface of the land has been changing in the along the northeastern side of the valley. Soils here are
Valley, the changes due to disturbance of the ground better watered during the rainy season, increasing the
water level and the hydrostatic system have been chances of successfully raising a crop. It was along the
changes in degree rather than changes in kind. Plants upper course of these barrancas that the taller montane
native to an area persist there because they have with- forest most closely approached the valley floor. The
in their genetic constitution the plasticity to accommo- early avocado fruit represented by cotyledons in Cox-
date to the fluctuating factors in their environment. For catlan Cave may have been gathered from trees grow-
example, a heavy growth on the fertile soil along a ing in the barranca forest. Similarly, the remains of
water course will slowly move to a more favorable pepper recovered from early levels of the cave
chili
environment if the ground water level persistently falls. may have come from plants growing at the margin of
Pressures of selection occasioned by the loss of ground the forest. Since the products of both plants are highly
water will eliminate individual plants which cannot desirable, they probably were particularly sought out
adapt. by occupants of the cave.
Every year, as regular as the weather cycles, the The other vegetational change in the Tehuacan
water level in the Tehuacan region rises during the region has occurred since the Spanish occupation. Dur-
222
PLANT REMAINS
ing the rainy season, most parts of the area receive tions. All of the species recovered from the excavations
enough precipitation to encourage grasses and herbace- were wild at some point in their ancestry. If there are
ous plants which grow, mature their seed, and die with- plants among the remains which are not now or never
in a single season. Open spots probably once became could have been native to the area under the climate
green with low herbaceous plants during the rainy which is believed to have prevailed over the past 9000
season. Masses of grass remains excavated from caves years, then they must have been under cultivation and
suggest that moist areas along the stream courses sup- are, therefore, cultivated plants.
ported stands of setaria and other grasses. In modern However, the term “cultivation” is relative. Many
times herds of sheep and goats have radically changed plant species native to the Tehuacan area are both
this. These animals crop all non-poisonous herbaceous wild and cultivated today. Mesquite is planted in yards
vegetation to the roots before it can mature seeds. The and border rows, where it is cultivated, but it is also
soil in many areas now remains essentially bare. wild in the thorn-scrub-cactus forest. Whenever an
The presence of setaria and other grasses in the cave Indian farmer cleared a small plot but saved the useful
deposits is not indicative of a persistently high water species and encouraged them for his own use along
table at any time in the past. It merely reflects the lack with the species which he planted, he was cultivating
of herds of animals to deplete the vegetation. Annual the plants. If he purposely planted the seed of a native
plantsdepend almost exclusively upon the moisture species, the stem joint of a prickly pear, or the offshoot
from precipitation to germinate, grow, and mature. of a maguey plant and gave it cursory care to eliminate
Most species have a delayed germination mechanism competing vegetation, he was cultivating the plants.
whereby not all seeds present in the soil will germinate Many of the native species do not show a marked in-
at any one time. If weather conditions cause a crop to crease in size or changes in morphology when grown
fail one year, sufficient seed persists to enable the under cultivation, and it is impossible to determine
species to grow and mature another, more favorable whether fragments representing these species may
year. Although cultivation, carried on in the vicinity of have come from cultivated plants. Therefore, we can-
all of the excavation sites, has removed the original not say exactly when such plants may first have been
plant cover at least once, the complex of species rep- cultivated.
resenting the native vegetation has largely managed to “Domesticated” implies far more than “cultivated.”
repopulate abandoned areas. It is fortunate that this Maize, for example, was cultivated before it was do-
is so, since an understanding of the recovered plant mesticated. Once the Indian farmer began actively to
material hinges on a knowledge of the local ecological select for certain qualities in a plant species, he began
situation. the process of domestication. In the case of maize,
Wemay assume that a major part of the plant ma- selection proceeded fairly rapidly, the result being a
terial was collected within a radius of twenty miles of a form of maize which could not propagate itself because
site. Locally available plants were certainly utilized, it lacked a means to disperse its seeds. It had to be cul-
but desirable plants known to grow a day or more’s tivated and was, therefore, completely domesticated.
journey away were sought out when needed. The pres- On the basis of fragments of plants, domestication is
ence in Coxcatlan Cave on the eastern side of the valley more difficult to prove. However, if recovered frag-
of fragments of “sotolin” Beaucarnea gracilis) and
( ments exhibit morphological changes due to active
palm ( Brahea dulcis which grow on basic soils on the
)
selection, largely increase in size, the plant may be
western side of the valley attest to this, as may the pres- said to be domesticated.
ence in El Riego Cave of pochote, which now grows on The value of permanent water sources in the valley
acidic soils of the northeastern side of the valley. Prob- must have been clear to the early agriculturists. Once
ably only the most precious commodities would be the idea of assembling useful species in areas con-
traded over a considerable distance. Even today, the venient to habitation sites became established, prob-
population of the Tehuacan Valley supplements the ably initially through the vegetative propagation of
tortilla and bean diet with wild plant foods collected in such easily cultivable species as prickly pears and
season. maguey, the limitation imposed by the loss of vegeta-
A
major problem in the discussion of the plant re- tion during the dry season would be apparent. Initial
mains revolves around the use of terms denoting the use of spring water probably was no more than the
conditions under which plants grow. The easiest of the planting of crop plants along the margins of permanent
words to define is “wild,” which refers to those plants streams. Even this is a form of irrigation, although it
that grow untended wherever they find suitable condi- consists of bringing the crop to the water instead of
supplying the water to the crop. By elementary ditch- interest here, because it was in this zone that aborig-
ing to divert small amounts of water from a stream, the inal people developed agriculture. Several facies of this
early farmers could grow’ increasing amounts of food forest occur according to the combination of altitude,
during the season when plant food was not plentiful and drainage. The umbrella-shaped
precipitation, soils,
in the thorn-scrub-caetus vegetation of the valley. crowns of most of the trees reach a height of six to
The introduction into cultivation and selection for eight meters, and their branching is frequently tortuous
larger fruit size in avocado indicate that the valley and intertwined. Shrubby vegetation rarely forms a
farmers were skilled in the use of irrigation. Examples continuous thicket among the trees. Several species of
of the vegetation found in the caves suggest that the maguey Agave (
and prickly pear Opuntia spp.)
spp.) (
climate has always been too drv to allow avocados to are to be found among the shrubs, but the most striking
grow naturally near tbe sites. By the Santa Maria phase feature of this forest is provided by the arborescent
(about 900-200 b.c.), the increase in size of avocado cacti, some of which reach a height of ten meters.
cotyledons provides unmistakable evidence that the These include the columnar Cephalocereus hoppen-
tree was not only in cultivation, but that it had been stedtii, the irregularly branching jiotilla ( Escontria
actively selected for increase in fruit size sufficiently chiotilla), the stiffly upright stems of garambullo (
Myr
long to prevent the large-fruited forms from being tiUocactus geometrizans), and the spectacular gigantic
completely swamped by the wild-type, small-fruited candelabra of the cardon (
Lemaireocereus iveberi).
forms in the system of open pollenation which must Most of these bear fruits which, like those of the opun-
have prevailed. Since it takes about seven years for a tias, are known as tunas. Among the trees of this forest
seedling avocado to come into bearing and trees con- is one of outstanding economic value, the pochote
tinue to bear for seventy years or longer, the period of (Ceiba parvifolia). This relative of the giant silk-cotton
selection prior to its actual documentation by larger tree produces large fleshy pods which later open into
cotyledons must have been extensive. Avocados must puffs of white fiber, among which are small black seeds.
have been grown under irrigation during the period of These are ultilized for food, not only as green, unripe
development and probably for some time before that. pods, but also as ripe seeds. The roots of the tree are
Irrigation can mean the use of water in many ways. enlarged, fleshy, and rich in starch, and they have
Probably the most primitive form was the planting of served as food for millennia. Other trees in the forest
crops at streamside so that short ditches could supply produce edible pods and seeds, and the magueys and
water during the dry season. The presence, near Purron prickly pears produce fleshy leaves or stems which were
Cave, of a dam whose earliest structure dates from used as food.
600-800 b.c., indicates that irrigation had progressed Two rather different aspects of this forest can be
far beyond the rudimentary stage at this time, for this seen, the one on the northeastern side of the valley
large structure would have furnished water for a siz- where reddish soils are derived from the prevailing
able area. Remains of old irrigation channels are evi- acidic rocks composing the Sierra Madre de Oaxaca,
dent from the mineral deposit which accumulated the other on the southwestern side of the valley where
from the water flowing in them. Most of these are diffi- the substrate consists of travertines, limestones, and
cult to date, but many date from pre-Conquest times. marls that allow rapid percolation of moisture and that
break down into prevailingly alkaline soils. In some
Ecology of the Excavation Sites localities water from saline springs has produced an
At an altitude of about 1800 meters, the vegetation environment suitable only to halophytes, species that
and lower slopes of the Te-
characteristic of the floor will tolerate salt.
huacan Valley gives way to oak and pine forest. Above In the principal barrancas draining the Sierra Madre
about 2500 meters, in an area of frequent and some- there is a more ample supply of water than found in
is
times persistent rain, the oak-pine forest changes to the surrounding terrain. Although this water cannot
montane rain forest. Although many of the plants grow- always be seen in the arroyo within the barranca, a
ing above 1800 meters would be valuable sources of considerable underflow exists, as is demonstrated by
food and raw materials, the archaeological plant re- the several methods of tapping it now employed by the
mains include no indication that the early inhabitants barranca-dwellers. If a “barranca forest” formerly ex-
of the valley regularly frequented the mountain slopes. isted along the lower reaches of these watercourses, it
Below 1800 meters a forest of spiny shrubs and cacti is difficult of demonstration within the area under con-
prevails. This thorn-scrub-cactus forest is of primary sideration because the sides of the barrancas are in-
PLANT REMAINS
tensively farmed by people who grow every available of plant species and larger amounts of remains were
fruiting tree, so that no trace remains to show what may preserved in Zones D and C. Remains of maize cobs, —
have been the aboriginal barranca forest. husks, and loose kernels, numbering altogether in the
—
thousands in some levels make up the largest portion
El Riego Cave of the remains. Leaf fragments and chewed quids of
El Riego Cave, a cavity in the encrusted travertine maguey are also consistently present in huge amounts.
covering the face of Cerro de la Mesa, is located about The hundreds of common beans in the upper four zones
four kilometers west of Tehuacan and near the present are notable, as is the greater proportion of Cucurbita
site of the spring called El Riego. It comprises two })epo among the cucurbits from the five zones. Rinds of
niches, an eastern and a western, which bear little tree gourds and bottle gourds are also present. The un-
stratigraphic relation to each other and are treated as usually large numbers of white sapote remains in Zone
two sites, Tc 35e, and Tc 35w. Recause of its situation, C in relation to remains of other fruits should be noted.
water has always been abundant near the cave, al- This pattern is not reflected in fruit remains from other
though it lies within the zone of least precipitation in sites. All the popular fruits of the valley are represented
the Tehuacan Valley. This water has been used for in two or more levels: avocado, chupandilla, cosahuico,
irrigation purposes since pre-Hispanic times, and al- ciruela, coyol,and black sapote.
though the water table appears to have dropped great- Besides cotton and maguey, which are present in all
ly, it is still a source of water for various purposes. levels, a number of plants utilized for fiber appear
Abundant supplies of water create oasis conditions among the remains of Zones D and C: Brahea dulcis
at the junction between the Tehua-
alluvial soil of the and Bcaucarnea gracilis, both of which grow primarily
can Valley and Cerro de Mesa, thus supporting an
la on limestone areas on the west side of the valley; Till-
atypically lush vegetation on the valley floor. Trees now arulsia spp.; the vine Cissus; and Yucca periculosa
growing in this zone include mesquite Prosopis juli- (
(izote). Among the edible cacti, prickly pear remains
flora), ciruela ( Spondias mombin), and various pod- are most prominent, but teteeho (
Cephalocereus hop-
bearing trees locally called guaje. The valley floor is ir- penstedtii) and cardon ( Lemaireocereus weberi ) are
rigated and intensively cultivated. Because of the culti- present in Zones D and C. Zone C also contained one
vation it is impossible to determine the original flora of the few' recovered seeds of the poisonous Thevetia
of the area, but it seems likely that it included trees and peruviana. Pochote (
Ceiba parvifolia) remains, either
shrubs of economic value that are not now normally pods or roots, appeared only in the three earliest levels,
found on the west side of the valley. and mesquite pods or quids were present in Zones A,
A few meters above the valley floor the face of Cerro C, and D. Fragments of chili pepper occurred in each
de la Mesa supports the succulents, agaves, and xero- of the Venta Salada levels, and the grains, amaranth
phytes of the southwestern side of the valley. A small and S etaria, w ere recovered from a few levels. The
;
barranca leads past the mouth of the cave and affords plant remains from each excavated component are re-
access to the nearly level surface of La Mesa. It seems corded in Table 26.
unlikely that it carries water except at times of ex- The appearance of certain plants among the remains
tremely heavy rain, because the travertine substrate tells us something about the seasons during which a
allows normal rainfall to penetrate readilv. There is sitewas occupied. To begin at the bottom of the deposit
now insufficient soil in the barranca to make cultiva- in the East Niche of El Riego Cave, the presence in
tion feasible, and its vegetational cover consists of Zone E of two pochote pods may indicate a spring visit,
cacti and spiny shrubs with only rarely a tree of low and the more abundant remains of chupandilla together
stature. Prickly pears, magueys, and a species of with two seeds of white sapote, which ripen in the
Hcchtia are much in evidence. Before the introduction autumn months, show occupancy in the fall. Since
of sheep and goats by the Spaniards, the floor of the numerous fragments of cultivated plants are present,
barranca was probably covered by a thin grass sward the occupants of Zone E w ere probably harvesting or
7
during the rainy season. planting fields during the times they lived in the cave.
The two lowest levels of the East Niche of El Riego The pattern revealed by the plants from Zone D is
Cave, Zones D
and E, represent the Palo Blanco phase similar. The presence of pochote pods, teteeho, and
and the upper three levels, Zones A-C, the Venta Sa- cardon indicate springtime occupancy, and seeds of
lada phase. The plant remains from each of the five w'hite sapote, plusone black sapote seed, show resi- 7
levels show a similar pattern, although a greater variety 7

dence in the fall.
225
Fig. 132. Tecorral Canyon; view from San Marcos Cave.
The plant remains in the three Venta Salada levels to the optimism of some bvgone farmer or to a once
lack most of the springtime indicators, but at this greater supply of water than is now available.
period, with irrigation well established, the cave may In the vicinity of the caves the vegetation has not
well have been occupied throughout the year. Cer- been recently disturbed except by grazing herds of
tainly large crops of com were being raised, and beans, sheep and goats, but a few hundred yards downstream
pumpkins ( Cucurbita pepo), squash, and cotton were someone cleared the sparse brush and lechuguilla from
probably planted along with the corn. Fruit trees were a small area. The hillside facing the caves is striped by
probably planted along irrigation ditches or in watered tall columns of arborescent cacti, among which Cephal-
fields. ocereus hoppensiedtii predominates. Darker tufts of

the leaves of the izote, prickly pear, and scattered
San Marcos and Tecorral Caves barrel cacti appear among the lighter columns of te-
These two sites are rockshelters in the southwest wall techo. An occasional bulbous-trunked sotolin ( Beau
of the headward reaches of Tecorral Canyon. They lie carnea gracilis ) projects above the spiny shrubs, but this
close together, about eight kilometers south-southeast tree is by no means common in this vicinity. The vege-
of El Riego Cave, and somewhat more than seven and tation in the upper parts of the canyon, dominated bv
a half kilometers from the Plaza of Telmacan, in one spiny shrubs, lacks some of the members of this associa-
of the most arid parts of the valley. Here the scant rain tionprominent on the northeastern side of the valley.
falls on porous sediments and sinks rapidly into the Herbaceous plants are few, since they succumb to the
ground. A small watercourse in the canyon shows that grazing animals. The vegetation appears to have
it supports an ephemeral stream, probably only during reached maturity, but it is largely second growth, the
times of very heavy rain, and remains of a very simple canyon floor and the less-steep slopes having once been
irrigation ditch below San Marcos Cave testify either cultivated.
226
PLANT REMAINS
No form of cultivation is undertaken in Tecorral Zone B. Tetecho, also appearing in two levels, is the
Canyon today. The small clearing mentioned above only other cactus represented.
looks as though it were an attempt to develop a small Besides maize, plants that might be attributable to
patch of com with dry-farming techniques. By means cultivation are not numerous and appear only in later
of deep, contour furrows and seed planted in holes in levels. Scraps of cotton fiber were recovered from
the furrow, the farmers of the dry, western side of the Zones D (Abejas phase), C (Ajalpan phase), and B (Palo
Tehuacan Valley manage to harvest a crop during the Blanco phase). The amaranth Zones B and C was in
rainy season. Both maize and beans are seeded in a probably derived from cultivated plants, and the se-
mixed planting. The cultivars grown by this technique taria seed from Zone C may have been cultivated. No
1
appear to stand without growth during periods of beans were recovered, and cucurbit fragments are few.
drought with little harmful effect. With the advent of Remains of fruit are scant, consisting only of one chu-
rain, rapid growth is resumed. Vestiges of old check pandilla seed, one cosahuico seed, three white sapote
dams and terrace walls demonstrate that the canyon seeds, and six seeds of Bumelia laetevirens— all from
floor was once cultivated by dry-farming techniques. Zone B. The only samples ground cherry, or “tomate"
of
There is at present no source of irrigation water up- ( Physalis sp.), recovered from any of the sites were 28
grade in the area. However, the presence of subsurface fragments in Zone C 1

and 25 in Zone B. These may or
water is attested to by the remains of an old tunnel may not come from cultivated plants.
system and fresh spoil-piles from a recently improved The plant remains from San Marcos Cave do not give
irrigation tunnel leading into the mouth of the canyon. as much evidence as we could wish about the seasonal
That part of Tecorral Canyon described by Flannery occupation of the sites, for the fall-ripening fruits are
(Chapter 8) lies in this lower part. absent from all but Zone B. However, the presence of
The plant remains from Tecorral Cave (Te 255) are pochote pods from Zone E (Coxcatlan phase) on indi-
few and appear late in the sequence of cultural devel- cates that the cave was visited during the spring, prob-
opment. Only Zone A (Venta Salada phase), contained ably when the fields were planted. Caches of mesquite
preserved plant fragments, including remains of cotton, seed could have been from the spring visits or may
left
squash, and maize. indicate summer crops of maize were grown

visits. If
San Marcos Cave (Tc 254) contained a deposit ex- without irrigation, they would have been harvested by
tending from Zone F (Coxcatlan phase) to Zone B groups living in the cave during the autumn.
(Palo Blanco phase). Plant remains were not very
abundant in the earlier levels, but extensive remains of Abejas and Purron Caves
maize and maguey show an unbroken sequence over These two caves in the Arroyo Lencho Diego are
the entire period covered by the deposit. The maize located on the eastern side of the Rio Salado, where the
further shows evolution from wild to domesticated reddish soil, primarily derived from acidic rocks, sup-
forms (Chapter 9). ports a fairlv uniform thorn-scrub-cactus vegetation.
Much of the plant material is representative of the Coxcatlan Cave also lies within this area. Abejas Cave
natural vegetation adjacent to the cave. Several species (Tc 307) lies in the southwestern side of the canyon,
of “guaje” (a generic term for pods of leguminous trees where a small branch to the northwestward leaves the
of the genera Acacia , Leucaena, Mimosa, and Pitheco- main canyon. Purron Cave (Tc 272) is only a short dis-
lohium) were present from Zone E (Coxcatlan phase) tance away, below a steep cliff on the eastern side of
on. A number of Jatroplia ncopauciflora seeds were this small branch and very close to the facing canyon
preserved in Zone B. Among the local fiber plants, wall. It must have been well sheltered from inclement
maguey, Bralica dulcis, Hechtia sp., Tillandsia spp. weather, and it is well heated by the afternoon sun.
(including T. dasilyriifolia), Beaucarnea gracilis, Yucca The canyon is now covered with spiny shrubs and small
periculosa, and Cissus sp. were represented in one or trees, punctuated by scattered columnar cacti, Cephal-
two of the Palo Blanco levels, and seeds of Beaucarnea ocereus hoppenstedtii. A large dam of earth and ma-
and yucca were found in Zone C (Ajalpan phase). From sonry was built across the arroyo at Mequitongo in
Zone E on, pochote pods are present, attesting to pre-Hispanic time, and this has materially altered the
springtime occupancy. (This plant was not observed in floor of the canyon by reason of sedimentation above it.
the vicinity of Tecorral Canyon in the survey.) Remains The unnaturally smooth surface of the ground is cov-
of mesquite — quids or seeds —are fairly numerous in ered with a growth of eryptogamous plants. In the
the last three levels, but stem fragments of Opuntia better-watered bottom of the ravine, the cover of trees
appear only twice, in Zone D (Abejas phase) and in of the families Leguminosae, Burseraceae, Anacardi-
227
cobs, one cosahuico and one coyol seed, a nopal ( Opun -

tia stem) fragment, and two maguey fragments. The
meager evidence hints at a summer or fall occupation.
Zones I-G represent the Santa Maria phase of Purron
Cave.The numbers of specimens and the species rep-
resented begin to increase in these levels. Remains of
maize, for instance, number 10 in Zone I, 12 in Zone H,
and 30 in Zone G. Maguey leaf fragments increase from
6 in Zone I to a startling 863 in Zone H, to 146 in Zone
Zone 1 Zone H Zone G

Avocado 0 1 5
Chupandilla 3 4 26
Cosahuico 3 1 194
Covol 4 0 7
White sapote 0 1 4
Fig. 133. Purron Cave as seen from the opposite side of Black sapote 0 1 0
the canvon.
G, and maguey quids increase from one to 42 over the
three levels. Fruit remains follow the same trend, and
aceae, and others is interspersed with arboreal cacti some species appear for the first time, as the accom-
and spiny shrubs of many As the ground be-
families. panying tabulation shows. The large earthen dam men-
gins to slope upward along the margins of the canyon, tioned earlier was apparently builtsome time between
the trees are more widely scattered. Where increasing the occupations of Zones I and II. The effects of its
slope allows more drainage, the arborescent cacti be- presence may be reflected by increases in the plant re-
come dominant, and the tree cover is sparse. Maguey mains in Zone II and later levels. Note that remains of
plants are usually found up the slopes on the better- black and white sapotes appear in Zone H, after the
drained areas. Among the native plants are many that dam would have provided water for irrigation. The
were gathered by the occupants of the caves —includ- absence of black sapote remains in all but one of these
ing pochote, guaje, prickly pear and other cacti, chu- later levels is surprising and is probably owing to poor
pandilla, cholulo,maguey, and the various grasses. preservation, since remains from Coxcatlan and El Riego
The excavation of Abejas Cave produced very few r
caves demonstrate this fruit's popularity.
plant remains. A small quantity of maize appeared in Among was
the fiber plants besides maguey, cotton
one level of the Palo Blanco phase. In Zone B three identified in and all subsequent levels, and
Zone II
maguey quids and ten fragments of maguey leaves Beaucarnea gracilis and yucca were present in Zone II
were found. alone. Annual crop plants in addition to maize and
Purron Cave, on the other hand, though preservation cotton include squash ( Cucurhita mixta and unidenti-
was spotty, yielded a variety of fragments, some few of fied Cucurbita sp. rinds).
which extend back to the El Riego phase of 6500- Since the plant remains in all three levels contained
5000 b.c. In Zone R two seeds of the fruit chupandilla both spring- and fall-ripening species, the pattern of
were preserved, along with one cosahuico seed and occupation was either throughout the growing season
two fragments of maguey leaves. or during spring and fall visits to plant and harvest
Preservation was poor through the next several levels crops.
(Q and Q of the Coxcatlan phase and N’-L of the

1
The Palo Blanco phase of Purron Cave is represented
Abejas phase). The only vegetal remains were frag- by Zones F through A. The trend of increasing quan-
ments of maguey and bits of wood and bark. tities in some cases is maintained throughout the six
The zones representing the Purron phase, K and K 1 ,

zones. Remains of maize, for instance, number 40 in
contained a few maize cobs, a few chupandilla seeds, Zone F and 198 in Zone A. The number of maguey
one coyol seed, two maguey fragments, two pochote leaves rises from a low of 193 in Zone F to a high of
pod fragments, and the usual wood and bark. These re- 1,396 in Zone B and drops in Zone A to 229. Zones E and
mains may represent spring and fall visits, or perhaps C also contained over a thousand leaf fragments each.
occupation throughout the rainy season. Most of the fruit remains follow a somewhat different
Zone J of
the Ajalpan phase contained twelve maize trend. After increasing to large —
amounts sometimes
228
PLANT REMAINS
Fig. 134. Coxcatlan Cave during tlie drv season. The deciduous nature of the vegetation is clearly seen. The man standing
on the talus demonstrates the scale of the shelter which extends across the face.
Zone Zone Zone Zone Zone Zone rron Cave and stems in Zone E
are the remains of fruit
F E D C B A and of stems in Zone D. Other naturally occurring
Avocado 0 0 6 0 0 0 species represented in these levels are Jatropha neo-
Chupandilla 48 923 249 61 54 124 pauciflora, pochote (pods and roots), yucca, tetecho,
Cosahuico 92 30 33 0 1 4 garambullo, and Mammillaria sp. Six mesquite quids
Coyol 37 68 68 18 0 1 from Zone B were the only specimens of this plant
White sapote 3 4 3 0 0 0 found in the upper levels of Purron Cave. The only
Black sapote 0 0 0 0 0 0 other mesquite remains were eight quids from Zone J.
Guava 0 4 0 0 0 0
The remains from the Palo Blanco zones show the
same seasonal pattern as the previous zones. Pochote
numbering hundreds (see accompanying tabula-
in the pod fragments indicate spring occupancy and the fall-
tion) — in Zones F
D, they then appear in diminishing
to ripening fruits are evidence of occupation in the fall.
quantities or not at all from Zones C to A. The fact The plant remains from Purron Cave show an un-
that the large dam in the area apparently no longer usually continuous pattern from the Purron phase
functioned after Zone D may account for the decreases through the Palo Blanco phase. Remains of plants at-
in the later levels. Of particular interest is the finding tributable to cultivation in every level are proportion-
of four guava fragments in Zone E —the only finds of ally larger than remains of wild plants —assuming that
this definitely cultivated fruit in the plant remains from most of the large amounts of maguey remains came
all the sites. from a permanent planting. This planting could have
Among the fragments of Opuntia recovered from Pu- covered an extensive area and would have been cut
229
Fitr.
O 135. View from Coxcatlan Cave showing
O the forest of thorax shrubs and cactus in the drx season. A forest of oak
and pine originallx extended to the lower edge of the montane fields. See Fig. 93 for view of the same countrx- in the wet
season.
annually, perhaps largely for the extraction of fiber. seed size noted for some of these plants. One should
Crops of cotton, maize, and squash were planted and note as well the abundance of most of these fruits in
harvested yearly. The area also seems to have contained the Santa Maria and Palo Blanco levels of Coxcatlan
orchards where such trees as coyol, avocado, ciruela, Cave (see Table 26).
chupandilla, cosahuico, and white sapote were grown
under The Coxcatlan Cave
irrigation. largest increases in remains of
these fruits occur during the late Santa Maria and earlv By far the largest amount of plant remains comes
Palo Blanco occupations —precisely when the archae- from twenty-four of the twenty-eight levels of Coxcat-
ological evidence shows that a dam provided standing lan Cave. This large rock shelter in a northward-facing
water in the area. This dam ceased to function some cliff looks out over the gently rolling slopes of the
time between the occupations of Zones D and C (Palo Sierra Madre de Oaxaca. Directly in front of the cave
Blanco phase). It is interesting to note that whereas a watercourse carries water only after heavy rains and
the amounts of recovered maize cobs increase in the shortly joins similar streams. This is an area of shallow-
later Palo Blanco zones and the amounts of maguey valleys and gentle hills adaptable to cultivation. That
remain large, the fruits —most of which had be to ir- the area was once extensively farmed is demonstrated
rigated to survive —consistently diminish after Zone D. by house mounds and foundations scattered through
the last occupation when the dam was functioning, and the rich stand of thorn-scrub-cactus forest that now-
further, they are almost nonexistent before Zones H
and covers the land. The native wild plants found in Cox-
I, when the dam was built. This is strong evidence for catlan Cave are today available in the nearby vegeta-
concluding that avocado, ciruela, cosahuico, coyol, and tion and certainly must have been available to the ab-
chupandilla were grown under cultivation and is in original inhabitants of the cave.
agreement with conclusions based on the increased In the valley bottoms and on the gentle slopes a thin
230
PLANT REMAINS
locactus geometrizans frequently have rows of tiny

glaucous along the crests of the ridges. The fuzzy
fruit
columns of Cephalocereus spp. and the huge candela-

bras of Lemaireocereus iveberi dominate some of the
better drained slopes.
Growing among the trees are numerous shrubs,
usually with intertwined spiny twigs, such as Castela
and J. neopauci-
tortuosa, Bursera arida, Jatropha dioica,
fiora. Large clumps of Opuntia hyptiacantha and O.
pilifera bear numerous edible, red prickly pears or
tunas. On the well-drained slopes above the cave and
on hills nearby, Agave rnacroacantha forms colonies of
rosettes, but the stalked leehuguilla, Agave karwinskii ,
and the Agave rubescens occur as solitary plants.

sessile
Scattered through the trees and shrubs are a surpris-

ing variety of herbaceous plants. During the rainy sea-
son numerous composites like Sanvitalia frnticosa
Hemsh, Tridax procumbens L., and Pectis canescens
HBK. intermingle with fragile plants of Oxalis neaei
DC., O. berlandieri Torr., Commelina erecta L., and
C. dianthefolia Del. After a rain, the resurrection plant,
Salaginella lepidophylla Spring., lies flat to expose its
bright green top, but it remains rolled into a tight

brownish ball when the weather is dry. Tillandsia re-
curvata L. is frequently seen as an epiphyte on the trees
IfHBri
and arborescent cacti. Grasses probably once formed an
important part of the herbaceous vegetation, but sheep
and goats have altered the composition of the her-
baceous flora.
The plant remains from zones of Coxcatlan Cave

dating to the Ajuereado phase are few but interesting.
The earliest level in which any perishable vegetal frag-
ments survived was Zone XXVI, probably dating to
Fig. 136. Garambullo (Myrtillocactus geometrizans) in
about 7000 b. c. In this level was one seed of mesquite,
flower.
a fragment of palm leaf, and two clumps of grass. Mes-
quite remains were found throughout the deposits of
cover of trees about eight meters tall is intermingled this cave and others. Brahea dulcis, the palm species
with arborescent cacti. Common trees growing across present, reappears only in Zone XVI, but it too is rep-
the hills are Acacia subangulata Rose, Cassia pringlei resented in other deposits and must have served as a
Rose, Acacia villosa (Sw.) Willd., Acacia cymbispina source of fiber from earliest times.
Sprague and Riley, and Acacia sericea M. and S. all — Zone XXV yielded no perishable remains at all, but
lacy foliaged members of the Leguminosae. Green and Zone XXIV contained identifiable fragments of food
red marble-like fruits readily identify cholulo, Z izy- plants which have been used by the inhabitants of the
phus pedunculatus. Scattered trees of pochote, Ceiba valley for several millennia. Among these were a few
parvifolia, project above the canopy, displaying puffs seeds of the wild grass setaria, one amaranth seed,
of silver at the end of the dry season. prickly pear seeds, one avocado cotyledon, and a seed
Intermixed with the trees, or sometimes in nearly of chupandilla. With the exception of setaria, these
pure stands, arboreal cacti raise their columns of yel- species are still widely used in cultivated or wild forms
low-green, gray-green, or blue-green above the canopy. today.
Escontria chiotilla is laxly, irregularly branched and Preservation was poor in Zone XXIII, the latest
bears both yellow flowers and fruit near the start of Ajuereado zone, which contained only a few seeds,
the rainy season. The straight, upright stems of Myrtil- identified as setaria, mesquite, and prickly pear. There
231
cent ia), Cucurhita pepo (probably a wild form), two

cultivated squash seeds (probably C. mixta), and sur-
prisingly, cotton (two unmistakable boll fragments
from Zone XVI). The next earliest cotton fragments
were found in an Abejas phase level of San Marcos
Cave. No cotton remains representing the intervening
Coxcatlan phase were recovered.
Fruits native to the valley or the barranca forests
were represented in the four El Riego zones with the
Z oneXVUl Zone XVI Zone XV Zone XIV

Avocado 8 7 11 5
Chupandilla 0 37 3 1
Ciruela ? 200 1 0
Cosahuieo 2 5 0 2
Fig. 137. Fluffs of pochote (Ceiba parvifolia) burst from the
best preservation, as shown in the accompanying tabu-
pods late in the dry season.
lation. Two of these fruits, avocado and chupandilla,
extend back to the very earliest zones with perishable
was also a questionable stem fragment of the latter remains. All of them appear in increasing amounts
cactus. throughout the remainder of the Coxcatlan Cave de-
Zones XXII through XIV represent the El Riego posit.
phase, which dates from after 6500 to about 5000 b.c. Little can be deduced from the plant remains con-
Preservation in the earlier of these zones was poor. The cerning the seasons of occupation of the El Riego phase
paucity of wood and bark fragments below Zone XVI other than that the cave seems to have been inhabited,
suggests that the deposit was subjected to moisture, or at least visited, during the summer and fall when the
resulting in deterioration of less durable plant frag- Jatropha ncojiauciflora seeds (Zones XXII-XX) and
ments. Most of the remains in levels XXII-XVIII chupandilla (Zones XVII-XIV) and cosahuieo seeds
(nothing perishable came from Zone XVII) were in the (Zones XVIII-XVI and XIV) were available. Increas-
form of seeds. These represent avocado, setaria and ing amounts of setaria seed may be interpreted as an
other grasses, amaranth, Acacia, Jatropha, and Conda- indication that setaria may have been sown and har-
lia spp., and two cacti, Opuntia sp. and Lemaireocereus vested near the which would require occupation
site,
hollianus. The earliest seeds of the fruit cosahuieo also at the beginning and end of the rainy season. In view
appeared in Zone XVIII. All the seeds would persist un- of the developing cultivation of crops in the succeed-
der adverse conditions because seed coats are more ing Coxcatlan phase, it is not unlikely that in later El
impervious to moisture than other plant parts. How- Riego times experiments in the planting of desirable
ever, fragments of prickly pear fruits and the earliest native plants such as avocado, maguey, and prickly
preserved fragments of maguey leaves also came from pear required attendance of the people for most of the
these levels. The pepper first appears as a ques-
chili rainy season from May through October.
tionable fragment in Zone XXI and again, this time The Coxcatlan phase, extending from about 5000 to
definitely, in Zone XIX. 3500 b.c., is represented by Zones XIII through XI.
The three later zones of the El Riego phase, XIV- Although as a group the plant materials from this
XVI, contained a variety of plant species, usually pre- horizon reflect marked dependence on wild plants for
served in increasing numbers. Caches of setaria and food, it is in these zones that the modern Mexican diet
grass seed in the three zones total about 180 ounces and based on maize, squash, beans, and chili pepper be-
show that these grains may have been stored for use comes discernible. Maize cobs appear for the first time
when other plant foods were scarce. Traces of am- in Zone XIII and again in the two more recent zones. In
aranth were also identified in these zones. Other plants all three levels, the cobs are small in size and few in
present were guaje (Leucaena esculenta), mesquite, proportion to other plant fragments. The squash re-
pochote ( Ceiba parvifolia pods (the earliest record for
)
mains are even fewer and consist of one or two seeds
this species), chili pepper, various cacti ( Opuntia Cep- ,
of Cucurhita moschata and a seed of C. mixta re-
halocereus hoppenstedtii, Lemaireocereus hollianus), covered from a eoprolite. Lagenaria, or bottle gourd,
maguey (both quids and leaves), the tree gourd (C res- rinds also appear in two of the three levels. The corn-
232
Table 26. Remains of Food and Fiber Plants from All Sites by Zone and Phase
VENTA SALADA
FOOD PLANTS FOOD PLANTS

1 7
Dioon edule Dioon edule
12.2oz. l3oz. 19oz. 16.5oz. 26.6oz. 9.2oz.
Selaria cl. macrostachya 2 oz. Setaria cf. macrostachya
33oz. IO. 802 . 4502. 90oz. 8 oz. lOoz.
Unidentified glass seed Unidentified grass seed
5 52
Grass quids 1 39 57 27 Grass quids
Zea mays (maize) 5 33 719 2497 1 685 4879 1151 Zea mays (maize)
4 7 6
Aciocomia mexicana (coyol) 46 23 5 1 Acrocomia mexicana (coyol)
Ainaranthus spp. loz. loz. 544oz. 2 oz. 78oz. Amaranlhus spp.
Peisea americana (avocado) 35 14 7 8 Persea americana (avocado)
Acacia spp.: pods 3
Acacia spp.: pods
seeds
seeds
Acacia sericea: pods
Acacia sericea: pods
Arachis hypogaea (peanut)
Arachis hypogaea (peanut)
Canavalia sp. (jack bean)
Canavalia sp. (jack bean)
Leucaena esculenta
Leucaena esculenta
Phaseolus aculilolius (tepary bean) 485
Phaseolus acutifolius (tepary bean)
P. coccineus (runner bean) lP
P. coccineus (runner beam)
P. lunatus (sieva bean)
P. lunatus (sieva bean)
1
P. vulgaris (common bean)
P. vulgaris (common bean)
Prosopis juliflora: pods
Prosopis juliflora: pods
(mesquite) seeds lOoz. 13oz. 29oz. 23oz. 25oz.
5oz. (mesquite) seeds
quids
46 quids
Casimiroa edulis (white bean) 2 Casimiroa edulis (white bean)
Malphigia sp. 2
Malphigia sp.
Jatropha neopauciflora 110 4 63
2 Jatropha neopauciflora
Jathropha sp. 1
Jatropha sp.
Cyrtocarpa procera (chupandilla) 37 3 1 78 1159 979 206 Cyrtocarpa procera (chupandilla)
Spondias mombin (ciruela) 200 1 4 57 19 Spondias mombin (ciruela)
Condalia mexicana 2 1000 78 176 10
Condalia mexicana
Ceiba parvifolia: roots 24 39 315 47 Ceiba parvifolia: roots
(pochote) pods 1 3 23 177 113 165 258 (pochote) pods
Cephalocereus hoppenstedtii 1 1 loz.s 3
1 Cephalocereus hoppenstedtii
(tetecho)
(tetecho)
Escontria chiotilla (jiotilla)
Escontria chiotilla (jiotilla)
Lemaireocereus hollianus: seed
Lemaireocereus hollianus: seed
stem
stem
L. stellatus: fruit
L. stellatus: fruit
L. weberi (cardon): fruit
L. weberi (cardon): fruit
Myrtillocactus geometrizans
Myrtillocactus geometrizans
(garambulo)
(garambulo)
Opuntia spp.: leaves 50 Opuntia spp.: leaves
(prickly seeds 0.03oz.
loz. (prickly seeds
pear) fruit 1
203 16 pear) fruit
Psidium guajava (guava) Psidium guajava (guava)
Bumelia laetevirens Bumelia laetevirens
Sideroxylon cf. tempisque Sideroxylon cf. tempisque
(cosahuico)
(cosahuico)
Oiospyros digyna (black sapote) Oiospyros digyna (black sapote)
Capsicum annuum (chili pepper)
Capsicum annuum (chili pepper)
Physalis sp. (tomate)
Physalis sp. (tomate)
Crescentia cujete (tree gourd)
Crescentia cujete (tree gourd)
Unidentified Cucurbita spp.
Unidentified Cucurbita spp.
(squashes & pumpkins) (squashes & pumpkins)
C. mixta C. mixta
C. moschata
C. moschata
C.pepo C. pepo
Lagenaria siceraria (bottle gourd)
Lagenaria siceraria (bottle gourd)
Apodanthera sp. Apodanthera sp.
FOOD AND FIBfcK FOOD AND FIBER

Agave spp.: quids Agave spp.: quids
leaves leaves
seeds seeds
fiber fiber
Yucca peiiculosa: leaves Yucca periculosa: leaves
seeds seeds
FIBER PLANTS FIBER PLANTS

Btahea dulcis (palm) Brahea dulcis (palm
Hechtia sp. Hechtia sp.
Tillandsia spp.: leaves Tillandsia spp.: leaves
spikes spikes
Tillandsia dasyliriifolia Tillandsia dasyliriifolia
Beaucarnea gracilis (sotolin) Beaucarnea gracilis (sotolin)
Agave kerchovei: stems Agave kerchovei: stems
Cissus sp. Cissus sp.
Gossypium hirsutum (cotton) Gossypium hirsutum (cotton)
OTHER USES OTHER USES

Selaginella sp. Selaginella sp.
Grasses (unidentified) Grasses (unidentified)
Agave sp.: spine Agave sp.: spine
Dioscorea sp.: inflorescence Dioscorea sp.: inflorescence
Quercus sp.: acorn Quercus sp.: acorn
Caesalpinia sp.
Caesalpinia sp.
Leucaena pueblana: pods
Leucaena pueblana: pods
Mimosa sp.: pods
Mimosa sp.: pods
Jatropha sp.: seed
Jatropha sp.: seed
Zizyphus pendunculatus (cholulo)
Zizyphus pendunculatus (cholulo)
Mammillaria sp. (bisnaga)
Mammillaria sp. (bisnaga)
Plumeria rubra (frangipani)
Plumeria rubra (frangipani)
Thevetia peruviana
Thevetia peruviana
Bark
Bark
Wood
V/ood
•Quantity to small to measure.

PLANT REMAINS
mon bean pod Phaseolus vulgaris ) found

(. in Zone XI tions of occupation in the spring, summer, and fall
is the earliest bean specimen from the Tehuacan exca- months.

vations. A fragment of chili pepper was found in each The next three levels, Zones X, IX, and VIII, repre-
of the three zones. sent the Abejas phase of about 3500 to 2300 b.c. Plant
Plants common form a far
to the native vegetation parts are fewer in number, although a variety of species
larger bulk of materials in the Coxeatlan phase than do are present, including fragments of such cultivated
remains of the corn-squash-bean-chili complex. Ma- plants as maize, squash, beans, and chili peppers. Be-
guey quids and leaves appear in amounts numbering in sides one pod each of common bean and Canavalia sp.,
the hundreds, and the quantities of setaria and grass there were a total of 118 tepary beans ( Phaseolus acuti-
seed remain large. Pochote pods are present in increas- folius).However, the proportion of maguey, prickly
ing numbers, along with fragments of the edible root. pear, and pochote remains to remains of maize and
Seeds of mesquite, Jatroplia neopauciflora, and prickly other annual plants is significantly greater. Fruit re-
pear and other cacti are well represented. mains are also proportionally less plentiful. The seven
A variety of fruit remains are present in the three species represented in the Coxeatlan phase levels are
Coxeatlan phase zones, some of them in really large present here, with an increase to fifty black sapote
amounts, as the accompanying figures shows. Added seeds in Zone IX the most notable occurrence.
to the four previously popular fruit species —ehupan- Pochote pods and fragments of the cacti, cardon and
dilla, avocado, ciruela, and cosahuico— are three new xoconoehtle, indicate spring residence in these zones,
ones: eoyol ( Acrocomia mexicana ), white sapote ( Casi - and mesquite seeds show summer, and assorted fruits
fall to early winter occupation.
Zone XIII Zone XII Zone XI Coxeatlan Cave was not occupied in early Formative
Avocado 0 1 32
times. There is a gap of at least two thousand years
Chupandilla 22 78 1159 between Zone VIII and Zone VII. The latter represents
Ciruela 5 0 4 the latter part of the Santa Maria phase, which ex-
Cosahuico 5 11 271 tended from before 900 b.c. to about 200 b.c. The most
Covol 0 4 7 significant change to be noted in the plant materials of
White Sapote 0 0 4 this level is the slight predominance of maize (580
Black Sapote 0 0 2 cobs), squash, and cultivated fruits over maguey, po-
chote, prickly pear, mesquite, and other plants repre-
miroa edulis), and black sapote (Diosj)ijros diguna). sentative of the native vegetation. Much of the maguey,
Although remains of these new species are not yet however, may have been cut from permanent plant-
numerous, their appearance is significant. These plants ings.
could not grow Tehuacan region without supple-
in the Besides maize, squash, and gourds, there was a single
mentary water, and the appearance of all three in Zone runner bean Phaseolus coccineus) in Zone VII, as well
(
XI indicates that the Coxeatlan phase people had man- as a few fragments of chili pepper, one ounce of ama-
aged to control the water available and supply it to ranth (probably cultivated), and two cotton plant
plants which were desirable but were by no means fragments. From Santa Maria times on, cotton remains
dietary necessities. Evidence of the progress of these appear consistently, along with cloth woven from cot-
people as agriculturists may also lie in the vast number ton fiber. Also present in Zone VII, even after maize
of chupandilla seeds recovered — 1,159 in Zone XI has become the staple grain, are 6.5 ounces of setaria
alone —which show an interesting range of sizes, per- and grass seed. Although these seeds never again ap-
haps indicative of experiments with cultivation. pear in the large quantities in which they were re-
The Coxeatlan phase people, then, were growing an covered from the El Riego and Coxeatlan horizons,
increasing number of cultivated plants, both annual they persist in smaller amounts through Zone I. They
and perennial, but the bulk of the plant material still were a source of food in the valley for some 8000 years,
came from a variety of native species. It is probable even though their role in the diet shifted drastically
that some of the latter, such as maguey and prickly after corn was domesticated.
pear, were being cultivated too. The plant materials as Among the fruit remains in Zone VII were 35 avo-
a whole from Zones XI-XIII indicate that a large num- cado, 979 chupandilla, 57 ciruela, 166 cosahuico, 6
ber of people were occupying the caves, with an in- coyol, 7 white sapote, and 13 black sapote seeds. Most
crease in number from the earliest zone to the latest. of the fruits indicate fall and early winter occupancy.
Among the diverse plant materials present are indica- Pochote pods, a single tetecho fruit, and mesquite and
233
Jatrapha neopauciflora seeds show that the cave was plant materials, but thev are not as numerous in rela-
occupied in the spring and summer as well. tion to the maize and maguey remains as formerly.
Zones VI, V, and IV are assigned to the Palo Blanco Chupandilla seeds are the most plentiful, as the ac-
phase, dating from about 200 b.c. to a.d. 700 or 800. companying tabulation shows.
Maize remains number in the thousands in these levels,
as do both quids and leaves of maguey. Squash and
Zone III Zone II Z one I
gourd specimens are present but not numerous. Bean Avocado 7 7 3

species are represented by 485 tepary beans and a Chupandilla 136 119 47
runner bean pod in Zone VI and a few common bean Ciruela 25 63 4
Cosahuico 58 43 9
seeds in all three levels. The large peanut seed in
Covol 5 4 4
Zone VI, the earliest fragment of Arachis hypogca re-
White Sapote 3 2 1
Zone VI Zone V Zone IV Black Sapote 3 7 1
Avocado 85 14 15 Among plants parts which can serve as seasonal indi-

Chupandilla 292 167 206 cators, pochote pods indicate spring occupancy of all
Ciruela 19 4 0
three zones, seeds of mesquite and Jatropha neopauci-
Cosahuico 177 74 483
flora show summer residence, and fruit remains are
Covol 46 23 0
evidence of fall and early winter occupancy.
White sapote 13 7 4
Black sapote 291 9 11 The plant materials recovered from twenty-four
zones of Coxcatlan Cave document man’s utilization of
covered, is obviously from a cultivated strain. Some of the botanical resources of the Tehuacan Valley and his
the huge amounts of amaranth recovered from Zones acceptance of introduced plants as additions to his diet.
IV and VI were found in large storage pits. The com- From a seed of mesquite, a fragment of leaf, and palm
mon fruits again make up a substantial portion of the two clumps Zone XXVI, the plant remains
of grass in
plant material in these zones, as the accompanying increased in quantity, quality, and variety to include
tabulation shows. thousands upon thousands of fragments representing
All the usual plant products representative of the well over fifty species. Among the remains are the re-
native flora appear in the earlier two Palo Blanco zones, sults of man’s first experiments in agriculture, but no
—
but the cactus species notably Opuntia are absent — one fragment can be singled out as proof of the be-
from Zone IV. The various zones were occupied in the ginning of agriculture. The early inhabitants trans-
spring, judging from the pochote pods in each. Mes- planted fruit trees to more accessible locations and, in
quite remains and Jatropha neopauciflora seeds show their cultivation, discovered that selection of better
summer occupancy, and fall-ripening fruits indicate propagules increased productivity. They probably cul-
fall to early winter residence. Perhaps these zones rep- tivated setaria, and they undoubtedly learned that
resent visitsby groups of agricultural workers through maguey and grown easily by
prickly pears could be
the growing and harvesting season. vegetative propagation. The first steps toward agricul-
The three upper levels of the cave, Zones III, II, and ture were probably made during the El Riego phase,
I, date from the Venta Salada phase of before a.d. 800 for in the subsequent period very definite products of
to about 1540. Maize cobs in these zones are more agriculture — cultivated squashes, the common bean,
numerous than in the Palo Blanco zones. Maguey quids early forms of maize, fruits needing cultivation for sur-
and leaves once again appear in huge quantities, but in vival — all appear, though in very small numbers com-
these levels do not equal the quantity of maize re- pared with the amount of material gathered from wild
mains. The various cucurbit species are well repre- species. From Coxcatlan times on the remains from
sented. Beans are rare— one small lima bean and two the cave present a history of the enlarging cultivation
common beans in Zone III are the only specimens. of food and fiber plants.
Guaje ( Leucaena esculenta ), a tree legume which may
have been cultivated, is substantially represented. It Description of the Plant Remains
has been consistently present since Zone VII and ap- The plant remains posed few problems in identifica-
pears sporadically before that. Peanuts appear again tion, since there had been no major change in the
in Zones I and II, amaranth in Zones II and III, and flora and, consequently, no change in the plants avail-
setaria in small quantity in each zone. able to the local populations. Fortunately, the pattern
The fruits, as usual, make up a large share of the of dietary preferences discernible today is the modern
PLANT REMAINS
equivalent of the preferences deduced from the archae-

ological material. Local men working for the Tehuacan
project were able to make tentative identifications of
many obscure fragments, which upon comparison with
known material proved to be correct. For this expert
help, I am very grateful. As with the debris in any
junk pile, some fragments defy identification, prin-
cipally because the plant parts necessary for classifi-
cation are missing. As groups of which this is partic-
ularly true are discussed, the fact will be noted. Be-
cause it is easier to make comparisons when related
plants are discussed together, the order of discussion
will be that set down in the Engler and Prantl system
of family arrangement.
Selaginellaceae (Club Moss Family) Fig. 138. Dioon edule growing on the mountain slope
above Teotitlan del Camino.
Sclaginella sp. (resurrection plant; helecho). Basal
portions of two plants were found in Coxcatlan Cave,
Zones VII and IV (Santa Maria and Palo Blanco
phases). The fragments were too incomplete to identify
specifically. The local flora today includes the resurrec-
tion fern, S. lepidophylla. The two specimens were
probably casual intrusions. No local uses or supersti-
tions connected with this plant were discovered, nor
does Martinez (1928) mention any use for Sclaginella.
Cycadaccae (Cycad Family)

Dioon edule Lindl. (chamal: Martinez 1937). The
broken seed coats of this plant were not abundant, but
showed that the plant had been fairly persistently
used over some thousands of years, primarily by the
occupants of Coxcatlan Cave (see Table 26). The only
other find of this plant was one fragment of seed coat
in the East Niche of El Biego Cave, Zone D (Palo
Blanco). Today Dioon do not grow near any of
trees Fig. 139. Cone of Dioon edule, in which edible seeds are
the archaeological sites, nor do local people seem to borne.
know the plant. I saw a few trees on slopes above Teo-
titlan del Camino, Oaxaca, near the edge of the oak and
pine forest where precipitation is somewhat greater

than in the thorn-scrub-cactus forest in the valley be- Grass brought into the caves, perhaps for padding or
low. Farmers plowing a nearby field used the name bedding, was sometimes found in large amounts. Gen-
“palmilla,” but said they knew of no use for the plant. erally, it was impossible to identify the species of grass
The oblong seed, about 2.0 by 3.5 cm., provides a involved, because critical decisions depend upon char-
starch food when cooked, but it could never have been acters of themature inflorescence. The above species
important in the diet of the Coxcatlan Cave people. were identified from inflorescence fragments culled
The cones and foliage are known to be poisonous to from some of the masses of grass.
animals, and the seeds might once have Ireen consid-
ered medicinal. A number of grass quids, consisting of chewed stalks,
were also found. The various grasses were probably
Gramincae (Grass Family)
gathered from expanses which must have developed
Chloris virgata Sw., Heteropogon contortus (L.) across open areas during the wet season. Today herds
Beauv., Trichloris pluriflora Foum. (grass; pasto). of sheep and goats keep the grass well trimmed.
235
ENVtRONMENT AND SUBSISTENCE
Fig. 141. Leaf of the palm Brahea dulcis, Tc 254, Zone B

(x 0.6).
posits, no method of preparation can be ascertained,

Fig. 140. Fruit of the covol palm (Acrocomia mexicana),
but it is obvious that grass and setaria seed constituted
Tc 50, Zone III, actual size.
an important dietary item in El Riego and Coxcatlan
times.The seed persisted to the top of the deposit in
Unidentified grass —stems only (reeds, grass; carrizo, Coxcatlan Cave as a supplementary food.
pasto). Numerous fragments of grass stems split to
various degrees are obviously discards from the manu- Zea mays L. (corn; mais). Maize was an important
facture of baskets or other items. A few cut sections of crop in the Tehuacan Valley from the Coxcatlan period
cane were charred on one end as though they might on. The material recovered is analyzed above in Chap-
have been used as cigarettes, but no charred leaf ma- ter 9.
terialwas found inside to constitute absolute proof. The
stem sections and split fragments are not complete Palmaccae (Palm Family)
enough to identify. They are probably stems of Phrag- Acrocomia mexicana Karw. ex Mart, (coyol). Coyol
mites communis Trin. Today, irrigation ditches of the palm fruits are a commonplace item in the deposits
Tehuacan area are frequently lined by masses of of Coxcatlan and Purron caves from Coxcatlan times
Arundo donax L., locally called “carrizo,” but a com- on. Some fragments also appeared in late levels of the
parison of the archaeological material shows that the East Niche of El Riego Cave. From the appearance of
ancient specimens are not this species. Arundo is def- the fragments, both the outer fleshy portion and the
initely a post-Conquest introduction. inner seed were used. The outside of the pieces is
usually devoid of flesh, and the bony endocarp is ir-
Sctaria macrostachya HBK. (pasto). Seed of this

cf. regularlv broken. The coyol is still used locally as a
species was found in most levels of Coxcatlan Cave, in fieshv fruit for eating out of hand and, in some parts of
caches or abundantly through the They were also
fill. its range, for the manufacture of wine. Coyol is found
present in two levels of San Marcos Cave. The small nowhere today in the Tehuacan region as a natural
yellowish to brown caryopses are about 2 mm. long and wild plant and therefore must have been a cultigen in
1.2-1. 5 mm. wide, flattened on one side and curved and prehistoric times.
rugose on the other. Sometimes they were found still
attached to small fragments of bristly inflorescence. Brahea dulcis (HBK.) Mart, (palm; palma). This
Nowadays, this species is not common in the valley, but palm is the source of leaf fibers presently used for the
it may once have been abundant along arroyos and on manufacture of hats in the area west of the Tehuacan
alluvial fans, together with other grasses. Under good Valley. The fruits are sweet and edible, but no fruit
conditions, the culms may reach a height of a meter remains were recovered. Small amounts of leaf frag-
with a panicle of fruit as long as 25 cm. A native of ments were found in El Riego, San Marcos, and Cox-
North America, this grass grows as far north as the catlan caves, and a single fragment of leaf was among
southwestern United States. the very earliest remains from Coxcatlan. The sparse-
Analysis of the pieces of inflorescence caught up with ness of material probably indicates few locally avail-
the seed reveals that inhabitants of the cave winnowed able sources or a preference for other kinds of fiber.
the seed by rolling the heads of grass between the The lack of fruit remains strengthens the opinion that
palms of their hands. From the seed found in the de- few Brahea trees grew near the cave sites.
236
PLANT REMAINS
Fig. 142. Leaves of soluchil (Tillandsia sp.) from which

the outer layers have been stripped. The ends were being
worked into fiber when the piece was discarded in San
Marcos Cave (x 0.6).
Bromeliaceae (Pineapple Family)
Hechtia sp. Colonies of spiny Ilechtia sprawl across

the dry slopes near all of the cave sites. They offer a
ready supply of leaves which, like many species of the
Bromeliaceae, have tough fibers. The leaves are not
very long and are stoutly armed with sharp spines,
which must have made unpleasant the gathering and
processing of them for fibers. For the same reason,
though, the leaves would probably not have been
casually gathered for discard in the caves. Small
amounts of leaves were found in Abejas, Santa Maria,
Palo Blanco, and Venta Salada zones of Coxcatlan or
San Marcos caves. Fig. 143. Sotolin (Beaucarnia gracilis) near Zapotitlan. The
leaves, when dried and split, furnish a very tough fibrous
Tillandsia spp., including T. (lasyliriifolia Bak. (air material.
plant; soluchil).Leaves of several species of Tillandsia
were present in Coxcatlan Cave from Coxcatlan levels could find no evidence that this material was being
on and in more recent levels of San Marcos and El used for fiber. Since present-day Mexicans use Spanish
Riego caves. The native species in the valley have a moss to decorate arches and other structures for fiestas,
heavy layer of scurfy scales on their leaves, which the the plant possibly had religious significance for earlier
people apparently managed to remove. If my interpre- peoples.
tation of the recovered fragments is correct, the inner
Liliaceae (Lily Family)
leaf tissue was then rolled into quills and worked (per-
haps by chewing) into soft fibrous material. The plants Beaucarnea gracilis Lem. (sotolin). Sotolin grows on
were probably brought into the sites whole, and then the well-drained soil overlaying limestone on the west-
were divided into individual leaves, for several pieces ern side of the Tehuacan Valley. It was probably once
of inflorescence were also found in Palo Blanco and quite abundant, but reproduction is apparently slow
Venta Salada levels. (I have seen no evidence of seedling plants even in
areas in which fruiting trees are usual). The leaves are
Tillandsia usneoides L. (Spanish moss; paxtle). Small tough and were probably a source of fiber. The earliest
amounts of Spanish moss were found in the deposits, record for this plant is four leaf fragments in Tc 50,
perhaps brought from nearby moist barrancas or from Zone XI (Coxcatlan). Other foliage fragments were in-
oak and pine forests farther up the mountain slopes. I frequent.
237
Fig. 145.Charred spines of maguev (Agave ghiesbrechtii)

from Tc 254, Zone F, actual size.
leaf tip spine, Tc 272-A, 1 (Palo Blanco); Agave ker-

chovei Lem. (rabo de leon) stem fragment, Tc 254-B,
11, and Tc 254-C 1 , 1 (Palo Blanco), Agave cf. ghies-
brechtii Koch, leaf tip spines, Tc 254-B, 5, and Tc 254-
Fig. 144. Izote (Yucca periculosa) in bloom in March at San
C 1
, 5; Tc 50- VI, 1 (Palo Blanco); Tc 254-D, 1 (Abejas);
Marcos Cave.
Tc 50-XIV, 1 (El Riego).
The frequency and distribution of the maguey re-
mains lead me to believe that the people of Coxcatlan

Yucca periculosa Bale (yucca; izote). Izotes serve and Purron caves in later periods may have been draw-
many purposes in present-day Mexico. The flowers and ing upon maguey plantings for the uses for which they
fruits of some species are edible, and a liquor is fer- were putting the plant. Consumption of maguey leaves
mented from the tissue inside the trunk of at least one (baked?) was high in both caves. The numbers of frag-
kind. Many izotes have usable leaf fibers. The spongy ments of maguey leaves in the two caves and the find-
tissue inside the trunk of some species is used for pad- ing of fiber material in San Marcos Cave may indicate
ding under the saddles of pack animals. The local izote a fiber industry among these people. They may well
was and is still used by the people of the Tehuacan have supplied fiber or finished woven used by
articles
Valley. Fragments of izote leaves were more plentiful other people.
in the later levels of the deposits, and in some levels Inflorescence material of certain species of maguey
quantities of seeds were present. was present because the flowers were used as food.
Although the plant remains do not demonstrate the
Amanjllidaceae (Amaryllis Family)
existence of an agua miel or pulque industiy among
Agave spp. (century plant; maguey). Maguey plants these people, it seems safe to assume that some among
furnished a major portion of the raw materials needed them discovered that drinking the fermented sap was
by the early people of the Tehuacan Valley. Chewed a pleasant experience.
leaf tissues in the form of quids indicate that maguey Besides the thousands of leaf fragments and quids
was a large item in the diet in all periods and a major recorded in the tables, the following amounts of fiber
item in the earlier levels. Numerous leaf fragments, were recovered: Tc 254-B, 125 oz.; Tc 254— C’, 282 oz.
spine tips, and other plant parts show that the plant was (Palo Blanco); Tc 254-C, 10 oz. (Ajalpan); Tc 254-D,
put to many uses. 44 oz. (Abejas).
Unfortunately, species of Agave are difficult to dis-
Dioscoreaceae (Yam Family)
tinguish unless there are ample leaf fragments with in-
tact margins and tips and, frequently, flowers or fruit. Dioscorea sp. (yam). One fragment of inflorescence,
Only a few of the recovered fragments could be identi- with dehisced seed pods adherent to it, was identified
fied to species. These are: Agave ef. kancinskii Zuec. as Dioscorea. The fragment may have fallen from a
238
PLANT REMAINS
piece of vine used to tie a bundle. Certainly, no other Fagaceae ( Oak Family)
evidence indicates that the people of the Tehuacan Val-
Quercus sp. (oak; roble). Three acorns were found,
ley knew or used edible yam tubers. The Dioscorea one in Tc 50-V (Palo Blanco) and two in Tc 50-XI
specimen, found in Tc 50-VII (Santa Maria), must be
(Coxcatlan).
an accidental intrusion.
Amaranthaceae (Pigweed Family)
Amaranthus cf. cruentus L. and A. cf. leucocarpus
Wats, (pigweed; quelite). Remains of Amaranthus spp.,
sometimes in large quantities, were found in Coxcat-
lan, El Riego, San Marcos, and Purron caves. The his-
tory of the ceremonial use of amaranth (frequently
called “alegria” or “bledo”) has been told in detail by
Jonathan Sauer (1950). It was undoubtedly used for
religious purposes in the Tehuacan area. However,
amounts of amaranth seed and inflorescence material
recovered from Coxcatlan Cave suggest that the plant
was also a part of the diet of the people living there.
The earliest remains date from the El Riego phase.
Workmen helping with the excavation of the caves
commonly used the name “quelite” for the Amaranthus
fragments, although thisname is probably more usually
applied to the weed amaranth, not to the cultivated
species whose seeds are used for food.
Lauraceae (Laurel Family)

Persea americana Mill. var. drymifolia (Schlecht. and
Cham.) Blake (avocado; aguacate). Remains of avo-
cados came from almost all levels of Coxcatlan Cave,
beginning with Zone XXIV of the Ajuereado phase.
Fig. 146. Inflorescences of amaranth with seeds removed, Seeds of this fruit were present in upper levels of other
Tc 50, Zone VIII, actual size. caves also. Variation in the size of the fruit of avocado
Fig. 147. Avocados (Persea americana). Left pit from Tc 50, Zone XVIII; center, two pits from Tc 50, Zone VII;
,
right,
a fruit of the anise-flavored Mexican race (var. drymifolia) purchased in the Tehuacan market. All x 1.5.
239
is normal, as it is for most plants. However, fruit col-

lected from native wild trees is consistently small. Be-
cause the seeds vary in shape as well as size, the aver-
age size of wild avocado seeds is best expressed as an
index figure derived from multiplying the length in
cm. bv the width in cm. For dried seeds of wild-type
fruit from the Tehuacan market / x tv = 3.47. When-
ever the size of dried seeds falls outside the usual range, Fig. 148. Peanuts (Arachic hypogaea). Left, seed from Tc
it apparent that selection for larger
is fruit is being 50, Zone VI; right, fragment of shell from Tc 50, Zone I.
undertaken. Both actual size.
In the deposits in Coxcatlan Cave, small seeds simi-
lar tothose from present-day native trees are found in
Zones XXIV (/ x tv = 4.62) and XVIII (/ x tv = 6.48), that they recognized their success with selection in
both of which are based on measurements of single other crops, such as maize. With the latter crop planted
cotyledons. The only intact cotyledon of the eight from annually, results of selection become apparent within
Zone XVIII is the largest found in this level. In the a very few years. Selection in a tree crop is tediouslv
Coxcatlan horizon, two measurable cotyledons yield slow to produce results for several reasons. Pollenation
/ x tv = 3.40, and in the Abejas horizon, two cotyledons in mixed groves of avocados is open and the chances
show l x tv = 3.78. In the Santa Maria horizon, a def- of a large-fruited tree being pollenated from another
inite increase occurs (/ x tv = 7.80, based on nineteen large-fruited tree are very tenuous. Further, the seed-
cotyledons), but it should be noted that the small coty- lings of selected large-fruited forms would not reach
ledons persist, the smallest being 1.4 x 1.5 (l x tv — bearing size for at least seven years after planting and
2.10). The size of the cotyledons increases greatly in the young treeswould not produce fruit of maximum size
Palo Blanco phase (/ x tv = 10.50) and decreases slight- until they had been bearing for several years. There-
ly in the most recent Venta Salada horizon (l x tv = after, the large-fruited selections would be genetically
9.92). Even in the latest horizon, though, the small wild- swamped by pollenation by small-fruited, wild-type
type avocado is found (the smallest measured, / x tv trees in the area. Thus, it must have taken a very long
= 2.55). time before the size increase due to selection became
Several facts are evident concerning avocados. The apparent in the cotyledons being deposited in the
other vegetal remains indicate that the climate of the caves, particularly since all fruit available would be
area has remained unchanged from at least the be- harvested for use irrespective of size.
ginning of El Riego time to the present. Under this Cultivation of avocados also implies water control
rainfall regimen, avocado cannot grow naturally in the in the valley, inasmuch as avocados cannot exist in
valley nor on the lower slopes of the mountains. Avo- the valley over the long dry season. Perhaps the best
cado is a thin-foliaged tree of the mesic forests of the evidence for a form of irrigation is the sudden appear-
barrancas higher up the mountains. The earlier coty- ance of coyol and black and white sapotes among the
ledons must have come from fruit gathered from the remains from Zone XI (Coxcatlan phase) of Coxcatlan
barranca These mesic forests probably never
forests. Cave. Because of its obv ious popularity, avocado must
extended to the lower areas where the sites are located, have been under irrigation cultivation also and lessons
since so few products from plants native to the bar- learned in the successful cultivation of avocado may
—
ranca environment such as Inga, Piper, Ficus, and have paved the way for the introduction of other fruit
—
Pcperomia are present among the remains. trees needing supplementary water.
It is obvious that the avocado was brought into
cultivation in the Tehuacan Valley. A major increase
Leguminosae (Pea Family)
in
fruit size would not occur in wild populations of av- Acacia sp. and A. scricea M. & S. (guaje). Fragments
ocado The marked increase in size of seeds found
trees. of Acacia pods were frequent in Coxcatlan Cave, but
in the Santa Maria horizon can be attributed only to they were seldom distinctive enough to be identified to
active selection of larger and better fruit for propaga- species. Some pieces were also recovered from San
tion. The act of selection presupposes that the cultiva- Marcos Cave. The green pods were probably eaten
tion of avocado had been sufficiently well established whole, although present inhabitants of the area con-
so that the valley people knew they would be successful sider them inedible. A few of the pods are definitely
when they planted avocado seeds. It further indicates fruit of A. sericea, locally known as “guaje bianco.” The
240
PLANT REMAINS
earliest Acacia specimens appeared in the Coxcatlan

phase.
Arachis hypogaea L. (peanut; cacahuate). Peanuts

were found only in Te 50, Zones I and II (Venta Salada
phase) and VI (Palo Blanco). The size of the fragments
leaves no doubt that these were remains of fully culti-
vated plants. The single seed from Zone VI is large
when compared with modern peanuts. This plant is not
native to the Tehuacan region.
Caesalpinia velutina (B. & R.) Standi, (guaje). The

single pod fragment of this species, from Tc 254-D
(Abejas), is probably an accidental introduction.
Canavalia sp. (jack bean). Remains of this and

Phaseolus spp. are described above in Chapter 10.
Leucaena esculenta (M. & S.) Benth (guaje). This

variety of guaje is presently widely cultivated and ap- Fig. 150. A cache of seeds of mesquite (Prosopis juliflora)
from Coxcatlan Cave.

pears to be a regular dietary item among the Indian
people of the Tehuacan Valley. The green pods, the
guaje proper, are used when available. When the trees
are not fruiting, small galls called “poloehoco” are
gathered from them and boiled to be eaten. Fragments
of pod and seed were found as early as the El Riego
phase of Coxcatlan Cave, and they appear fairly con-
A few remains were found in
sistently in that site.
Purron Cave and a moderate number in San Marcos
Cave.
Leucaena pueblana B. & R. (guaje). Pod fragments

were found in a few Venta Salada and Palo Blanco
levels. It is not certain that they were used as a food, Fig. 151. Chewed quids of mesquite, from Tc 254, actual
hut a total of 15 fragments probably are not an acci- size.
dental intrusion.
Mimosa sp. A few fragments of pods came mainly Prosopis juliflora (Sw.) DC. (mesquite). Remains of
from Venta Salada and Palo Blanco zones. Two speci- mesquite, primarily in the form of seeds, appeared in
mens came from the Abejas phase of San Marcos Cave. considerable abundance in the deposits. Chewed pods
constituting quids and whole pods were also discov-
ered. Mesquite remains come from the earliest level at
which there was preservation and continue to the top of
the deposits. This plant obviously furnished one of
the regular items of the prehistoric diet. During later
times, when techniques of cultivation were well known,
mesquite may have been cultivated, although the re-
mains show no morphological evidence of this. At the
present time, fruit is harvested from wild trees as well
as from cultivated trees in the Tehuacan Valley.
Rutaceae (Citrus Family)
Casimiroa edulis Llave & Lex (white sapote; zapote

bianco). Zapote bianco or zapote duermelon is culti-
Fig. 149. Pods of Leucaena esculenta from Tc 254, actual vated for its fruit. It is one of the plants whose remains
size. in the Tehuacan excavations are indicators of some
241
Fig. 152. Seeds of chupandilla (Cyrtocarpa procera) from

Tc 50, Zone XIII, actual size.
kind of water control in the valley. It cannot survive

in the Tehuacan climate outside of cultivation. Because
Fig. 153. Chupandilla fruit, with dried flesh and broken
it has to be cultivated, and the fruit is a supplemental
seeds, Te 50, Zone II, actual size.
rather than an essential item in the diet, its appearance
in the deposit is significant.
Had there been some change in the species compo-
Euphorhiaceae (Poinscttia Family)
sition of the plant remains which could be interpreted
as having been caused by a climatic change or a per- Jatropha neopauciflora Pax. Seeds of this species are
manently higher ground water level in the past, the conspicuous in the deposits from the earliest to the
appearance of white sapote could be accepted as a pos- latest levels. Since the seeds are never cracked, but
sible component of a more mesic forest in the area. clearly show the marks of rodents’ teeth whenever they
Because the indications are otherwise, species among are found open, the seeds were apparently not used
the remains requiring a more mesic environment for — by the gatherers. From the abundance of these seeds,
instance, white sapote, black sapote, coyol, avocado, though, it is obvious that some portion of the plant,
—
and ciruela are evidence that local agriculturists were probably the foliage, constituted one of the regular
aware of the conditions under which they could be dietary items. Most of the seeds appear in Coxcatlan
grown and had sufficient knowledge of irrigation tech- and Purron caves, but one collection represents the
niques to be able to furnish water for these plants dur- Palo Blanco phase of San Marcos Cave.
ing the dry season. Jatropha sp. In addition to the nearly spherical seeds
Although there is variation in the sizes of the seeds, of Jatropha neopauciflora, two seeds of another species
the fragments of fruit and the seeds do not show any with a three-seeded capsule were found: one in Tc 50,
clear indication of selection for larger fruit in the upper Zone III (Venta Salada), and one in Tc 50, Zone VII
levels of the deposit, in marked contrast to fruit trees (Santa Maria). Foliage of the plants from which these
like avocado. two seeds came may also have been eaten as greens,
White sapote remains appear regularly but rarely but the seeds probably were accidental intrusions or
in large numbers in the middle and upper zones of were for medicinal use. Many of the Jatropha species
Coxcatlan, El Riego, and Purron caves. Because each are known to have a cathartic principle in the seeds.
fruit has a number of seeds, the excavated sample may
represent fewer whole fruits than do remains of single-
Anarcardiaccac (Poison-ivy Family)
seeded species. The largest number of seeds recovered Cyrtocarpa procera HBK. (chupandilla). Chupan-
was sixty-six in Zone C of the East Niche of El Riego dilla occurs in the Tehuacan Valley as a native mem-
Cave, a Venta Salada component. ber of the thorn-scrub vegetation. A tree growing only
to 6 or 8 m. in height, it has a twisted trunk, gray bark,
Malpighiaceae (Malpighia Family)
and up to about seventeen leaflets to each compound
Malpighia sp. (nanche). Nanche, the fruit of at least leaf. The fruit is pubescent and said to have a resinous
two species of Malpighia in the Tehuacan area (the flavor, hut it is used where it occurs from Jalisco to
name is applied to Byrsonima crassifolia in other parts Puebla and Oaxaca. It is not a prominent member of
of Latin America), is locally collected today, but only the Tehuacan thorn-scrub, usually growing as scat-
one from Zone XIII of Coxatlan Cave, was iden-
fruit, tered individuals. According to Miranda (1948), the
tified among the plant remains. No seeds were found. bark is a substitute for soap.
PLANT REMAINS
Fig. 154. Fruit of ciruela (Spondias mombin from Tc 50,

Zone XIII.
One in Zone XXIV

seed of chupandilla was present
of Coxcatlan Cave (Ajuereado phase), and from the
El Riego phase on the seeds were present in increas-
ing quantity, in all but one instance outnumbering the
Fig. 155.Cotton (Gossypium hirsutum). Above: “wild”-
remains of any other fruit. In Purron Cave the seeds
tvpe bolls:left, from Cape Sable, Florida, and right, from
appeared sporadically at first, beginning with Zone R
Tc 50, Zone V. Below: left segment of large boll from Te
(El Riego phase), and they increased from Santa Maria
50, Zone XVI, and right, from the cultivar Deltapine grown
number from a single level of this
levels on; the largest
at Stoneville, Mississippi. All approximately actual size.
cave was 923 Zone E (Palo Blanco phase). Chupan-
in
dilla seeds were much less common in the East Niche
of El Riego Cave, and only one seed was recovered tive of a warmer climate which would enable ciruela
from San Marcos Cave, from a Palo Blanco level. An to grow in the upper reaches of the valley.
analysis of the seeds show a subtle but definite increase Because the ciruela does not grow in the valley as
in size through time. Seeds measured for length and a native member of the thorn-scrub flora, the oldest
width (in cm.) averaged (/ x to ) 2.31 inZone XI of Cox- find of ciruela seeds is highly significant. A lot of 200
catlan Cave. Seeds from Zone III of the Venta Salada seeds from Zone XVI of Coxcatlan Cave (El Riego
phase of the same cave had the largest average, 3.09. phase) indicates a nearby source of fruit, since this
is not a primary food item, but a secondary food which
Spondias mombin L. (Spanish or hog plum; ciruela).
contributes variety without furnishing essential food
Ciruela is another of the fruits which do not grow as
values not obtainable from other foods. Chupandilla,
a part of the native vegetation of the Tehuacan area.
which ripens later in the year, is a similar fruit in size
It is susceptible to frost injury as well as requiring
and flavor.
more, and more evenly distributed, rainfall than is
found in the valley. Once well established, ciruela Rhamnaceae (Buckthorn Family)
under somewhat drier conditions than
trees will persist
Condalia mexicana Schleeht. Seeds of Condalia are
avocado, but in the Tehuacan region they cannot grow
found in the early levels of Coxcatlan Cave, reach
unless they are tended and irrigated.
Again, the question arises as to the effect of a higher
maximum abundance in the intermediate zones, and
fall off in number in the upper levels. They are not
ground water level in the area in earlier periods. At the
present time, the tree
present in other excavations. The fruits are small,
is apparently not persistent out-
bright red, and edible. One species of Condalia yields
anywhere in the region of the upper
side of cultivation
a blue dye, and the root bark of several species is a soap
Papaloapan drainage, since it is not mentioned by Mi-
randa (1948). Even in the frost-free lower reaches of the
substitute. The shrub is a member of the native thorn-
scrub complex.
region, then, rainfall is not sufficiently well distributed
during the year to enable ciruela to grow naturally. Zizyphus pedunculatus (Brandg.) Standi, (eholulo).
None of the information available from the plant re- Cholulo is a native of the thorn-scrub vegetation and
mains indicates that the dry period was ever shortened is gathered locally for the saponin content of the fruit,
so that an elevated ground water level persisted year- which is be particularly good for washing the
said to
round during the time represented by the deposits. hair. Seeds of cholulo appear scantily in the upper
Furthermore, no vegetational change occurred sugges- levels of Coxcatlan and Purron caves.
243
Vitaceae (Grape Family)
A number of stem fragments apparently

Cissus sp.
of thisgenus were found with marks of usage which
suggested that the stems were used whole for tying
bundles and were sometimes pounded to obtain the
rather harsh fiber. The earliest fragment appeared in
an El Riego level of Coxcatlan Cave. Fragments were
found in later levels of that cave and in Purron and
San Marcos caves.
Fig. 156. Pochote pods from Tc 254, x 0.6.
Malvaceae (Malloiv Family)
Gossypium hirsutum L. (cotton; algodon). Cotton is
well represented in the excavations by artifacts rang-

ing from small bits of string to sizable pieces of fabric.
The artifacts are primarily from the Palo Blanco and
Venta Salada horizons, but some date from the Ajalpan
and Santa Maria phases. There is even a small, ques-
tionable bit of cotton yarn from an Abejas level of
Coxcatlan Cave (Zone IX). In addition, boll fragments
were found scattered through the deposits, and from
these it was possible to determine that only G. hirsutum
is present. The earliest cotton specimens found were
two segments of bolls from Zone XVI of Coxcatlan

Cave (El Riego phase). These are the earliest cotton
remains yet recovered, but the possibility that they may
be intrusive should be noted. A gap of about two
thousand years separates these and the next boll frag-
ments, recovered from Zone D of San Marcos Cave
(Abejas phase). Cotton was present in all subsequent
phases. The fragments range in size from small seg-
ments resembling the “wild” type of G. hirsutum bolls
to those which are indistinguishable from modern cul-
tivated varieties.
Although it is difficult to assess color in old, long-
Fig. 157. Pochote fiber used to enclose what may have been buried samples, the natural color of the fiber recovered
a paste of ground seeds, Tc 50, Zone III, actual size. seems to range from white to brown. In some samples,
the fiber is obviously heavily colored from the sur-
rounding matrix. Artifacts of cotton fiber are described

in Volume II, Chaps. 12 and 13.
Bomhacaceae (Kapok Family)

Ceiha parvifolia Rose, (pochote). Pochote trees are
a characteristic feature of the vegetation in the Te-
huacan Valley. These trees are much smaller than the
giant ceibas or kapok trees of the lowland tropics. They
have served as a source of food from early times, and
present-day residents of the valley seem to relish the
small dark seeds, which are gathered from wild trees
and are eaten raw or boiled. Botli the pods from the
tree and the bulbous storage roots (“jicama de po-
Fig. 158. Fragment of stem of Lemaireocereus hollianus chote”) appear in the deposits. The floss which sur-
from Tc 50, Zone XVI, actual size. rounds the seeds also appears to have been used. In
244
PLANT REMAINS
at least one case, it was employed to enclose a mass of

material which may once have been a dough ground
of
seeds. Much of the floss found in the deposits may have
been discarded from the gathering of seeds.
The earliest fragments of pods were found in Zones
XVI and XIV of Coxcatlan Cave (El Riego phase). In
subsequent phases in this cave pod remains are nu-
merous. Pods and root fragments were also found in
Purron, San Marcos, and El Riego caves.
Cactaceae ( Cactus Family)
Cephalocereus hoppenstedtii (Weber) Schum. (tete-
cho). This columnar cactus grows more abundantly

tall,
along the west side of the valley than along the east
side, but remains of the fruit were found in excavations
Fig. 159. Cardon (Lemaireocereus weberi) does not fruit
on both sides. The fruit appears with some consistency
regularly, but in fruiting years it bears heavily.
in late El Riego and Coxcatlan levels of Coxcatlan
Cave, and again in Zone VII, the Santa Maria com-
ponent. It is absent thereafter. It was also found in a
Mammillaria sp. (biznaga). A few whole plants of the
few of the later zones of Purron, El Riego, and San
small biznaga came from upper levels of the excava-
Marcos caves.
tions. They do not appear to have been modified in
Echinocactus grandis Rose (barrell cactus; biznaga). any manner, and it is impossible to judge for what
A few woolly mats from the growing tip of the barrel purpose they may have been collected. They were
cactus were found in two El Riego phase levels of Cox- found in Palo Blanco and Venta Salada zones of Cox-
catlan Cave and in two Palo Blanco zones of Purron catlan and Purron caves.
Cave. This material may have been gathered for use as
Mijiiillocactus geometrizans (Mart.) Cons, (garam-
packing or padding.
bullo). The fruit is currently used wherever this
Escontria clxiotilla (Weber) Rose (jiotilla). Jiotilla
cactus grows. In the excavations, this species is repre-
fruits are currently gathered in season and appear in sented by several fragments of stem tissue that span a
the local markets in some abundance, but they appar- period from El Riego to Venta Salada times. All were
ently never played a significant part in the diet of the
recovered from Coxcatlan and Purron caves.
people The remains of six fruit were rec-
in early times.
ognized Tc 50, Zone V (Palo Blanco), and one fruit

in
Opuntia sp. (prickly pear; nopal, tuna). The stems
was found in Tc 50, Zone XI (Coxcatlan). and fruit of the prickly pears have long been important
dietary items for the occupants of the Tehuacan
Lemaireocereus hollianus (Weber) B. & R. A single
Valley. Today, every dooryard has a few prickly pear
large spinewas found in Tc 50, Zone XVI (El Riego). A
plants for family use. The people of some villages, such
stem fragment was recovered in Zone II.
as Azumbilla to the north of Tehuacan, grow extensive
Lemaireocereus stellatus (Pfeiff.) B. & R. (xoconoch- fields for nopal (the flat-jointed stems) and tuna (the
tle). Fruit remains are represented by four fragments fruit) to supply local markets. In addition, fruits of many
in Tc Zone VI (Palo Blanco) and four pieces in Tc
50, of the wild species are regularly gathered. The plants
50, Zone IX (Abejas). Although this tree cactus is now are started by planting a joint or two in rows without
cultivated on the western side of the Tehuacan Valley, irrigation or subsequent cultivation. The plants soon
it apparently was not very significant in the diet of the
form a thicket which excludes weedy shrubs. In time,
ancient people of the valley.
an abundant crop of tunas is borne on the upper pads.
Lemaireocereus weberi (Coult.) B. & R. (cardon). The varieties now grown are known as “tuna blanca,”
The fruit is gathered and used today in quantity when which remains a pale green when mature, and “tuna
it is in season. As with the other arboreal cactus fruits, amarilla,” which turns a pale yellow. The cultivated
it did not appear extensively in the cave deposits, being fruits are substantially larger than fruits gathered from
present in five zones of Coxcatlan Cave and one zone wild plants. Some of the latter are red when mature.
of El Riego Cave. The earliest remains date back to In general, it is impossible to distinguish between the
the El Riego phase. vegetative parts of the wild and cultivated stem joints
245
Fig. 161. Fruit and leaves of cosahuico (Sideroxylon cf.

tempisque).
Sapotaccae (Sapote Family)
Fig. 160. Opuntia fUifera in flower and with immature Bumelia laetivirens Hemsl. (tempixquistle). Fruits
fruits. The young pods are eaten as a vegetable. from several genera of Sapotaceae are now grown in
the Tehuacan Valley. Such was not the case in earlier
periods. Tempixquistle is one of the two sapotaceous
of the species with edible fruits. Similarly, unless whole fruits whose remains have been found in the excava-
fruits are available, it is impossible to distinguish be- tions. Seeds appear sporadically in upper levels only
tween wild and cultivated fruit fragments or seeds three in Zone VII of Tc 50, one in Zone A of Tc 272,
from these fragments. Cultivated plants are probably six in Zone B of Tc 254, and one in Zone III of Tc 50.
represented among the archaeological remains, but I The late appearance of this fruit in the deposits indi-
have found no way to recognize or separate them. I cates that it was probably introduced as a cultivar.
suspect that Opuntia may have been one of the earliest There is no marked difference in the sizes of the seeds
cultivars of the Tehuacan Valley farmers because of found, indicating that there was no active selection for
the ease with which the plants are vegetativelv fruit of better quality. If this fruit had been gathered
propagated. from wild trees, remains should have appeared in
The earliest Opuntia remains were mere traces of earlier levels.
seed in levels of the Ajuereado phase. Fruit and stem Sideroxylon cf. tempisque Pitt, (cosahuico). Cosa-
fragments and seed appear regularly thereafter, some- huico has long been one of the more popular fruits of
times in large quantities. Besides almost all levels of the Tehuacan region. The tree was apparently native
Coxcatlan Cave, Opuntia remains were present in Pur- along watercourses in the barrancas, since seeds of
ron, El Riego, and San Marcos caves. in levels of the El Riego phase,
cosahuico appear early,
and continue in increasing numbers throughout the
Xlyrtaceae (Myrtle Family)
sequence. Cosahuico trees were observed in full leaf
Psidium guajava L. (guava; guayaba). Guava is now in some rather dry sites in the winter of 1966, so the
commonly cultivated in the Tehuacan Valley as a door- species must be able to persist under fairly inhospitable
yard tree. It apparently was a late introduction into the conditions.
region, since the only fragments of fruit found are four Some whole seeds were measured
of the recovered
pieces from Purron Cave, Zone E (Palo Blanco). These and an index figure derived by multiplying the length
appear to be from fruit identical with modern seedling- (in cm.) by the width (in cm.). Seeds recovered from
tree fruit. At the period this species appears in the exca- Coxcatlan Cave, Zones XVIII (El Riego phase) and
vations, the Tehuacan Valley was under intensive cul- XII (Coxcatlan phase), averaged respectively 3.12 and
tivation with irrigated fields and fruit trees. Guava was 3.20. These seeds probably were collected from wild
undoubtedly brought into cultivation elsewhere before trees. In the large sample from Zone XI (late Coxcatlan
being introduced into the Tehuacan Valley. phase) of the cave, the average / x tc increased to 3.46,
246
PLANT REMAINS
Fig. 162. Dried fruits (left) and seed of cosahuico, Tc 50,

Zone IX, actual size.
an indication that fruit of large size may have been Fig. 163. Black sapote (Diospyros digyna): fruit, calyx,
selected for propagation. The increase in size con- skin, and separate seeds.
tinues but becomes more gradual in late horizons. The

largest average, 3.88, is for a lot of seeds from Zone D trees requiring supplementary water during the dry
of Purron Cave (Palo Blanco phase). seems obvious
It season, must indicate that the local people knew this
that irrigated orchards of superior fruit were being and were prepared to provide water for these trees.
grown near this site. Black sapote seeds were found in all but one level of
Fruits like cosahuico are not of primary importance Coxcatlan Cave from Zone XI onward. Quantities were
in the diet of the valley people, but are seasonal items not very large except for two instances 50 seeds in —
which provide palatable differences in texture and Zone IX (Abejas phase) and 291 seeds in Zone VI (Palo
flavor from primary diet foods. The appearance of such Blanco phase). Small numbers of black sapote remains
foods in abundance may indicate either that the food were recovered from one level of Purron Cave and the
economy has developed sufficiently to permit time to upper four levels of the East Niche of El Riego Cave.
be spent gathering secondary food items or that the Because each fruit contains a number of seeds in a
plant has been brought into cultivation to provide a thick flesh, the amount of seeds recovered does not re-
plentiful supply of fruit. The presence of 271 cosahuico flect the exact quantity of fruit used.
seeds, together with remains of black and white sapotes
Apocynaceae (Dogbane Family)
and coyol, in Zone XI of Coxcatlan Cave, seems to be
such an indication and suggests that at least a primi- Plumeria rubra L. var. acutifolia (Poir.) Woods, (red
tive form of irrigation may have been practiced as early jasmine). A portion of inflorescence of this lanky
as the Coxcatlan phase. native shrub was recovered from Tc 50, Zone IV (Palo
Blanco). Called “flor de mayo” in some parts of Mex-
Ebcnaceae (Ebont/ Family) ico,the flowers are used as religious decorations. No
Diospyros digyna Jacq. (black sapote; zapote negro). such tradition could be found for this plant near Te-
Black sapote is found today in almost every dooryard huacan, and I suspect the single fragment was an acci-
in the valley where fruit trees are grown. It has long dental intrusion.
been a popular fruit. Black sapote does not occur today Thevetia peruviana (Pers.) Merr. (venenillo). Ven-
as a native member of the flora of the valley, and it enillo is known as a poisonous plant in the Tehuacan
cannot survive in the area without irrigation during region today. Five seeds of this tree, one from Tc 35e,
the dry part of the year. In spite of the fact that the Zone C (Venta Salada), and four from Tc 254, Zone B
ground water level must once have been higher during (Palo Blanco), may have been gathered as medicinal
the wet seasons, the remains of the native flora in the plant material — or they may have been brought into
excavations show that the climatic regimen has appar- the caves for more sinister reasons. The species has no
entlyremained unchanged from about 7000 b.c. to the significance as a food plant, nor is any other useful
present day. Thus, at no time in the past would there product known to be derived from it.
have been sufficient water available during the dry

season to enable black sapote to persist in the valley Solanaceae (Nightshade Family)
without irrigation.Its appearance in the Coxcatlan Capsicum annuum L. (chili pepper). Fragments of
phase (Tc 50, Zone XI), along with other fruits from chili peppers are not particularly numerous, but they
247
Fig. 164. Pod of chili pepper (Capsicum annuum) from Tc

50, Zone II, is indistinguishable from dried chilis sold in
local markets today.
were found as far back as Zones XIX, XVI, and XV of Fig. 165. Cache of tomate or ground cherrv (Phijsalis sp.)
Coxcatlan Cave (El Riego phase) and in coprolites from from Tc 254, actual size.
Zone XIV. Chili is another of the plants which are

native to a more mesic habitat than is available in the
valley now or has been available during the period
covered by the plant remains. Early fragments may Cucurbitaceae (Squash Family)
have come from fruit gathered in the more mesic areas Cucurbita mixta Pang., C. moschata Duchesne, C.
along the barrancas higher in the mountains, in areas pcpo L. (squash, pumpkin; calabaza). Remains of these
where avocados could be gathered. All of the wild
also species and of Lagenaria siceraria (Mol.) Stadl. (bottle
or feral plants of chili which
I have seen bear small gourd), are discussed in Chapter 11 above.
fruit. The large fragment found in Tc 50, Zone XI
(Coxcatlan phase), cannot be distinguished from mod- Compositae (Sunflower Family)

ern cultivated varieties. The plant was almost certainly
Several dried heads of unidentified species of this
in cultivation in Coxcatlan time along with squash,
family were recovered. None of them were the sun-
gourd, and the common bean, which are also repre-
flower which was cultivated by Indians in the south-
sented in the remains of that phase. Chili remains came
western United States. They were probably brought
from Coxcatlan Cave and the late levels of El Riego
into the caves as medicinal plants. The following frag-
Cave. They were not recovered from other excava-
ments were recovered: Tc 35e-C, 3 (Venta Salada);
tions.
Tc 35e-D, 1; Tc 254-B, 3; Tc 254-C 1; Tc 50-V, 2 1
,
Phijsalis sp. (ground cherry; tomate). Two small (Palo Blanco); Tc 50 VII, 1 (Santa Maria).
caches of fruit were recovered from Palo Blanco com-
ponents of San Marcos Cave, 25 fruits from Zone B The Vegetal Remains, Meteorology and Geology
and 28 fruits from Zone C The fruits are small, little
1
. The plant remains from the Tehuaean Project com-
resembling the large-fruited varieties found in the prise one of the most complete histories of the agricul-
Tehuaean market today. They could have been tural pursuits of a group of people which has yet been
gathered from wild plants or they could have been undiscovered. Furthermore, they summarize the local
der cultivation, but they were obviously not highly vegetation and its history as no other method can. From
selected for size. the evidence furnished by the plant remains, inferences
concerning the climate of the area over the last 9000
Bigononiaceae (Bignonia Family)
years can be drawn.
Crescentia cujete (tree gourd; calabaza). Fragments The usual means of estimating the stability of or
of tree gourds are described in Chapter 11 in conjunc- changes in the climate of an area depend upon records
tion with remains of Cucurbita. In addition, a single left in geological formations by the movement of water.
winged seed of an unidentified member of the Bigno- Less frequently, biological evidence in the form of
niaceae was found in Tc 50, Zone III. fossil pollen or the remains of animal skeletons may
248
PLANT REMAINS
provide some index to past climate. Rarely has the Patterns in agricultural development
continuous record of local vegetation been available to From the abundant plant remains, the development
furnish direct evidence of the climate of an area. of agriculture in the Tehuacan Valley may clearly be
From the many reports on the post-Pleistocene of the seen. One most striking examples is the modifica-
of the
north temperate regions of the world, a picture of cli- tion of maize (Z ea mays) under cultivation. It is evident
matic fluctuations may be drawn. Alternating periods that maize was first gathered in the valley as a wild
of more or less rainfall have left their records in varved plant, (seeabove Chapter 9). All evidence suggests that
clays, old shorelines, terraces, and changing pollen to itwas an annual plant of the canyons and barrancas
such a marked degree that Pleistocene geologists have whose seed survived the dry period in dormancy in the
drawn general conclusions stating that post-Pleistocene sun-baked soil. With the coming of the rains in May
climate has fluctuated markedly in North America. and June, the seed germinated, and the plant grew and
With such background, the flora of the Tehuacan area matured with the culmination of the rainy season in
shows no agreement. September. Once the maize plant was brought into
The earliest plant materials from the deposits in cultivation during the Coxcatlan phase, selection
Coxcatlan Cave can be dated roughly 9000 years be- forced rapid modifications in the structure of the grain-
fore the present. From the few remains at this level bearing spike.
and the increasing numbers of native plant species The maize remains tell something of the impact of
found in the later levels of the deposit, it is evident cultivation of a staple food on the diet of a people.
that no change in the species composition of the local For several thousand years (from about 5000 to 1500
flora has occurred. Since the flora of the area is largely while the development of modern maize was tak-
b.c.)
controlled by the lack of available rainfall, even the ing place, the new crop obviously took only a limited
smallest increase in mean annual precipitation would place in the diet of the people. Once the high pro-
result in the elimination of some species at the expense ductivity of maize in an essentially modern form was
of others. In more mesic environments, minor fluctua- well established, the use of maize quickly expanded
tions in rainfall would not have so marked an effect, until it became the staple diet food.
and a small climatic change might go unnoticed in the Setaria macrostachya was another widely used
species composition. The long-sustained uniformity in starch source among the people of the Tehuacan
climate has resulted in the development of a surprising region. This grain was first gathered along with other
amount of endemism in the local flora. Climatic uni- grass seed and then possibly was cultivated in abun-
formity is probably one factor permitting the early dance. From the large, compact heads, thousands of
development of agriculture among the people of the small fruits (caryopses) would be winnowed for grind-
valley, since they would have long known the limita- ing and conversion into gruel or tortillas. Long before
tions of the local rainfall pattern. maize appeared among the plant remains, setaria was a
Dramatic changes in the valley’s flora would have primary starch source. After maize came into cultiva-
occurred had the rainfall pattern of the valley altered tion, setaria continued
to be an important starch source.
from its present regimen. While it is true that the cur- Since there no apparent modification of setaria
is
rent rainfall is not great when the mean annual tem- through selection for larger fruit, no character of the
perature and the rate of evaporation for the area are recovered remains provides a clue to the date at which
considered, the same amount of precipitation spread it may have been brought into cultivation. The first
more evenly through the year would be much more remains of this species were found in late Ajuereado
effective. In such conditions, a nearly complete vegeta- and early El Riego levels of before or about 6500 b.c.
tional cover would probably prevail and most of the At this stage, it was probably gathered from wild stands
rainfallwould then be used by the plants. Under pres- of grass in the barrancas and canyons or along the
ent conditions, which the native plant remains certify watercourses in the valley. I believe that it was prob-
as having prevailed from the beginning of the botanical ably brought into cultivation considerably earlier than
record, much of the precipitation falls within a relatively maize, perhaps as early as the middle of the El Riego
short period of the year. During the prolonged dry phase. The abundance of setaria and other grass re-
season, all of the annual herbaceous plants dry up, and mains Zones XVI-XI of Coxcatlan Cave indicates
in
most of the shrubs and trees lose their foliage. Much that it was available in such quantities that the people
of the precipitation fails to penetrate the soil but rushes could be somewhat wasteful in its use.
off the slopes and down the watercourses of the valley Other plants in the traditional modern Mexican diet
where it is completely unavailable to the plants. of maize, beans, chili, and squash appear in the Tehua-
249
can deposits. Beans and squash are discussed sep- Maguey plants furnish many raw materials for the
arately in Chapters 10 and 11, but a few general obser- some species and bases of
local people: the flowers of
vations may be of interest here. Three species of young leaves of others are foodstuffs which may be
Cucurbita were known to the Tehuacan people as well used raw or cooked; the leaves, split and dried, make
as the bottle gourd, Lagenaria siceraria, and the tree small, very tough boards; the sap is a refreshing bev-
gourd, Crescentia cnjete. Cucurbita mixta and C. erage as agua miel or a mild intoxicant as pulque; the
moscliata are the first cultivated species definitely to inflorescence stalk is a long, straight pole in a region
be identified in the deposits. Two seeds which are otherwise destitute of poles; finally, the fibers of the
apparently C. mixta came from a late El Riego level of leaves may be stripped out in coarse bunches for rough
Coxcatlan Cave and both C. mixta and C. moscliata tying or completely retted and cleaned from young
appear in components of the Coxcatlan phase, as do leaves to furnish fine fiber which may be twisted into
remains of the bottle gourd. C. pepo as a cultivar ap- thread, woven into fabric, or worked into sandals.
parently entered the area later, perhaps in the Ajalpan From the archaeological evidence, it cannot be proved
period and definitely by Palo Blanco times. that maguey was put to all of these uses from its earliest
Bean remains appear under contexts which indicate appearance in the deposits. However, chewed leaf ma-
that all species were known only as cultivated plants. terial in the form of quids appears as early as Zone
Pliaseolus vulgaris fragments, the oldest bean spec- XVIII of CoxcatlanCave, proving the use of maguey
imens, appear first in the Coxcatlan phase and re- as food. The use of maguey sap was almost certainly
occur throughout the sequence. Tepary beans (P. known quite early, and since it does not take agua miel
acutifolius ) appear in an Abejas level and again in the long to become pulque in the Tehuacan climate,
it can
Palo Blanco period. Remains of the Canavalia bean be assumed that pulque did not remain long undis-
appear briefly during the Abejas phase. Runner beans covered.
(P. coccineus) appear in Santa Maria and Venta Sa- Maguey and Opuntia are both gathered and culti-
lada levels. Remains of the small lima bean (P. lunatus ), vated at the present time in the Tehuacan area. Every
were uncovered in Venta Salada levels. dooryard has a few plants. Frequently, they are planted
The chili pepper has long been an established dietary as borders to fence cultivated fields and terraces. Both
item in the Tehuacan area. Fragments of Capsicum sp. are propagated vegetatively, maguey from offshoots and
appeared in the refuse as early as the El Riego phase. Opuntia from stem joints. The botanical characteristics
Although the earliest chili remains might have been of the strain of plants is thus transmitted unchanged
gathered from wild plants growing along the water- from generation to generation, so long as a plant does
courses, Capsicum annuum is not a native plant in the not die completely. Both plants also may be raised from
thorn-scrub forest of the valley. None of the characters seeds, but this form of reproduction is not used in-
of the earliest fragments indicate cultivation, but I be- tentionally by Mexicans today. Selection of superior
lieve that chiliwas among the earlier cultigens in the varieties from a variable cultivated seedling popula-
area, since it would have grown near the occupation tion has never been extensively practiced. Fruits of
sites only with human care. It probably was brought cultivated Opuntia varieties do differ in size and some-
into cultivation in the middle or latter part of the El times in color from fruit of wild forms. Otherwise the
Riego phase. yardstick by which domestication is indicated so clearly
In addition to the maize, beans, chili, and squash in some crops is lacking in Opuntia and maguey.
normally associated with indigenous food preferences, On the other hand, the ease with which both plants
maguey and Opuntia fruits and stems must be con- may be vegetatively propagated leads to the suspicion
sidered. These items are important supplementary that the idea of cultivation resulted from someone’s
foods today; evidence from the Tehuacan cave deposits observing that a carelessly discarded Opuntia stem
indicates that they may once have been major items, joint rooted and grew. From such an observation it is
at least during a portion of the year. One Opuntia a small step to the transplanting of maguey offshoots to
fruitfragment and traces of seeds appear first in the a place where they would be more accessible. Further-
Ajuereado phase. Nopal or stem fragments appear more, both plants are succulent natives of dry hillsides
in the following phase. Fragments of maguey first and porous canyon where they receive a mini-
walls,
appear early in the El Riego phase. Both maguey and mum of moisture. They would survive under conditions
tuna are sources of sugar, and the nopal pads are a that would destroy more tender plants. In spite of
green vegetable. These foods are available during the lack of evidence to prove it, I am convinced that nopal
dry season, when many other foods are not. and maguey are the earliest cultivated plants of the
250
PLANT REMAINS
Tehuacan region. They could have been, and probably found and the custom of preparing and eating them
were, in cultivation by the middle of the El Riego cannot be accurately dated. Since the practice is a
phase. local one known only to the Indian inhabitants of the
Mesquite is another plant which today is both a wild Tehuacan area, it might be a tradition with its roots
food source and grown as a cultigen. Remains of this far in the past. Guaje fragments appear only once in the
plant from the Ajuereado phase place it among the El Riego horizon, and it is by inference that cultiva-
earliest food plants of the Tehuacan people. Along with tion is assumed for the tree by the close of that phase.
setaria,mesquite was proportionally more important in Other early remains are sporadic and very few. In the
the Tehuacan diet before about 4500 b.c. Despite the Santa Maria phase the amount of guaje increased, and
sudden increase in the importance of maize, mesquite the tree was certainly in cultivation from this period
—
unlike setaria maintained its popularity and its place to the present.
in the diet. Although a date for the cultivation of The use by the people of the Tehua-
of fruit crops
mesquite cannot be determined from the remains in can area is well documented by the remains of fruit
the deposits, I believe that mesquite trees probably found in the deposits. Avocado is one of the two earliest
were first cultivated in the El Riego phase, when other fruits used. (The other, as we have noted, is chupan-
native plants of the Tehuacan region may have been dilla.) The avocado pit recovered from Coxcatlan Cave,
brought into cultivation. Zone XXIV (Ajuereado phase, about 7000 b.c.) is the
Accompanying mesquite in the deposits of the earliest record known for this fruit. It probably was
Ajuereado phase is a seed of chupandilla, Cyrtocarpa gathered from a native tree standing along a water-
procera, a fruit tree native to the thorn-scrub forest in course, perhaps in a nearby barranca. The avocado, as
the Tehuacan Valley. From El Riego times chupandilla I have pointed out, is not found in the thom-scrub-
is persistently present among the plant remains and cactus association across the floor of the valley and
formed an important portion of the diet throughout the will not grow away from a permanent water supply.
sequence covered by the deposits. The striking in- The high oil content of the fruit must have made it
crease in the amount of chupandilla remains in Cox- particularly desirable, and it is probable that in the
catlan times may have been
implys that the tree El Riego phase trees were transplanted to watercourses
brought into cultivation form a major constituent of
to near dwelling sites. From this period to the present
the pre-maize food complex. Even after maize became time, the avocado remains a popular dietary item. Note
the staple food plant of the valley, chupandilla main- that most of the avocado seeds were found in levels
tained a prominent place in the diet, as the data for of Coxcatlan and Pun-on caves.
Coxcatlan and Purron caves show. The marked fluctua- Another fruit whose remains were recovered in early
tion of and increase in the size of chupandilla seeds levels is Spondias mombin, or ciruela. Ciruela, once it
from Coxcatlan times to the most recent levels shows is established, will tolerate somewhat drier soil and
that selection for fruit of better quality was actively more adverse conditions than avocado. The earliest
practiced by the farmers of the area. lot of ciruela seeds, 200 number, suggests the gather-
in
The immature pod of the leguminous tree, Leucaena ing of crop fruit. Ciruela is found in relative abun-
csculenta, a rather minor food plant, was gathered dance throughout the deposits. Again the largest
from native trees in El Riego times and is still gathered amounts were recovered from Coxcatlan and Purron
today, although the larger part of the modern crop caves.
comes from cultivated trees planted in dooryards. The The only other fruit with an ancient distribution
pods are called “guaje.” From the same trees, during pattern that undoubtedly included the watercourse
the last few months of the dry season (March, April) margins of the barrancas is cosahuico Sideroxylon cf.
(
people gather “polochoeos,” which are galls or ab- tempiscpie). Cosahuico first appears in levels of the
normal growths of plant tissue caused by flies of the El Riego phase. The remains of this fruit indicate that
group Cecidomyiidae ,“ which lay eggs in the plant it probably was gathered until middle Coxcatlan times,
tissue. Today the galls are prepared by boiling. No when a sudden increase in the number of cosahuico
archaeological remains identifiable as polochoeos were seeds suggests that trees were planted near the sites.
The fruit of cosahuico remained popular throughout

the sequence. During the Coxcatlan phase, cosahuico
° Identified by G. Steyskal, U.S.D.A. Cecidomijiidae are, in
seeds increase in size significantly, suggesting that the
turn, often parasitized by a wasp of the group Torymidae,
Tortjmus sp. (identified by B. D. Burks, U.S.D.A.). Both insects selection of seed for planting may have been underway
are present in the polochoeos when they are eaten. for a long time. Proportionally, the greatest amount of
251
cosahuico material came from Coxcatlan and Purron Finally, peanuts (

Arachis hypogaea) appeared in
caves. Coxcatlan Cave in a Palo Blanco component tentatively
A significant change in the quantity and kind of fruit dated about 100 b.c. A few later remains were present,
remains found in the deposits occurred in the middle but never abundance. The peanut was evidently
in
of the Coxcatlan phase. At this point, the amount of a useful secondary addition to the Tehuacan diet be-
material recovered becomes significantly larger. Also, cause it persisted. The avocado, however, seems to have
three species are added to the list of fruit remains; remained the primary source of vegetable fat for the
coyol (
Acrocomia mcxicana), white sapote ( Casimiroa people of the region.
edulis), and black sapote ( Diospyros digyna). None The late appearance of three of the plants we have
of these species are native to the valley flora, been discussing provides an interesting point for specu-
nor are they known to he present along the water- lation. Peanuts, guava, and a species of Bumelia are
courses. None become truly abundant in the recov- known from archaeological excavations in coastal Peru
ered plant remains (both black and white sapotes at a much earlier date than they were found in Tehua-
have several seeds per fruit so that a number can. It is not certain, but every possibility exists that the
of seeds may represent fewer fruit than do seeds three species were introduced into Mexico from Peru
of avocado, ciruela,and coyol). None
cosahuico, shortly prior to the time at which they appear in the
of these species can grow
Tehuacan Valley with-
in the Tehuacan deposits. I expect to see these plants appear
out a supplementary water supply. The conclusion, in other dry deposits in southern Mexico under con-
then, is that all three species coyol, white sapote, and— texts datable within 200 to 300 years earlier than they
—
black sapote were brought into the area in cultivation appear in the Tehuacan Valley and represented by
and under such circumstances that they were not sufficiently distinct varieties that they can be firmly
needed as a primary food source. The fruits of these connected with Peruvian archaeological varieties.
plants apparently were a pleasant supplement, a luxury The final cultivated plant to be considered is cotton
enjoyed by the people because the development of ag- (Gossypium hirsutum). The earliest cotton finds yet
riculture had progressed sufficiently to permit easier known are the segments of cotton boll from Coxcatlan
accumulation of basic food materials. Furthermore, I Cave, Zone XVI, an El Riego component which can
interpret the presence of these fruits as evidence of the tentatively be dated between 5500 and 6000 b.c. It
controlled use of the valley’s water resources, per- cannot be proved that the fragments were picked from
mitting year-round watering of plants which would a cultivated cotton plant, but neither can it be dis-
not otherwise grow in the Tehuacan region. proved, because other plant remains can be interpreted
Several other plants make brief or late appearances to show that some species apparently were in culti-
in the archaeological remains. One of these, a minor vation in the Tehuacan area at this time. There is,
fruit called tempixquistle Bumelia laetevirens), ap-
(
however, a possibility that the boll fragments are in-
pears first in the Santa Maria horizon and then sporad- trusive, since cotton plant fragments are absent from
ically to the top of the deposit. It is still grown in the the Coxcatlan horizon. They next appear in Abejas and
valley as an occasional tree, but it cannot be said to Ajalpan phase levels of San Marcos Cave and are
be important or popular. The currently ubiquitous present in all subsequent horizons. Artifacts made of
guava appears only once, in a Palo Blanco level of the fiber appear from the Ajalpan phase throughout
Purron Cave, and cannot be classed as an important the remainder of the sequence.
addition. Fruit of a native Physalis sp. was found in
San Marcos Cave in two Palo Blanco levels, but was Significance of the
not found elsewhere. Since the cultivated “tomate,” Tehuacan Vegetational Remains
Physalis, is a prominent plant in the Tehuacan area It is obvious from the evidence presented by the
today, the brief appearance of a related plant raises the vegetational remains found in the Tehuacan excava-
question of the early introduction of tomate into the tions that archaeological plant remains present a true
region. I believe that the San Marcos material was record of the vegetational history of the local area. This
probably a chance cultivation of a local plant which bears out the pattern found in the previously discov-
was sometimes gathered, but it was not found desir- ered dry sites in which plant remains have been pre-
able enough to continue its cultivation. It probably served. In New Mexico, the Bat Cave deposits (Smith
has nothing to do with the currently cultivated species 1950) showed that the inhabitants used only the local
of Physalis, which I feel was brought into the area after plant resources which reflected the local climatic varia-
the Spanish conquest. tions during the time of occupation of that site. Simi-
PLANT REMAINS
larly, the vegetational remains from the Huaca Prieta rect. While these excavations show that the develop-
inPeru (Bird 1948) reflect the local vegetation. Another ment of agriculture among the people of the Sierra
example is provided by the long sequence reported for Madre progressed to a more refined state than it had
Danger Cave, Utah (Jennings 1957). The Tamaulipas farther north and east, the cultigens were still limited.
caves (MacNeish 1958) also disclosed a long sequence Besides the early appearance of Cucurbita pepo (which
in which vegetational remains correlated closely with seems to have been domesticated here), later horizons
the local ecology. In every instance so far reported for showed that these people cultivated C. moschata ,
C.
the American area, no abrupt change from local to mixta, three kinds of beans (
Phaseolus coccineus, P.
foreign plant species has occurred. vulgaris, and P. lunatus ), maize, and chili peppers. In
Man today is notably resistant to changes in habits addition, cotton became a crop plant whose fibers were
and particularly to changes in food preferences. In the utilized for the production of fabrics and so on. It is
United States, vast budgets are expended to present evident from the cultural remains that these people
new products to a sophisticated public well calloused attained a higher cultural level than their neighbors in
by the impact of change, yet the acceptance of any the Sierra de Tamaulipas, but they also accepted no
radically different food product is slow. Far in the past, luxury food items into their agriculture.
when depended on a satisfactory equilibrium
existence No other excavation covering the change from a
establishedby man with his immediate environment, gathering economy to an agricultural economy has
acceptance of introduced and unusual foods would come to my attention. The two Peruvian excavations in
have been even slower. Beyond the fact that no leisure which abundant plant remains have been recorded at
time was available for people to explore widely for and Evans 1952) and
early levels, Viru Valley (Strong
new foods, human inertia forbade the undertaking. Huaca Prieta (Bird 1948; Whitaker and Bird 1949), ex-
Some men may have been, and still are gamblers, but pose sequences in which the lowermost cultural level
few will exchange a sure, but poor, existence for a already represents an agricultural-fishing people. The
grand, but improbable, surfeit. Even were he willing lowest levels of the Huaca Prieta dig contained gourd,
to take the chance, a man’s dependents frequently squash ( Cucurbita moschata ), and cotton remains from
would not let him try. cultivated plants. Huaca Prieta may represent an
Once cultivation of plants has become established earlier period than the Viru Valley levels, although this
as a surer means of producing food, people are released is very difficult to determine from the respective re-
from the tense struggle for basic subsistence. The exca- ports. At any rate, the transition to agriculture was not
vations made to date covering in detail the change from found; the progress of agriculture in later levels of both
a gathering economy to an agricultural economy have sites seems to have been slow. Beans (
Phaseolus vul-
not been many. The excavations of the Tamaulipas garis), maize, avocado, peanuts, chili, and calabaza were
caves covered two such sequences (MacNeish 1958). introduced into the local agriculture, but over-all agri-
However, there were major gaps in the temporal se- cultural advancement apparently was limited. Guava
quence in the excavations in the Sierra de Tamaulipas remains were found in both of these sequences, and it
area which can only be filled by extrapolation from the is probable that this plant was in cultivation in coastal
vegetal remains in the intervening levels. The first cul- Peru. Considering the small complex of local plants
tigen used by these people was the squash, Cucurbita which made up the diet in the lowest levels, the peo-
pepo. The increased food value of cultivated strains of ple of Huaca Prieta and the Viru Valley became quite
this plant could do little immediate
to alleviate the adventurous in accepting such luxury items as avocado
needs, and was not until maize became an established
it and guava into their diet. Whenever it appears, maize
crop that the Tamaulipas people were able to forego cannot be considered a luxury as it rapidly assumes the
some of their gathering activities. Perhaps because of role of prime food source. Both beans and peanuts are
poorer soil and water resources, the Sierra de Tamau- secondary foods in most primitive diets, but they can-
lipas farmers never developed a highly successful sys- not be considered truly luxury since they fill an impor-
tem of farming into which there were repeated tant dietary position through their individual food
introductions of new, largely luxury food items. values. From the standpoint of flavor, under the primi-
The Ocampo caves of Tamaulipas (Whitaker, Cutler, tive conditions under which they were undoubtedly
and MacNeish 1957; Kaplan and MacNeish 1960) pro- prepared, neither could be considered a luxury.
vide longer continuous sequences from which it can be The sequence of development shown in the Tehua-
judged that the extrapolations used in the Sierra de can excavations is unparalleled. From the earliest levels
Tamaulipas interpretations could not be far from cor- of the excavations, the people obviously practiced a
253
gathering economy utilizing a limited number of local (see Chapter 9). Once the value of maize as a diet
plants. The number of plant species involved may be staple and as an efficient yield producer was estab-
deceiving, for it is likely that other plant parts were lished, it rapidly became disseminated. An earlier
utilized in season at locations other than the site of the cultigen may have been the chili pepper, which was
excavation. Even from the vicinity of the excavation apparently among the first cultivated plants. Chili
site, tuna or other might not have found its way
fruit provided flavor in a diet which
must have
originally
into the cave in sufficient quantity to have been recov- been somewhat monotonous. This plant also soon be-
ered in the excavation. came widely disseminated or was quickly brought into
Once agriculture became established in the Tehua- cultivation from wild plants once the techniques of
can region, the development of crops and the introduc- cultivation became known. I am convinced that
tion of new items into cultivation seems to have Opuntia and maguey were introduced into cultivation
proceeded at a rapid pace. I believe the earliest very early in the valley, although the lack of modifica-
Tehuacan cultigens were probably maguey and Opun- tion through vegetative reproduction precludes recog-
tia, owing to the ease with which these species are nition of these as cultivated plants among vegetal
vegetatively propagated and because they are regular remains.
dietaiy items in the area. How long thereafter or under Avocado was first cultivated in the Tehuacan region
what conditions the people discovered that the grain, perhaps as early as 5000 or 6000 b.c., but the influence
setaria, could be cultivated is not shown by the evi- of selection on fruit size is not apparent until the Santa
dence in hand, but probably this species, avocado, and Maria phase of 900-200 b.c. The much longer time
chili peppers must soon have been added to the list of between generations of fruit trees precludes the rapid
cultigens. Thereafter, innovations in local diet and ad- developmental picture which is shown by maize.
ditions to the list of cultigensproceeded at an un- It is possible that ciruela and cosahuico were first
precedented rate. Cucurbita mixta, gourds, ciruela, brought into cultivation in the Tehuacan Valley, and
and perhaps cotton soon were grown in the valley. that mesquite and chupandilla may have been among
The evidence in support of the tremendously impor- the very earliest tree crops. As in the case of avocado,
tant innovation entailed in the controlled use of the ciruela shows little change through a long period after
water supply in Coxcatlan times is secondarily impor- it apparently was brought into domestication. Mesquite
tant in that the introduction of luxury foods occurred remains are essentially the same throughout the de-
at this time. The squash, Cucurbita moschata, which posits. Cosahuico shows an increase in size in early
appears can be considered a

in the deposit at this time, levels, which continues, although more slowly, in later
secondary food with enough value to remove it from levels. Chupandilla shows a definite increase in size
the luxury category. The same cannot be said for black in later horizons, but it has not been found in any
and white sapotes and coyol, which are fruits with archaeological deposit outside of the Tehuacan region,
widely different flavors and little standing as necessary and was probably not widely distributed.
it
dietary items. There can be no doubt that the Tehua- The introduction of peanuts and guava and the use
can people very early arrived at a high stage of sophisti- of tempixquistle about 200 b.c. are exciting and almost
cation in their dietary preferences. Through a rapid certain proof that the people of southern Mexico were
development of agricultural techniques, coupled with in contact with the culture which had developed in
the partial release from climatic variables provided by coastal Peru (Fig. 73). Peanuts and guava were found
irrigation, the early bondage to full-time search for a in Peruvian excavations at a much earlier date, but
bare subsistence was broken. under conditions in which it is certain that peanuts
All of the later additions to the list of Tehuacan were cultivated. Guava could also have been cultivated
cultigens were luxury items, with the possible excep- in Peru during the period in which it is represented.
tion of small lima, or sieva, beans (Phaseolus lunatus). Contacts between the major areas of high culture
The three cultigens, peanuts, guava, and tempixquistle, in North and South America are evident at a much
are items which supplement previous cultigens provid- earlier period, though. With the recovery of Cucurbita
ing essentially the same food values or flavors. The moschata fragments at a very early level in Tehuacan,
fruits, particularly, were superfluous, as there were I am certain that this species originated in North
already several kinds of fruits in cultivation. America and was transported to Peru by 2000 b.c.
Although it probably was not one of the earliest cul- Maize and common beans appear in the archaeological
tigens, maize undoubtedly was brought into domestica- record inPeru about 1000 b.c., after a long period of
tion in the valley during Coxcatlan and Abejas times, use in Mexico. All of these crops represent either a
254
PLANT REMAINS
prime staple (maize) or prominent secondary foods rewarding experience. Many previouslyunknown and
(squash and beans) whose value is imme-
certainly unpredicted facts have been disclosed. More important,
diately apparent to the recipient. This would lead to the entire history of American plant cultigens has been
their appearance in archaeological sites very shortly advanced immeasurably. What had once been empty
after they were known. Regular contacts between conjecture concerning the areas of origin or the routes
North and South America can be assumed by at least of dissemination of several important plant species has
1000 b.c. and, perhaps, as early as 2000 b.c. been proved. It is now possible to say with confidence
The study of the archaeological plant remains from that cultivated plants had many different points of
the Tehuacan Valley has been an exciting and highly origin and differing routes and rates of distribution.
REFERENCES
Anderson, Edgar, 1950. MacNeish and Nelken, 1961.

Baines, 1835. Mangelsdorf et al., 1964.
Bird, 1948. Martinez, 1928, 1937.
Bird and Mahler, 1951-52. Massey and Osborne, 1961.
Blake, 1920. Meggers and Evans, 1963.
Brooks, R. H., et al., 1962. Merrill, 1950
Cardenas, 1948. Miranda, 1942, 1948.
Carter, 1945b. Popenoe, 1941.
Castetter, et al., 1938. Popenoe and Williams, 1947, 1948.
Coe and Flannery, 1964. Rzedowski, 1962.
Coit, 1948. Sauer, 1950.
Cook, O. F„ 1925. Schroeder, 1947.
Cook de Leonard, 1963. Schuchert and Dunbar, 1947.
Cutler, 1960. Sears, 1953, 1955.
Cutler and Whitaker, 1961. Shamel, 1938.
Dewey, 1943. Sharp, 1953.
Engle’, 1963. Skovsted, 1934, 1937.
Gerstel, 1953a, 1953b. Smith, C. E., 1950, 1963, 1965a, 1965b.
Griswold, 1946. Smith, C. E., and MacNeish, 1964.
Gulati and Turner, 1928. Standlev, 1920-26.
Heiser and P. G. Smith, 1953, 1958. Strong and Evans, 1952.
Hill,1952. Towle, 1961.
Hitchcock, 1913. Webber, 1935, 1939.
Howard, 1961. Whitaker, 1956, 1957, 1961.
Howard and Norlindh, 1962. Whitaker and Bird, 1949.
Hutchinson, 1962. Whitaker and Carter, 1961.
Hutchinson et al., 1947. Whitaker et al., 1957.
Jennings, 1957. Whitford, 1943.
Johnston, 1962. Willey and Corbett, 1954.
Kaplan, 1960. Williams, 1952.
Kaplan and MacNeish, 1960. Woodburv, 1961a, 1961b, 1961c.
MacNeish, 1958, 1961a, 1961b, 1962, 1964b.
255
CHAPTER 13
A Cotton Boll Segment from Coxcatlan Cave

Stanley G. Stephens
wo cotton boll segments were uncovered in estimate the number of locks from the relationship 2-n-r
T Zone XVI of Coxcatlan Cave, a level representing

the El Riego phase and dating to about 5500 b.c.
In their report on this material. Smith and MacNeish
= ns.
The ratio s/r

for a four-lock boll,
should be 2.10 for a three-lock
and 1.26 for a five-lock boll. How-
boll, 1.57
(1964) thought that the two segments were apparently ever, during drying and storage the tissues of the boll
from the same boll. Their illustrations show that both may undergo considerable shrinkage and curvature,
segments were as long as or longer than those of a and the question arises whether the foregoing the-
modern Upland cotton. One of the segments was un- oretical relationships would be valid for dried boll
curled with boiling water, and they concluded that segments or whether they would be vitiated by differ-
it “represented one-fifth of a boll about 3 cm. in diam- ential shrinkage.
eter.”Presumably this implies that they interpreted the Some material available in the Genetics Garden at
segment as being part of a five-lock boll. Raleigh, North Carolina, in 1966 was used to test the
The following observations deal with the second relationship. It consisted of an interracial hybrid, RV,
boll segment from Zone XVI, which has not been belonging Gossipium hirsutum. RV in
to the species,
processed or otherwise manipulated. general is similar to modern Upland Cotton varieties,
but is more variable in boll shape and lock number.
Lock Number Bolls varied in shape from nearly round to broadly
A
mature but unopened cotton boll is roughly circu- ellipsoidal in vertical section; the number of locks
lar in cross section,and its interior is divided by vertical varied from three to five. Random samples of dried-
partitions or “septa” into chambers (“locks”). The out bolls were collected from this material, three-,
number of locks varies from three to five. When ripe,
the boll splits vertically along the middle of each lock
from tip to base, and so becomes divided into three
to five segments. If, as is usually the case in archae-
ological material, the segments become completely sep-
arated from each other, it should be possible, the-
oretically, to detennine by measurement whether an
isolated segment was derived from a boll of three, four,
or five locks. This follows from a consideration of the
simple geometry of the boll (Fig. 166). The width of
the septum (r) provides an estimate of the radius of the
boll; the width of the segment (s) multiplied by the
number of locks (n) measures the circumference of the Fig. 166. Geometry of cotton bolls seen in cross section:
boll. Thus, if one can measure both r and s, one can r, width of septum; s, width of segment.
256
A COTTON BOLL SEGMENT FROM COXCATLAN CAVE
Table 27. Maximum Width of Septum (r) and removing the foil, and measuring it flat. The sum of the
Segment (s) and the s/r Ratio of 10 Samples Each two measurements plus the thickness of the septum
of Dried 3-, 4-, and 5-Lock Bolls of RV (hirsutum) (measured by calipers) gives the estimate of 19 mm.
(in mm.) shown in the table. The calculated s/r ratio (1.62) lies
between the values expected for four and five locks
Calculated Theoretical
and closer to the former. This suggests that the Tehua-
r s s/r s/r can boll segment came most probably from a four-lock
boll, possibly from a five-lock boll, and almost certainly
RV 3-lock 8.2 ± 0.35 20.1 + 0.57 2.45 2.10
RV 4-lock 10.3 ± 0.28 17.8 ± 0.55 1.73 1.57 not from a three-lock boll.
RV 5- lock 11.4 ±0.22 15.9 ± 0.35 1.39 1.26
Tehuacan boll 11.8 - 19.0 - 1.62 - Boll Shape

The shape of a green boll is usually expressed in
terms of “boll index” (maximum diameter of the boll
four-, and five-lock bolls being collected separately. expressed as a percentage of boll length). Similarly one
From these random samples, ten segments from each could calculate an “internal lock index” (maximum
—
lock-number category each from different bolls width of the septum, r, expressed as a percentage of
were selected for measurement. The mean values of lock length). Theoretically, twice the “internal lock
r and s measurements of segments sampled from three-, index” should equal the “boll index.”
four-, and five-lock bolls are compared in Table 27. Dried were collected from material available
bolls
Differences between these means are highly significant, in the field. These included:
showing that even after drying, characteristic differ- —
M 8 an Upland variety, similar to Deltapine, with
ences in the dimensions of three-, four-, and five-lock exclusively four-lock bolls, round in shape with a boll
bolls are preserved. Thus, although the samples are index of about 80.
very small, the differences would appear to be real AS 15 —an Egyptian (barbadense) inbred line with
and reliable. exclusively three-lock bolls. Bolls are oval, with a taper-
From these measurements the ratios, s/r, can be ing apex, and a boll index of about 60.
calculated for comparison with the theoretical values. RV — an interracial hirsutum hybrid, described in the
It can be seen that the calculated ratios are of the right previous section, with lock number varying from three
order, but that there is a consistent discrepancy from to five and a boll index intermediate between those of
the theoretical ratios, the calculated values being larger M 8 and AS 15.
than expected. The discrepancy is readily explained From the bolls collected, 10 segments of each variety
by the fact that the calculated ratio does not take into were selected measurement, an attempt being made
for
account the thickness of the boll wall. (Width of septum to choose segments in which the boll walls were strong-
was deliberately measured from the edge of the septum ly recurved, matching as nearly as possible the re-
to the inside of the boll wall, because this can be curved condition of the Tehuacan boll segment. The
measured quite accurately. It is a more practically use- measurements are shown in Table 28. It can be seen
ful measurement than a measurement which includes that the internal lock index ( lOOr/L ) of the Tehuacan
the thickness of the boll wall, since the latter cannot
be measured when the wall is strongly reflexed by dry-
ing.) If one adds the arbitrary but reasonable correction Table 28. Comparative Average Measurements of
of one mm. to each of the r values to allow for thick- 10 Dried Boll Segments Each of Three Cultivated
ness of boll wall, the s/r ratios would have the follow- Forms of Cotton
ing values: three-lock, 2.18; four-lock, 1.58; five-lock, (in mm.)
1.28. These agree remarkably well with the theoretical
Length Width of Length
ratios. of lock septum of beak
(L) 100 r/L 100 b/L
At the foot of the table are shown the measurements (0 (b)
of the boll segment from Coxcatlan Cave. It was pos- AS 15 30.09 ± 0.66 8.69 ± 0.22 7.58 ± 0.28 28.9 25.2
sible to measure r accurately; s could only be measured 3-lock 23.64 ± 1.08 8.25 ± 0.37 5.89 ± 0.47 34.9 24.9
crudely, because of the strong reflexion of the boll wall. RV 4-lock 30.70 ±1.98 10.80 ± 0.29 6.40 ± 0.44 35.2 20.8
'
5-lock 32.49 ±1.55 11.67 ±0.44
It was measured by molding a strip of thin aluminum 5.84 ± 0.39 35.9 18.0
foil to the reflexed surfaces on either side of the septum,

M 8 24.44 ± 0.64 10.04 ± 0.30 6.59 ± 0.34 41.1 27.0
Tehuacan - - -
marking the limits to be measured with a scalpel, 33.30 11.75 9.75 35.3 29.3
257
Fig. 167. Dried segments of cotton bolls. The segment from Coxcatlan Cave, Zone XVI, is at extreme left of middle
row. Top row: 3-lock bolls of AS 15 (barbadense). Middle row: 4-lock (left) and 5-lock (right) bolls of RV (hirsutum).
Bottom row: 4-lock bolls of M 8 Upland (hirsutum).
boll is M 8 and larger than that of

smaller than that of boll segment (see Fig. 167) does not indicate that the
AS 15. The index between the four- and five-lock
lies boll was tapered (maximum width of the septum is
values for RV. It appears that the Tehuacan boll was nearer to the apex than to the base of the boll). More
probably similar in shape to RV; that is, in its original probably the apex of the boll was rounded with an
state it would be ellipsoidal with a boll index of about “abrupt” break.
70. In Fig. 168 are photographs of a range of cultivated
In the same table, apical beak length is also com- forms of hirsutum. The upper row shows Central
pared. The Tehuacan boll has a long beak —consider- American types. The two bolls at the extreme left are
ably longer than in the other material examined and types from Guatemala which resemble closely modem
also relatively longer ( index lOOb/L). In general, bolls Upland The lower row consists of Mexican door-
bolls.
with a tapered apex tend to have longer beaks than do yard forms exclusively. The right-most boll in this row
rounded bolls, but the conformity of the Tehuacan has what I would consider to be a good match for the
258
A COTTON BOLL SEGMENT FROM COXCATLAN CAVE
Fig. 168. Range of hirsutum bolls in present-day primitive cultivation. Top row: Central America. Lower row: Mexico;
the boll at the far right probablv has the same shape as the boll from Coxcatlan Cave.
shape of the Tehuacan boll, including the long beak. septum are a few long fibers which appear to have
It was collected in Laureles, Michoacan, by Dr. L. L. been trapped in the boll. They are flattened and con-
Phillips. voluted and thus presumably could be small samples
of lint from the original boll. It would probably be
Other Observations worthwhile to have these examined by a competent
In all New World cottons, both wild and cultivated, fiber technologist.
with which I am familiar, there is a sharp contrast in In some dried bolls the edge of the septum still bears
texture between the tough boll wall and the thinner, traces of the funiculi (“stalks”) to which the mature
flexible, fibrous septum. In the completely dry boll seeds were attached, and it is thus possible to estimate
the flexibility and the septum often becomes
is lost the number of seeds per lock. In the Tehuacan seg-
brittle. In the Tehuacan boll segment the septum ap- ment, unfortunately, the edge of the septum is worn
pears to be tougher than in present-day New World away and only one attachment can be identified.
cottons. The edge of the septum (the thickest part)
varies from 1.0 mm. to 1.8 mm. as compared with 0.7 General Comments
mm. to 1.4 mm. in the M 8 material I have examined. The Tehuacan cotton boll is comparable in size to
Whether the thickness of the septum is due to accre- some of the largest hirsutum bolls in cultivation today.
tion or impregnation of foreign material or represents a Many types in primitive cultivation today have con-
real difference from modern forms, I am unable to siderably smaller bolls (see Fig. 168). However, we do
guess. Possibly it could be determined by sectioning find large-boiled types in apparently primitive culti-
the septum. vation — in hirsutum, the large-boiled types in Tuxtla
The inner surfaces of the Tehuacan boll segment, par- and Acala, Mexico; in barbadense, the long kidney-
ticularly in the less exposed parts or “pockets” toward seeded bolls which appear to be native to Central
the base and apex seem to be covered
of the segment, Brazil and the Guianas.
with hairlike structures, while the ribs of the exposed The Tehuacan boll was found at a cultural level that
septum and the reflexed surfaces of the boll wall ap- preceded settled cultivation, which would imply that
pear to be smooth. It is possible that the “hairy” sur- the boll was borne on a wild plant. If true, this is a
face may be nothing more than the scuffing and fray- totally unexpected situation. All wild species of Gos-
ing out of the fibrous tissues of which boll wall and sypium today, including wild forms of the cultivated
septa are composed.The internal boll surfaces of all species, have small bolls with small seeds (boll size and
present-day New World cottons are smooth, but hairy seed size are highly correlated). The Tehuacan boll
septa are found in several wild relatives. would be expected to have proportionately large
Toward the base of the segment, on both sides of the seeds, and if the fibers trapped at the base of the seg-
259
ment are original, it had convoluted (i.e., true lint) hairs, before other materials identifiable as cotton appeared;
distinct from the fibers of exclusively wild species. these were found in Zone D of San Marcos Cave (see
Thus, if the boll segments are correctly assigned to Table 26 above). One wonders if the two segments may
Zone XVI, one must assume that about 5500 b.c. there not have fallen into the refuse of Zone XVI from an
was a wild form of Mexican cotton with an exception- overlying level and hence represent a type in primitive
ally large boll quite beyond the range of boll size in all cultivation rather than a bona-fide wild form. In the
other wild species today. However, I understand that absence of a radiocarbon date on the boll fragment
the two segments were found at the back wall of the itself, only the discovery of an independent sample of
cave, in a disturbed area where stratigraphy was diffi- comparable age can settle the question satisfactorily.
cult to establish. Also, there is a gap of about 2000 years
ed. note: Since the above was written, Dr. Stephens has had that the ‘furry’ surfaceis due to scuffing of the long fibrous cells
an opportunity to examine a second segment of cotton boll found of which the lining of the lock is composed. They are often
next to the one he has described. The second specimen had been irregularly branched and pass into the structure of the wall.
pickled for preservation. He notes that agreement between meas- “In Segment I the septum seemed to be thicker and tougher
urements on the pickled boll and similar measurements on the —
than in modern cultivated forms it was not clear whether this
dry boll are not particularly good but observes that there would was due to a heavier structure or impregnation during storage.
be variations between segments from the same boll. He also “In Segment II the septum is translucent and flexible in sharp
reported: contrast with the boll wall — as in modern cultivated cottons.
“In Segment I the inner surface of the segment had a furry “On the whole I can find nothing which would be atypical in
appearance and I wondered whetherthis was due to hairs or a present day cultivated hirsutum cotton. The size is comparable
‘scuffing’ of the surface. The
surface of Segment II appears quite with that of a modern, commercial Upland variety, though prob-
smooth when wet — when it
is surface dried, the fibrous surface ably relatively narrower (ellipsoidal rather than rounded). There
curls up in places, giving the same surface appearance as in are types in native cultivation today in Mexico which would
Segment I. I think it is quite clear that there are no hairs and appear to be very similar.”
REFERENCES
Smith and MacNeish, 1964.
260
CHAPTER 14
Analysis of the Tehuacan Coprolites
Eric O. Callen
he dry conditions which prevail in certain archae- one coprolite if they had been wrapped together in
T ological sites of the

for
New World are responsible
and of
preservation of botanical remains
fecal material, both human and animal. Such desic-
aluminum foil at the excavation
derived from a single large piece.
and appeared
Before treatment the coprolites were assigned suc-

to have
cated material has been found repeatedly during the cessive numbers, starting at one, there being a separate
last fifty or sixty years in archaeological sites of North series for each phase of the Tehuacan cultural se-
and South America, but it is only comparatively re- quence. Each coprolite was inspected and the follow-
cently, with the development of new techniques, that ing information recorded: (1) the location from which
its value has been realized. At first, only seeds and it came— cave, culture, level, and square; (2) general
bones could be identified. Today, plant tissues of many appearance, shape, andsize, color; (3) possible identity
kinds, as well as various animal and insect remains, of donor— human or animal; (4) any vegetal, animal, or
can be identified with certainty. mineral remains adhering to the surface and obviously
Since coprolites contain the actual materials con- not originally part of the coprolite, such as stones,
sumed, analysis of them can give a more precise picture twigs, or bits of reeds or grass; (5) any material incor-
of the actual diet of ancient peoples than has been pos- porated in the coprolite which could be recognized
sible before. Therefore, provided a sufficient number of from the exterior or broken surface, such as bones,
coprolites are available, it should be possible to study seeds, and hair; (6) whether pock-marked with holes
changes in the diet from site to site and from culture usually the emergence holes of coprophagous (dung-
to culture. It should also be possible to obtain informa- inhabiting) insects.
tion on subsistence activities and patterns of the inhab- Thereafter the coprolite was dropped into a wide-
itants of a particular site or of the people of a particular mouth, screw-topped, sixteen-ounce glass jar half-filled
culture. It will be the object of this chapter to try to with a half-percent aqueous solution of trisodium phos-
answer such questions as fully as possible, using the phate. This chemical was used by Van Cleave and Ross
evidence furnished by more than two hundred copro- (1947) to reclaim dried tapeworms, so that they became
lites found in the Tehuacan excavations; 116 of these softand pliable and regained their original size. Callen
were determined to be of human origin. Other evidence and Cameron (1955, 1960) successfully adapted this
concerning subsistence is discussed in the chapters on method to reconstitute coprolites a few years later. The
plant remains, on the three major agricultural plants, coprolite was allowed to soak for a minimum of sev-
and on animal remains. enty-two hours, though more usually for a week or
more, to insure maximum softening. With gentle
Methods and Techniques shaking it would then fall apart.
In this study each portion of feces was considered to Though many of the coprolites regained what must
be one coprolite, whether large or small. Occasionally, have been their original consistency, those that con-
however, one or two small pieces were considered to be sisted largely of bone, hair, cartilage, and meat re-
261
mains remained hard, even after prolonged soaking. terial, the following information was recorded for each
They had a dirty, claylike appearance and had to be coprolite as the extraction was completed: (1) the
broken open with scalpel and forceps. color of the trisodium phosphate after the coprolite
When the coprolite was ready for analysis, the color had been soaked, the smell when the jar was opened,
of the liquid was noted, together with the presence or the presence or absence of a “chemical skin,” and
absence of a scum or “chemical skin” covering the sur- whether or not fungus had started growing on the sur-
face and the smell, if any. After removal of most of the face; (2) the presence or absence of seeds, in what
scum, the liquid was gently decanted, to be used later. amounts and of what type, as well as their possible
The solid matter was then suspended in fresh water identity; (3) the presence or absence of bone and its
and visually inspected bit by bit with the naked eye, identity if possible; (4) the presence or absence of
or by means of a low-power binocular microscope. wood or stones (sand); (5) the different types of ma-
Representative samples of all materials were transferred terial present in the coprolite and their relative abun-
to clean water, to be mounted later. All bones and bone dance; and (6) unusual material present.
chips were extracted, washed in clean w ater, and em- 7
Two card-index systems have been maintained. The
bedded in phenolized glycerine jelly in vials. first contains one card for each microscope slide, on
After inspection, the remaining material plus the which is recorded the identity of all the material on
originally decanted liquid was poured into a conical that slide. The second, a punch-card system, has one
flask with a small spout at one side of the neck. Ben- card for each piece of different material on each slide.
zene was added until the liquid was within a quarter Frequently four pieces of material were mounted on
inch of the side arm. Vigorous shaking for thirty the same slide. If all four pieces were of identical ma-
seconds thoroughly mixed the two liquids and brought terial, say Opuntia epidermis, they would be repre-
the solid matter into contact with the benzene. In sented by one punch card for that slide. If all differed
theory, all insect cuticle wetted by benzene will float from one another, they would be represented by four
at the benzene-water interface, but in practice small punch cards.
pieces of cutinized plant material appear there as well. The abilitv of trisodium phosphate to reconstitute
Thus the material obtained by carefully pouring off material, apparently to its natural size, raises the ques-
the top clear layer of benzene and collecting the inter- tion of whether this “natural size” is actually smaller
face may contain not only small insects, parts of insects, or somewhat larger than normal. This would be im-
pupae, and larvae, but also small seeds, grass glumes, portant if the identity of some pieces of material
pollen grains, and other spores, as well as small pieces hinged on actual measurements. In order to find out,
of epidermisand “chemical skin.” the seed and fruit epidermis of year-old, sun-dried, ripe
Representative samples were extracted from the ben- chili peppers, (Capsicum spp.) of seven varieties, which
zene-water interface material and placed in fresh had been bought in the town markets of Tehuacan and
water. Thereafter, all extracted materials w ere
r
Coxcatlan, were soaked in water and in trisodium phos-
mounted on microscope slides in phenolized glycerine phate for a week. Other seeds from each sample were
jelly, with the benzene extracted material clearly sown in pots in a greenhouse, in order to obtain fresh,
marked as such. Usually a few small bones or bone ripe material for comparison. Statistical analysis of the
chips were mounted, as well as a few chips of stone or measurements of epidermal cells, of seed size, and of
grains of sand, if present. The remaining coprolite seed coat (testa) cell size, showed that there was no
material was considered exhausted for the purpose of difference between living material and dried material
this analysis and was discarded. subjected to soaking in water or trisodium phosphate
In some previous analysis attempts were made to solution.
extract parasite eggs by means of salt-flotation tests Each coprolite, after being described, was placed in
(Marsh 1965). However, so much plant material, espe- a sixteen-ounce glass jar. The jar was then half-filled
cially grass glumes and seeds, floated to the surface with the half-percent aqueous solution of trisodium
that any eggs that might have been present were com- phosphate, and the cap screwed on tightly to preserve
pletely obscured. The results were therefore negative. any odors that might develop. Within thirty seconds
Attempts to keep the slides for even three weeks with- after being immersed, some coprolites started leaking a
out their drying out were quite unsuccessful. I made no brown color into the clear, colorless solution. Others
attempt, therefore, to subject the Tehuacan material to took a little longer, but within ten minutes 95 percent
salt-flotation tests. of the coprolites began to color the liquid. The remain-
Before the actual mounting of the extracted ma- ing 5 percent were putty- or ivory-colored coprolites
ANALYSIS OF THE TEHUACAN COPROLITES
and bone.
that consisted almost entirely of meat, hair, others unstable. Among these latter is tryptophan,
Three or four days of soaking were necessary for the which breaks down into 3-methvl indole, which is the
liquid to develop the full color, ranging from a clear substance responsible for the strong smell. This finally
straw yellow through varying shades of brown to a breaks down into the stable substance indole. There-
deep, opaque orange-brown-black. After about a week fore, it appears that meat coprolites, assisted by the
of soaking, with the minimum of disturbance, a scum of laxative powers of the chili pepper, which rushes them
varying texture frequently developed on the surface of through the alimentary tract before all the amino acids
the trisodium phosphate solution. Occasionally fungi have been extracted, are responsible for the offensive
developed on the scum, or chemical skin. The scum de- smell.
veloped only when coprolites containing a mixture of The human stomach empties in from two to four
meat and vegetal materials were soaked, not when the hours. Any materials eaten during that period will be
contents of the coprolite were almost entirely meat. churned up together, and in due course, will appear
Subsequent analysis has shown that coprolites color- together in the same coprolite. Food normally takes
ing the solution in which they were soaked an orange- twenty-four hours to pass through the alimentary tract,
brown-black color are almost certainly of human origin. but may take up to forty-eight hours to do so. The in-
The pale ivory coprolites consisting mainly of meat re- testine has many folds, in which pieces of material may
mains were interpreted as being of nonhuman origin. be sidetracked for twelve hours or longer. Consequent-
Other criteria used to distinguish human from animal ly, if only a single piece of a certain plant material
coprolites were size and shape,
and content.
smell, turns up in a coprolite, it isquite likely to be a rem-
The question of pH (acid-base balance) was also in- nant of a previous meal. I might add that there is no
vestigated, and while many systems in the human body reason to believe that the human body functioned
maintain a balance within narrow limits, this does not differently 8,000 years ago.
apply to feces. It proved impossible to find data about No foolproof method of distinguishing between
animal material, so that this line of investigation was human and animal coprolites has been discovered. A
negative. combination of the following four criteria has been
The smells that developed from the soaked copro- used: (1) the shape and size of the coprolite; (2) the
lites varied from a musty cave-soil smell, with a slight color of the trisodium phosphate solution after the cop-
hint of a chemical, to the most intense fecal smell pos- rolite has been soaked for seventy-two hours or longer;
sible. Human coprolites containing a good deal of meat, (3) the smell; (4) the contents of the coprolite.
and particularly those containing chili in any quantity,
were the most malodorous. This at first suggested that Identity of Hairs
perhaps intestinal bacteria had survived in spore form, Many of the coprolites containing meat and bone
had germinated during the soaking, and were continu- debris also contained mammal hairs, most of which
ing the job of putrefaction. However, Dr. H. P. A. it was possible with varying degrees of cer-
to identify
Snead (1962), formerly of the National Institute of tainty.** Mammal
have a definite structure, with a
hairs
Medical Research, London, while investigating the central core or medulla surrounded by an outer layer
longevity of bacteria, analyzed some coprolites about or cortex. Externally they show a distinctive scale pat-
3500 years old from Tamaulipas which I sent him. He tern. The medulla contains air pockets or cavities,
found that they were absolutely sterile, as far as in- which are arranged in definite patterns. As the hair
testinal bacteria were concerned. His work showed that becomes older, the membranes separating these air
bacteria could survive only some two hundred to three pockets tend to break down and leave a continuous
hundred years in soil. hollow in the hair. The identification of the hairs
Therefore, the actual decomposition products of the rested, therefore, on the following: the width of
(1)
food eaten, irrespective of the causal agent, must cause
° I am greatly indebted to the following gentlemen who
the smell. In the process of digestion, carbohydrates are
permitted me to obtain hairs as comparative material from
broken down into simple sugars, and none of these has Mesoamerican specimens in their care: Mr. P. M. Youngman,
an unpleasant odor. Fats are broken down into fatty National Museum of Canada, Ottawa; Dr. J. C. Moore, Chicago
Natural History Museum; Dr. George Goodman, American
acids (which Snead 1962 recorded as being present in
Museum of Natural History, New York; and Dr. Kent V. Flan-
the coprolites he tested) but they too would not ac- nery, Smithsonian Institution, formerly with the Tehuacan
Project. I am also indebted to several student assistants in this
count for the unpleasant fecal smell. The third class of
work, particularly to Robin H. M. Cross, who was responsible
foodsis proteins, which the body breaks down into the
for the bulk of the final hair determinations. This work has been
component amino acids. Some of these are stable, aided by a grant from the National Research Council of Canada.
263
Fig. 169. a, Bassariscus astutus, ring-tailed cat, medium hair, Venta Salada coprolite. 2. b,
Hairs identified in coprolites:
B. astutus, coarse hair,Venta Salada coprolite 2. c, Sylvilagus audubonii, Audubon cottontail, medium hair, Venta Salada
coprolite 62. d, S. audubonii, coarse hair, Venta Salada coprolite 62. e, Dipodomys ordii obscurens. Old’s kangaroo rat,
medium hair, Venta Salada animal coprolite 28. f, Lynx rufus, bobcat, medium hair, Venta Salada coprolite 28.
the hair, (2) the width of the cortex, (3) the width of the ber of hairs involved was small, and some were usually
medulla, (4) the pattern of the medulla, and (5) in fine lost in the course of each such operation. However, the
hairs mostly, the external scale pattern. For most hairs glycerine jelly mounts showed up the medulla pattern
the order of importance of the criteria was 4, 1, 3, and very clearly, and the scale pattern on the fine hairs
2; for fine hairs was 5, 1, and 4.
the order showed almost as well without staining.
Four types of hair were distinguishable: (a) coarse, Hairs of the following animals were found in the
the guard hairs; (b) medium, the normal fur; (c) fine, Tehuacan coprolites (nomenclature of The Mammals
the under hair or fur; and ( d the whiskers. The me-
) of North America, by Hall and Kelson, 1959): Sylvila-
dulla pattern was clearest in the medium hairs. In the gus audubonii, Audubon cottontail; S. cunicularius,
coarse hairs the pattern had either developed further Mexican cottontail; Ueterogeomys sp., pocket gopher;
to give very distinctive patterns, or it was in the proc- Dipodomys sp., kangaroo rat; Canis latrans, coyote;
ess of breaking down. The whiskers do not generally Bassariscus astutus, ring-tailed cat; Frocyon lotor, rac-
resemble the medulla pattern of the animal they come coon; Lynx rufus, bobcat; Tayassu tajacu, collared pec-
from, and the scale patternis much less distinct, having cary; and Odocoileus virginianus, white-tailed deer.
been almost worn off in an adult animal. The whiskers
closely resembled human hair. Rev iew of the Literature
Hairs of genera of animals differ from each other, Harshberger (1896) was perhaps the first to remark
though those of the genera of mice, for example, bear a that undigested seeds extracted from human excrement
general resemblance to each other, and are quite dis- might disclose which plants had been used for food.
tinct from those of other genera. Two cottontails have Apparently he did not himself report on any fecal
been, and still are, common in the Tehuacan Valley, analysis. Young (1910) noted quantities of sunflower
both belonging to the genus Sylvilagus. One is S. cuni- seeds in human feces from Mammoth Cave in Ken-
cularius, the Mexican cottontail; the other is S. audu- tucky, and noted especially in the feces of Salts Cave
bonii, the Audubon cottontail. Surprisingly, the two the presence of sunflower seeds, watermelon (?) seeds,
sets of cottontail hairs are quite distinct from each and fragments of hickory shell. He concluded that the
other, though they do have a family resemblance. sunflower was a plant of major economic importance in
The material to be examined had already been the diet. Loud and Harrington (1929) found human and
mounted in phenolized glycerine jelly on microscope coyote feces in Lovelock Cave, Nevada, although the
slides. Attempts to extract the hairs and stain them be- size of the feces (two inches in diameter) rather sug-
fore remounting had to be discontinued since the num- gests something larger than human. In this case the
264
human diet was given as seeds, hulls, and tough plant well as bones of pygmy mouse Baiomys ), deer mouse (
fibers. ( Peromijscus ), lizard, snake, and deer. The coprolites

Volney Jones (1936) found seeds of marsh elder ( Iva ), also contained unidentified bone, eggshell, and feath-
canary grass ( Phalaris ), sunflower ( Helianthus ), and ers. Insects eaten included grasshoppers, bees, and
pigweed or goosefoot ( Chenopodium ), as well as frag- wasps. In addition, found remains of a number of
I
ments of acorn shells and hickory nuts, in human feces other insects, dung feeders or parasites on dung feed-
from Newt Kash Hollow Shelter, Kentucky. Feces ers, which had colonized the feces after they had been
taken from a desiccated body recovered from a bluff deposited at the back of the caves. Among these were
shelter in Arkansas were found by Wakefield and Del- Fantiia scalaris, the latrine fly; Musca domestica, the
linger (1936) to contain seeds of sumac (Rhus), acorns, house and Drosophila, the fruit fly. Also found were
fly;
and a fair amount of charcoal. R. L. Fonner (1951) gave beetles, ants, spiders, mites, and termites. Bird (1943)
a paper before the Michigan Academy of Science, Arts, identified remains of cereal and meat in offering bowls
and Letters, entitled “Human Feces as a Source of by the flies and beetles known to have a dietary affinity
Archaeological Data,” but this was never published. for these substances.
Also in 1951 he delivered another paper, “Plant Re- examined feces from Salts Cave, Ken-
R. A. Yarnell
mains from the Proether Bluff Shelter,” which is men- tucky, and while confirming Young’s (1910) identifica-
tioned in the University of Michigan Occasional Con- tion of sunflower seeds and hickory shells, he was able
tributions of the Museum of Anthropology. It too was to identify Young’s questionable watermelon as squash,
never published. However, a more comprehensive ac- and to identify from seeds some fourteen other food
count of the contents of human feces is that of Jen- plants. He also found fish scales, arthropod remains,
nings in a preliminary report on Danger Cave, Utah, and bone (Watson and Yarnell 1966). In 1964 Martin
who recorded seeds of burro weed ( Allonrolfea ), the and Sharrock gave an account of a pollen analysis of
remains of desert bullrush ( Scirpus ), of mesquite pods human and nonhuman feces from various sites in the
( Prosopis ), hair of deer or antelope, feathers, and small southwestern United States. Frequently some ten or
chips of bone (1953). Further work by Charles C. more genera of plants were identified from each speci-
Sperry and by Robert L. Fonner was done on mam- men tested, some of which were cultigens and others
mal feces from Danger Cave and from Juke Box Cave. wild plants.
This material was incorporated into Jennings’ final Finally, I supplemented my account of the insects
Danger Cave report (1957). Webb and Baby (1957) ex- from the Ocampo Caves, Tamaulipas (Marsh and Cal-
amined human feces from some Kentucky caves, and len 1964, Callen 1965), by adding Tlujlodrias contractus
reported finding seeds of sunflower (
Helianthus ) and Mots., the odd beetle, a scavenger of dried animal ma-
goosefoot (
Chenopodium ), as well as bone, a feather, a terial, which appears to have followed man over the
grasshopper leg, and a fragment of beetle. MacNeish Bering Land Bridge as he migrated to the New World.
(19.58) records breaking open fecal masses from the I have a preliminary account (Callen 1965) of the
also
Ocampo Caves in the Mexican state of Tamaulipas and plants and animals identified from the human feces
finding maguey fibers (Agave), squash seeds (Cucur- recovered from the caves of the Tehuacan Valley of
hita),the remains of grasshopper, and snail shell frag-
ments.
In 1960 Callen and Cameron reported on human
feces from the Huaca and were able to
Prieta, Peru, Table 29. Number of Coprolites per Phase and Cave
identify chili pepper (Capsicum) seeds and tissue, and
Tc 35e Tc 50 Tc 254 Tc 255 Tc 272 Tc 307
tissue of the jack bean Canavalia ), the lima bean
(
(Phaseolus lunatus), and squash ( Cucurbita ), as well as Animal Animal Animal Animal
Human
Animal
Human
Animal
Human Human Human Human
mussel, crab, starfish, and fish. In a subsequent paper,

sea urchin, arthropod, snail, diatoms, and algae were _ _ _ _ _ _
Venta Salada 5 21 9 44 i
added to that list (Callen 1965). Palo Blanco 7 3 33 17 8 11 - - 11 6 - -

- - - - - - - - -
In a preliminary account of the contents of human Santa Marla 18 9 2
-
Ajalpan - - - - 2 3 - - - - -
feces from the Ocampo Caves of Tamaulipas, Mexico
Purron
(Callen 1963), I identified maguey (Agave), aloe (Aloe), Abejas - - 4 - 1 1
- - - 1
chilipepper (Capsicum), safflower (Carthamnus), sun- Coxcatlan - - 12 - - 1

- - - - - -
- - - -
flower (Helianthus), prickly pear cactus (Opuntia), El Riego 6 1
Totals 12 24 82 71 11 16 0 i 11 8 0 1
lima bean ( Phaseolus ), and foxtail millet (S etaria), as
265
Mexico. I will not list these materials here, since they

form part of the substance of the pages that follow.
Sources of Material and Data

Coprolites were obtained from seven of the nine
phases into which the Tehuacan cultural sequence is
divided. All 237 specimens recovered have been proc-

essed,and representative fragments from them have
been mounted on 6,904 microscope slides. The distri-
bution of these materials in the Tehuacan cultural se-
quence appears in the accompanying tabulation. Bones
Human Animal
Cops. Slides Cops. Slides
Venta Salada 14 282 66 2085

Palo Blanco 59 2132 37 735
Santa Maria 18 505 11 329
Ajalpan 2 64 3 28
Abejas 5 152 2 37
Coxcatlan 12 214 1 36 Fig. 170. Setaria, greatly enlarged to show characteristic
El Riego 6 296 1 9 glumes.
Totals 116 3645 121 3259
and bone fragments and other materials too large to

root of the pochote tree Ceiba parvifolia Rose), stem
(
mount have been preserved in vials. These materials
tissue and seeds of various species of cactus, and ma-
are presently housed in the coprolite laboratory of the
guey (Agave) tissue.
Department of Plant Pathologv of McGill University,
Even though the six human coprolites came from
at Macdonald College, Province of Quebec, Canada.
three different zones, they proved to have setaria
all
They will be turned over to the Departamento de Pre-
seed or pochote root, or both, as dominant or co-dom-
historia of the Instituto Nacional de Antropologia e
inant materials. Table 30 shows the more important
Historia in Mexico.
materials mentioned above, plus the chili pepper
The 237 coprolites from the Tehuacan excavations
Capsicum ) and mesquite ( Prosopis ).
were obtained from the following sites: El Riego Cave, (
Naturally it is the seeds and fruits in the diet which

East Niche (Tc 35e), Coxcatlan Cave (Tc 50), San Mar-
will indicate most reliably the season during which a
cos Cave (Tc 254), Tecorral Cave (Tc 255), Purron
cave was occupied. In this case, therefore, since setaria
Cave (Tc 272), and Abejas Cave (Tc 307). Details of
and other grasses including maize, chili pepper, cactus,
the number of coprolites obtained from each cave and
mesquite, and the seeds of a composite (not sunflower)
from each phase are given in Table 29. The totals are
are all present, they indicate wet-season meals. The
broken down as to origin, either animal or human. Al-
presence of grass flowers with stamens in Cop. 2 would
together 116 coprolites were considered to be of hu-
indicate a mid-wet-season meal, while the others are
man origin and 121 of animal origin. Well over half
probably late wet-season ones. Cop. 7, with no seeds
of the total sample came from a single site, Coxcatlan
and only the pochote root and maguey, could
tissues of
Cave.
very well represent a dry-season meal. It should be
El Riego Phase noted also that remains of meat were included in four
Nine from Coxcatlan Cave are as-
distinct levels of the coprolites and that pochote and maguey were
signed to this cultural phase, from earliest to latest, also included in wet-season meals.
Zones XXII-XIV. A total of six human coprolites were Setaria, along with other grass seeds, was the main
recovered from the three latest of these levels, Zones plant represented in the single coprolite from Zone
XVI-XIV, and a single animal coprolite came from XVI. There is one piece of leaf whose trichomes and
Zone XIX. stomata suggests maize. Taken together this may in-
The principal plant materials in the human copro- dicate that seeds of all grasses including wild maize
lites proved to be Setaria seeds, tissues from the starchy were gathered, but that setaria was the most abundant
266
be identified. Cop. 4, which did not appear to contain

meat debris, contained fragments of egg shell, but they
were too small to identify further.
In all the coprolites there were scraps of insect cuti-
cle, probably from small insects which are unlikely to
have formed part of the diet. A single head appears to

be that of an ant (Cop. 4).
Summary
In Coxcatlan Cave during the El Riego phase the
dietary pattern as revealed by the coprolites consisted
of seeds of setaria and other grasses (including maize)
and poehote root as the dominant materials, along with
cactus and maguey and, generally, some meat. The
Fig. 171. Root of poehote shows starch grains, cells, and
setaria and the maguey tissue in some cases were def-
fiber.
initely roasted. The poehote roots were apparently
charred on the outside. The six surviving coprolites of
the El Riego phase seem to represent wet-season meals.
one. The presence
of some unidentifiable roasted plant The amount of setaria and other grass seed being
tissueshows that the grass seeds were not eaten alone. used Zone XIV suggests incipient cultivation and
in
Setaria, with some other grass seeds, again formed also that there may have been more surface water in
the dominant material of the single coprolite from the valley than there is today. Nevertheless, the inhab-
Zone XV, but with the addition of poehote root and itants of Coxcatlan Cave must have traveled a good
cactus tissue. mile or so to reach the Rio Salado where the grasses
The four coprolites from Zone XIV came from the probably grew. Other plant species identified in the
same excavated square and contained the same prin- coprolites were probably wild, but there is some possi-
cipal materials as the coprolite from the previous zone, bility that maguey and Opuntia, which are easily prop-
but with the addition of maguey tissue in some. The agated from offshoots or cuttings, could have been un-
identification of chili pepper ( Capsicum ) seeds in Cop. der primitive cultivation.
6 is notable, since this marks the earliest appearance of It is interesting to note that the seeds of setaria and
this plant in the coprolites. The presence of cactus other grasses were also found in small caches in the
seeds, including Lemaireocereus, in three of the copro- cave floors, suggesting that they were stored for future
probably further indication that they represent
lites is use and then forgotten. The seeds were also scattered
wet-season meals. These four coprolites of Zone XIV —possibly accidentally — on the floors. A preliminary
show that a definite pattern of diet, based on wild study of the breaking of the grains and glumes suggests
plants growing near the cave, was adhered to by the a stirring or a pounding motion (both of which produce
inhabitants of Coxcatlan Cave at the end of the El
Riego phase.
The grass seeds were gathered and prepared in such
way Table 30. Contents of El Riego Phase
a that many were incorpo-
pieces of grass leaves
rated into the mixture eaten. This mixture appears to Coprolites from Tc 50
have been roasted (Cops. 1, 2), but not burned. In at

pepper
least one other coprolite (no. 6), the maguey was very no.
Mesquite
Setaria Poehote
Cactus
As for the poehote root, larger pieces

definitely roasted. Zone
Cop.
Maguey
Chili Meat
containing definitely undigested starch grains do not

XVI 1 t t t
appear to have been roasted, but in most cases charred
XV 2 t t t t
material mixed with it. This rather suggests that after
is XIV 4 t t t*
roots were dug up they were tossed on the fire for a few XIV 5 t t f t t
moments to burn off the outer layers. The heat of this XIV 6 t t t f t
XIV 7 t t t
charring, however, had not swollen the starch grains.
Four coprolites show evidence of meat, but there tPresent.
tDominant or co-dominant.
were no bones and no hairs by which this meat could Tissue and seeds.
267
early dr)' season. Cop. 6 contained, in addition to po-

chote tissue, several kinds of seeds and the remains of
a caterpillar. The seeds would indicate a wet-season
meal. It is interesting to note that black sapote first
appears in the coprolites in exactly the same zone in

which it first appears among the plant remains from
the refuse (Chapter 12).
Cops. 8-10, from another level of Zone XI, also had
pochote root tissue as a dominant material, with se-
taria seed co-dominant in Cop. 8 and Lemaireocereus
tissue co-dominant in Cop. 10. Setaria was present in
Cops. 9 and 10 along with cactus tissue and (in Cop.
9) also some cactus seeds. Cops. 9 and 10 may represent
an early dry-season diet. However, Cop. 8, containing
Fig. 172. The epidermis of Opuntia.
pochote and meat, rather suggests a wet-
root, setaria,
season meal, if the setaria were eaten fresh, as does
the same type of fractures), but not a rolling one. A Cop. 4, with setaria dominant among pochote tissue
stone mortar and pestle, recovered from the El Riego and meat. Cop. 16, the final sample from this level, con-
phase, was used for the study (Callen, in press). tained a dominant amount of setaria along with meat,
black-sapote remains, and chili pepper ( Capsicum )
Coxcatlan Phase tissue. This combination represents a late wet-season or

The twelve human eoprolites representing this phase an early dry-season meal.
were recovered from Zones XIII-XI of Coxcatlan Cave. Cops. 11 and 12, from a third level of Zone XI, con-
The single animal coprolite came from Zone F of San tained pochote root tissue as dominant plant material,
Marcos Cave. The principal plant materials in the hu- with meat co-dominant in Cop. 12.The presence of
man eoprolites proved to be setaria seeds; the starchy setaria seeds in both, along with maguey and cactus
root of pochote; stem tissue and fruits of cactus, in- tissue, probably indicates early dry-season meals.
cluding Opuntia and Lemaireocereus; maguey tissue; The single coprolite from Zone XIII showed the same
black sapote fruits; the fleshy leaves of yucca; and pattern of contents established in coprolites from Zone
maize seeds. XIV of the El Riego phase, with pochote root tissue
The single coprolite (no. 5) from Zone XIII contained the dominant vegetal material among setaria and other
pochote root, setaria seeds, what appeared to be part grass seeds, including part of a maize kernel. Cop. 7
of a single grain of maize, and meat. If eaten fresh, the from Zone XII demonstrated a common dry-season
setaria and maize would represent a late wet-season meal pattern that continued into the Santa Maria phase
meal. — a mixture of maguey, pochote root, and meat.
Zone XII is also represented by a single coprolite
(no. 7). Maguey and pochote tissues are present in
about equal quantities along with meat. The lack of
other materials would seem to indicate a dry-season
meal.
The remaining eoprolites are from Zone XI and can
be subdivided according to different levels within the
zone. Coprolites 6, 14, and 15 came from the same level.
All contained a large amount of pochote root tissue.
However, Cop. 14 also contained black sapote ( Dio
spyros) as co-dominant material, as well as a single
setaria seed, some maguey tissue, and meat. The fruit
of the black sapote ripens in the fall, and its presence
definitely indicates a late wet-season or early dry-sea-
son meal. Cop. 15 contained a little black-sapote tissue
as well as cactus, meat, and two small unidentified Fig. 173. Tissue of Lemaireocereus. Note large crystal at
seeds —a combination that indicates the late wet or left.
268
Table 31 . Contents of Coxcatlan Phase debris was present in all except Cop. 8. This does not
Coprolites from Tc 50 mean had been eaten at each
that large quantities
meal, but rather that there were traces of meat in each
coprolite. There were no bones or hairs by which the
pepper sapote
very definitely roasted meat of Cop. 12 could be identi-

cactus
Lemaireocereus
no.
fied. The only coprolite with bones was Cop. 11, a po-
CL) Pochote
Setaria
ss
Maguey Chili Maize
MO
Other Black
Cop.
chote and maguey meal, where the bone pieces sug-

XIII 5 t t t t gested an animal larger than a rabbit. A few hairs were
XII 7 t t t
present in Cop. 16 which proved to come from an
XI
level 7 6 t t t t* t
Audubon cottontail.
14 t t t t t Insect remains were recovered from four coprolites
15 t t* t t
of Zone XI. Cop. 4 contained pupae of a coprophagous
level 6 4 t t t
insect. Cop. 6 contained the remains of a medium-sized
8 t t t
9 t t t t* t
butterfly caterpillar, which was apparently responsible
10 t t t t* t for much of the meat debris found in the coprolite. It
16 t t
t t
also contained an entire small insect, as yet unidenti-
level 4 11 t t t t
fied. Cop. 11 contained a Drosophila, or fruit fly, larva,
12 t t r* t
and some insect cuticle. Cop. 12 contained many ant

tPresent.
heads, some ticks, and a number of both small and
(Dominant or co-dominant.
"Tissue and seeds. medium-sized larvae of unknown type. This was the
coprolite that also contained the roasted meat. The ant
heads present a tantalizing question: Are they from
The Zone XI coprolites, like those of Zone XIII, honey-pot ants? A special study of ant heads at a later
showed the pattern carried over from El Riego times. date may answer this question. Nothing suggesting the
Pochote root is the dominant or co-dominant material remains of the swollen abdomen of ants was found.
of the diet, appearing along with other plant tissue. Cop. 16 also contained a little insect cuticle of unknown
There is only a little cereal in the form of setaria and origin.
other grass seed. The other plant tissue might be Le-
maireocereus, Opuntia, or maguey. Usually some meat Summary
is included in the diet. Coxcatlan Cave during the Coxcatlan phase was
The two coprolites from Zones XIII and XII both occupied in both the wet and dry seasons. The single
contained charred material in small quantities. Some coprolite from Zone XIII probably represented a wet-
of the pochote tissue in both was obviously charred. It season meal. The one from Zone XII, on the other
seems certain that, as in El Riego times, the pochote hand, was probably the residue of a dry-season meal.
rootswere flung on the embers before being eaten, in The other coprolites, all from Zone XI, represented
order to burn off the outer corky layers. Some of the meals from both wet and dry seasons. The late wet-
maguey epidermis in Cop. 7 also had definitely been season diet of pochote root and setaria noted in late
roasted. El Riego times continued in use. At the same time a
The eight coprolites from Zone XI that contained maguey and pochote
dry-season diet of root, along with
charred material also contained pochote root tissue. some meat, became evident during the Coxcatlan
The one that lacked pochote (Cop. 16) contained very phase.
definitely roasted and charred setaria seeds, as well as Pochote root was the most important single plant
some other unidentifiable charred plant material. The human coprolites from the Cox-
material found in the
co-dominant pochote and setaria materials in Cop. 8 catlan phase. The pochote tree Ceiba parvifolia) grows
(
were definitely roasted. Although meat was present in wild in the valley, mainly in the thorn-scrub vegetation
most coprolites, only in Cop. 12 were there definite on the eastern Lemaireocereus and Opuntia are
side.
signs that it had been roasted. The seeds of Lemaireo- also native to the vegetation around Coxcatlan Cave.
cereus from Cop. 6 were definitely roasted, and in fact However, setaria, the second most important plant in
were almost charred. the diet during this phase, must have grown along the
Meat debris was present in the two coprolites from Rio Salado, a mile or two from Coxcatlan Cave. Ma-
Zones XII and XIII, but there were no bones or hairs guey seems to have played a much less important role
to indicate its origin. In the Zone XI coprolites, meat in the diet of this phase than the various cacti did.
269
The black sapote (

Diospyros ) appears for the first bris. This combination suggests a late wet-season or
time in the coprolites of this phase. In fact, the black early dry-season meal.
sapote must have been brought into the Tehuaean Although pochote root played an important role in
Valley at this time. Chili pepper appears again in this the diet of the El Riego and Coxeatlan phases, it is
phase and may have come from a cultivated plant. not so prominent in coprolites of the Abejas phase, pos-
Surprisingly, none of the annual plants believed to have sibly because the sample is too small. In Zone IX two
been first cultivated —beans (
Phaseolus and squash kinds of cactus and a little maguey were the plants in
( Cucurbita ) —
appear in the human coprolites of this one meal, and squash was almost entirely the substance
phase.® of the other. It is worth noting that the whole squash
Methods of preparing food do not seem to have had been eaten, as both fruit tissue and chewed frag-
changed. In this period, too, the pochote roots appar- ments of seed were found.
ently were flung on the fire to burn off the outer layers, In Zone VIII the combination in Cop. 4 was maguey,
but the heat did not affect the starch grains in the cen- cactus, a very little pochote root tissue, and some meat
ter. As much of the starch was still contained within the — a diet that was firmly established by Santa Maria
cells, there appears to have been no beating to break times.Cop. 6 from this zone contained predominantly
up the cells to allow the starch to escape. The setaria meat few fragments of mesquite and other
tissue with a
grains and glumes, on the other hand, had been broken tree-legume pods. The pods indicate a wet-season meal.
down mechanically, probably with mortar and pestle as The Cucurbita in Cop. 7 from Zone IX had definitely
in the previous phase. The resultant material had been been roasted. Both the exocarp and mesocarp showed
roasted and even charred a little —or perhaps the seeds the typical browning associated with this type of cook-
were first parched for better storage and were later ing. from the same zone, contained roasted
Cop. 5,
ground. Lcmaireocereus seeds also were roasted, and maguey and some charred Opuntia tissue which
tissue
maguey was roasted occasionally. In only one coprolite contained concentrated amounts of druses, suggesting
was there evidence of the roasting of meat. There was that the subepidermal tissues had been scraped to-
no evidence of food prepared by boiling. gether for eating purposes and the epidermis discarded,
as none was present in the coprolite. Both coprolites in
Abejas Phase Zone VIII contained roasted or charred plant material
This phaseis represented by seven coprolites, five
which could not be identified. Perhaps some charred
bits of pochote root were eaten with the uncharred tis-
of which are of human origin and two of animal origin.
Four of the human coprolites came from Zones IX and sue in Cop. 4.
VIII of Coxeatlan Cave. One human and one animal Bone occurred in three of the four coprolites. The
coprolite were recovered from Zone D of San Marcos small bones in Cop. 5 seem definitely to be mouse
Cave. A single animal coprolite came from Abejas bones, but there are no hairs to confirm this identifica-
Cave. tion. In Cop. 6 the bone chips and fragments are those
of an animal larger than a rabbit, but again no hairs
Coxeatlan Cave were present to help identify the victim.
The principal plant materials in the human coprolites

from this site provedbe cactus tissue, including
to
Opuntia tissue; maguey tissue; pochote root tissue; and
squash ( Cucurbita ) fruit tissue and seeds. Table 32. Contents of Abejas Phase Coprolites
Cop. 5 from Zone IX contained mainly cactus tissue,
zone
sapote
including Opuntia and meat debris. There was also

,
cactus
leaf
and no.
Cucurbita
some maguey tissue, a very little pochote root tissue,

Mesquite
Opuntia
Pochote
Site
Maguey Other Black Maize
s
Cop.
and a small sliver of wood. The contents together indi-

cate a dry-season meal. Cop. 7, from the same zone, Tc 50, IX 5 t t t t t t
consisted of Cucurbita material, with small scraps of 7 t t
charred grass or cane epidermis, plus a little meat de- Tc t t t

50, VIII 4 t t t
6 t t
6
A seed identified as Cucurbita mixta (see Chapter 11 above) Tc 254. D 3 t t t t
appears in a coprolite from Zone XIII. This coprolite forms part
of the exhibition of the archaeological museum in Tehuaean tPresent.
and therefore was not available for this study. ed. tDominant or co-dominant.
270
Fig. 175. Maguey epidermis.
bones or hairs to give a clue as to the identity of the

animal. None of the foods eaten showed evidence of
charring or roasting or other indication of preparation.
The only insect remains recovered from the San
Marcos coprolites, human and animal, were some chi-
tin of unknown origin and an ant head and leg.
Summary
Human coprolites representing the Abejas phase
were recovered from two caves. They represent a diet
Fig. 174. Maguey tissue. Above: unroasted; note clear of maguey and meat, supplemented by cactus. This
elongated crystals. Below: roasted, with shattered crystals. combination, as a dry-season diet, persists into Santa
Maria times. Pochote root tissue was not prominent,
perhaps because the number of coprolites was too small
The coprolites from Coxcatlan Cave in the Abejas
for an adequate sample of the diet.
phase, then, suggest mainly a dry-season diet in which
Roasting appears to have been the main method of
pochote root did not play a major role. This diet, con-
cooking. The squash in Cop. 7 had been roasted, al-
sisting mainly of maguey and cactus tissues, continued
though the seeds were quite untouched by the cook-
to be used in succeeding phases. Cucurbita appeared
for the first time in the coprolites which I examined.
ing. The seeds had been chewed —
one of the very few
occasions when they have been found in this condition
The whole fruit was roasted and eaten, including the
in the Tehuacan coprolites. Maguey and Opuntia had
seeds.
also been roasted. The absence of the Opuntia epi-
San Marcos Cave dermis suggests that the starchy interior tissue of the
stems were eaten and the charred epidermis discarded.
The principal plant materials recovered from the
The subepidermal layer of druses was obviously
human coprolite in Zone D were cactus tissue, grass
scraped together before eating. There was no evidence
leaves,and maguey tissue. The presence of a very few
of roasting of meat.
grass and Opuntia seeds suggests a late wet-season or
even an early dry-season meal, a diagnosis confirmed Ajalpan Phase
by the presence of black-sapote tissue. The cactus and
All five coprolites representing the Ajalpan phase are
maguey combination is the fairly standard dry-season
from San Marcos Cave.* Two are of human origin and
diet noted in the Coxcatlan Cave coprolites of this
three of animal origin.
period. The presence of the few seeds, however, would
make it either a late wet-season or early dry-season
"No material representing this phase was recovered from
meal here.
Coxcatlan Cave, which was unoccupied during this early
Meat was eaten with this meal, but there were no Formative period. ed.
The principal plant materials in the two human cop-

rolitesproved to be maguey tissue, Lemaireocereus
seeds, and black sapote fruits. Both coprolites con-
tained roasted and charred maguey tissue as domi-
nant material, along with Lemaireocereus and other
cactus seeds. One also contained some black sapote re-
mains and an unidentified seed, indicating a late wet-
season meal, or perhaps an early dry-season one. The
other contained some additional unidentifiable plant
material, but was most likely a late wet-season
it too
meal. The maguey tissue had been roasted and charred,
but the other plant material was apparently eaten raw.
Fig. 176. Maize pericarp shows few distinctive features.
In one coprolite the teeth and lower jaw of a mouse
were found, along with other bones. The other copro-
lite contained meat debris but no bones or hair by 21 from another level had the same co-dominants. In
which the meat could be identified. There were also a the former there was also a little Cucurhita tissue, in
few insect remains: material that looked like pseudo- the latter a little setaria. These would probably repre-
scorpion chitin, as well as insect chitin which could not sent late wet-season meals.
be identified. In level 7 there were two coprolites. Cop. 1 contained
Lemaireocereus and setaria as co-dominant materials,
Summon/ and Piu/salis (ground cherry) as sub-dominant. It also
Plants generally regarded as cultivars which were contained several other seeds, such as bean and ama-
eaten during this phase include maize and black sa- ranth. This combination points to a wet-season meal.
pote. Lemaireocereus and maguey were probably from Cop. 14, by the time it reached the laboratory, had al-
wild plants. The almost complete absence of setaria ready broken down into dust. It contained Cucurhita,
seeds and pochote root in the Ajalpan zone of San but no setaria seed or pochote root. Some tree-legume
Marcos Cave, as in the Abejas levels of both San Mar- pod tissue and maize pericarp suggest a wet-season
cos and Coxcatlan caves, is of interest, but the inade- meal.
quateness of the sample makes theorizing unsatis- In the next level there were two coprolites (nos. 2 and
factory. 11) with eight plants in common. Four of these plants
With regard to food preparation, maguey was were represented by seeds and fruits. Setaria was the
roasted and charred as in previous cultures. The pres- dominant material in Cop. 2 and pochote root in Cop.
ence of mouse bones but no hairs might indicate that 11 with maize subdominant, but both must have been
the animals were skinned or singed before they were wet-season meals.
eaten. From level 5 eight coprolites were recovered, of
which five came from the same square. Here pochote
Santa Maria Phase root was dominant or co-dominant along with setaria or
All but two of the twenty-nine coprolites from this cactus (including Lemaireocereus). Black-sapote fruits,
phase were recovered from Zone VII of Coxcatlan mesquite pods, and various seeds indicate late wet-
Cave. Of these, eighteen are of human origin and nine season meals as well.
of animal origin. The two others, from Purron Cave, The five coprolites from level 4, containing setaria,
are of animal origin. Lemaireocereus, mesquite, black sapote, and other
The principal plant materials in the human coprolites seeds, all must have been late wet-season meals.
proved to be setaria seeds, tissue of pochote root and It would appear, then, that Zone MI was occupied at
maguey, Lemaireocereus and other cactus tissue, least during the wet season.
black-sapote mesquite pods, and the starchy
fruit, With regard to the use of plants in this phase, fair
tissue of what appears to be cassava ( Manihot ). amounts of setaria were still eaten. It was the dominant
Zone VII is the only Coxcatlan Cave component of or co-dominant material in five coprolites and was
the Santa Maria phase. It was excavated in eight levels, present in all the others except one. It was eaten in com-
five of which contained human coprolites. In level 8 bination with pochote root or maguey. This has no
there was a single coprolite (no. with pochote root
12) seasonal significance, as black sapote and Lemaireo-
and Lemaireocereus as co-dominant materials. Cop. cereus were eaten with both combinations, which
272
Table 33. Contents of Selected Santa Maria Phase

Coprolites from Tc 50, Zone VII
I
,
I
11.11
1 I I I 1 I
17 t t t t t
18 t t t t
19 t i t } t
20 t t t
21 t t t t
5 3 t t
4 t I t t t
5 t t t t
6 t t t t t
Fig. 177. Bird or biting louse (Mallophaga)
7 t t t* t t
8 t t t t t t
9 t t t* t t t
10 t t 1* t
fPresent.
tDominant or co-dominant.
‘Tissue and seeds.
means they come from late wet-season meals, or possibly

meals of the early dry-season. It is worth noting that
black sapote and Lemaireocereus had by now become
regular items of the diet in Coxcatlan Cave.
The continued use dominant
of pochote root as the
or co-dominant material in a meal in what we shall call
the “cave diet,” is of interest, as this must represent a
diet based on less sophisticated plants. Squash, and
particularly maize, had become important cultigens by Fig. 178. Feather, possibly of a turkey.
this time, but they make only a token appearance in the
0
coprolites.
The appearance of an epidermis resembling that of
Five of the eighteen coprolites lacked remains of
Physalis fruits is of interest here, as is a starchy tissue
meat. Of the five, one was a wet-season coprolite con-
that greatlv resembles Manihot (cassava). The former taining fruits and seeds; another was probably a dry-
species does not appear among the refuse remains un-
season one containing chiefly Lemaireocereus tissue.
til the subsequent phase.
Cops. 6 and 8 contained mouse bones, including verte-
The pattern of food preparation noted in El Riego
brae, but no hairs were present to confirm the diagnosis.
and Coxcatlan times maintained in this period.
is still
Cop. 11 contained bone chips of an animal larger than
The pochote root, still a dominant food in most copro-
a rabbit, but again no hairs were present.
lites, shows definite evidence of roasting and charring.
Meat debris of insect origin can generally be dis-
The setaria in Cop. 1 had been charred, and in some
tinguished from that of mammal origin, and some was
of the others there is a suggestion that it had been
present in the human coprolites.
roasted.
Feathers, possibly those of turkey, appear in two
coprolites, but chiefly in Cop. 14, along with meat
°
Evidence provided by the plant remains among the refuse
indicates that maize was probably used universally by the Santa
debris and unidentified bone.
Maria phase. If maize kernels were boiled with ashes or lime to The insect remains recovered from the human and
remove the pericarp and the resulting hulled corn ground to a animal coprolites were mainly in the form of pupae of
very fine meal, or paste, as in the making of tortillas in historic
time, there would be no identifiable trace to appear in the dung-inhabiting insects. The remains of small adult
coprolites. — ed. beetles were recovered from two coprolites, Drosojihila
273
larvae from two others,

and the larva of Fatmia scalaris, charring, and that at least occasionally setaria seed is
the latrine from another. Chewing lice were re-

fly, still parched. Although maguey was roasted, it does not
covered from one human coprolite (no. 14), containing appear frequently in this phase.
bird remains, and from one animal coprolite (no. 25). Physalis was apparently eaten raw, as a fruit, but
These lice belong to the Order Mallophaga, which gen- at times evidence of roasting seemed to be present. The
erally parasitize birds. material identified as cassavashowed no sign of being
In summary the coprolites from Zone VII of Coxca- roasted, norwas there evidence of the roasting of meat.
tlan Cave suggest: (1) a wet-season diet of a variety of The absence of hairs from coprolites containing meat
fruits and seeds (including setaria) with a little plant debris and bone suggests that skinning had become the
tissue and occasionally supplemented with meat; (2) fashion and that greater care was now taken in the
an early dry-season diet with an occasional fruit or a skinning.
few seeds, but mainly plant tissues and meat. It ap-
pears to have been the usual custom to eat maguey Palo Blanco Phase
with meat; frequently such a meal was supplemented
Secular and religious cities had come into existence
by pochote root. Setaria and other grasses were prob-
at this period, and the caves were probably used only
ably eaten when ripe. Cucurbita and maize appeared
occasionally for human habitation. It is therefore not
only three times each in eighteen coprolites, tree-legume
surprising to discover that a large proportion of the
pods (other than mesquite) also appeared three times,
coprolites recovered from the caves are of nonhuman
and amaranth and chili pepper twice each.
origin. Fifty-nine coprolites of human origin were re-
Summary covered, of which thirty-three came from Coxcatlan

Cave, eight from San Marcos Cave, seven from El
The pochote root and setaria diet, seen in coprolites
Riego Cave, and eleven from Purron Cave. Thirty-
from the El Riego, Coxcatlan, and Abejas phases of
seven animal coprolites were recovered.
Coxcatlan Cave, is still represented in this phase, even
after the lapse of 2000 years or so during which the cave
Coxcatlan Cave
apparently was unoccupied. Lcmaireocereus and the
black sapote, which had appeared in the coprolites in The principal plant materials from the thirty-four
the Coxcatlan phase, are still important plants in the human coprolites proved to be the starchy root of po-
diet. On the other hand, maguey and cactus, which had chote, setaria seeds, maguey tissue, black sapote fruit
constituted the major portion of the diet represented by tissue, Opuntia tissue and fruit, and Lcmaireocereus
coprolites from the Abejas and Ajalpan phases of San tissue and fruit. This phase is represented by Zones VI,
Marcos Cave, play a relatively unimportant role in the V, and IV of Coxcatlan Cave, from all of which copro-
diet during the Santa Maria phase represented by lites were secured.
coprolites from Coxcatlan Cave. Twenty-one human coprolites were found in Zone
Maize and the cucurbits have an insignificant role in VI, of which nineteen were from the same square and
the diet, which is therefore designated the “cave diet,” level. In view of the contents, they must represent the
to distinguish it from the diet which must have pre- meals of sundry individuals over a period of several
vailed in the villages and towns at this period. Plu/salis months. Pochote root tissues formed the dominant or
and what is apparently cassava make their first appear- co-dominant material in twelve of them. Although se-
ance here in the coprolites. taria seedwas present in all but two, it was dominant
The meat in coprolites cannot always be identified in onlytwo of the nineteen (Table 34). These latter two
because hairs, on which identification is based, mav not probablv represent late wet-season meals. Those con-
be present. However, mouse bones were identified sev- taining black sapote represent either late wet-season
eral times. In a few other coprolites chips
bones in- of or early drv-season meals. The cactus seeds present in
dicated animals larger than rabbits. Feathers from two a number of these coprolites again indicate wet-season
coprolites are possibly those of turkey. Identified in- meals: Lcmaireocereus possibly early to middle wet
sects include chewing lice (
Mallophaga ), beetles, and season, and Opuntia possibly late wet season. Opuntia,
larvae of Drosophila and Fannia ,
and pupae of other however, is not considered such a reliable indicator as
insects. other cacti, because it fruits over a long period. Cop.
The pattern of roasting, first noted in coprolites of 42, alsofrom Zone VI, contained cactus tissue without
the El Riego phase, is maintained. This means that po- seeds and no meat, and was probably a dry-season
chote root still shows definite evidence of roasting
O and meal. Cop. 81 consisted mainly of mesquite pods and
Table 34. Contents of Selected Palo Blanco

Phase Cop rolites
zone
pepper sapote
cactus
seeds
Lemaireocereus
and no.
Composite
Mesquite
Pochote Opuntia
Setaria
Site
Cop.
Maguey Other Chili Black Maize Other
s
Tc 35e, D 45 4 t
49 t 4
50 t 4 4
51 t 4 4 4 4
52 4 4 4
53 t 4 4 4 4
86 t 4 4 t
Tc 50, VI 19 t t 4
20 4 t
21 4 t t Fig. 179. Cucurbita exocarp.

22 t t t* 4 t
23 t 4 t 4 4
24 t t f* 4
25 t t 4 4
seeds, none numbers. This meal would rep-
in great
26 4 t 4* 4 t resent the latewet or early dry season.
27 t t 4
The other two eoprolites of Zone V are of particular
4
28 t t 4 4 4
29 t 4 t interest. Cop. 94 contained Phaseohis (common bean),

30 t 4 t*
pineapple, chopped lily buds, and roasted maguey
t t
31 t t t* |
32 t t 4 4 tissue and had a smell which suggested that large

33 4*
t t t t
quantities of “beer” had been consumed. The flower
34 t t 4 4
35 t 4 4 4 4 4 buds may have been those of maguey or yucca, which

36 4"
t t t 4 4
are eaten today. Cop. 74 contained maize and roasted
37 t 4 t 4* 4 f
maguey tissue as co-dominant materials, plus squash
Tc 251 B 6 t t* 4 4 4 4
92 4 4 4 4
and cactus tissues. The meal represented by Cop. 94
12 4 t 4 had very likely not been eaten in the cave, but rather
87 4* 4
88
at home the day before. That represented by Cop. 74
4 4
90 4 4 had perhaps been eaten away from home, but with

91 t
maize and squash brought from home and supplemented
Tc 272, B 55 4 t t 4 4 4
by semidomesticated or wild maguey and cactus tissues.
56 t 4 4 4
57 4 t t 4 4 4 This suggests someone more than twenty-four hours

58 t 4 4 4
away from home, most probably nearer thirty-six hours.
68 4 t 4 t 4
Zone IV contained six eoprolites, five of them from
Tc 50, IV 95 t t t t 4 4
96 t t t t 4
the same location. Pochote root and cactus were co-
97 t t 4 4 dominant materials in four (Table 34). The presence of
98 t 4 t 4
99 t
small quantities of setaria in all four suggests a late
4 4
wet-season or earlv drv-season meal. The fact that two

tPresent.
^Dominant
of these eoprolites contained at least three different
or co-dominant.
"Tissue and seeds. types of cacti suggests a party living off the land. The
two other from Zone IV both suggest dry-
eoprolites
season meals. Maguey and pochote tissue were co-
dominant materials, along with meat, some tree-legume
contained no meat and was most likely a wet-season remains, and empty grass glumes. The other coprolite
meal. contained a piece of wood, some unidentified fruit tis-
Zone V contained five human eoprolites, two of sue, and meat; this was probably an early dry-season
which represented maguey meals-one maguey alone, meal.
the other maguey with fruit and meat. The former was During the Palo Blanco phase, therefore, it would
probably a dry-season meal; the fruit in the latter points appear that Coxcatlan Cave was occupied from the
to the late wet or early dry season. A third coprolite mid-wet to the mid-dry season.
consisted of cactus tissue, plus five different types of Zone VI shows that pochote root was still the dom-
275
immature stamens in Cop. 94 of Zone V represent a

delicacy still eaten in the Tehuacan Valley today. This
is the only time such material has been found in the
coprolites. Other plant materials eaten with it were
beans, pineapple, and maguey tissue. Another copro-
lite (no. 46) with this latter combination of three plants
appears in the Venta Salada phase. The presence of
pineapple (identified by the characteristic appearance
of seeds and bracts) is actually of great interest, as this
seems have been an imported fruit at that time
to —
true luxury item —
there being no evidence of its culti-
vation in the valley.
Cop. 74 of the Palo Blanco phase contained two eulti-
Fig. 180. Cucurbita mesocaip, showing the occasional gens, with maize a co-dominant material and squash
grouping of cells in rosettes. almost as common. The plants eaten along with them
were cactus tissue and maguey.
In Zone IV the use of at least three different cacti at
one meal (Cop. 96) is of interest. Although some cacti
may have been cultivated at this time, it would appear
that those eaten were probably wild, since the other
plants involved undoubtedly are. Another Zone IV
coprolite (no. 101) contained pochote and maguey
tissues plus meat. This combination probably repre-
sents a hunting diet.
Pochote root, Opuntia, and maguey tissues appear
stillto have been roasted, and even charred on oc-
casion, but not regularly so. On the other hand, there is
no evidence of boiling.
At the present time, when chopped maguey or yucca
Fig. 181. Cucurbita testa has a characteristic wavy pattern. buds are eaten, they are served with eggs, but no evi-
dence of egg was detected in Cop. 94. Actually the bud
material appeared slightly brown as if lightly roasted,
but this color could be attributable to the retention of
bile pigments by actively dividing tissue.
inant plant material in the cave or wild diet of Palo The presence in Zones IV to VI of coprolites from
Blanco times. Fragments of chili pepper occurred what are probably peasants tending fields throws some
sparsely on five occasions, Cucurbita once, and maize light on their dietary habits, since much of their food
not at all in the nineteen coprolites from the same was roasted. The squash exocarp (Cop. 74) was very
square and level of Zone VI. If the individuals who definitely roasted, as was the chili tissue; tissues of
used the cave had brought food from home, one or more maguey, Opuntia Lemaireocereus, and, of course, po-
,
of these eultigens should have shown up in quantity.* chote all show evidence of roasting.
Furthermore, if they had come from home within the Of the twenty-one coprolites in Zone VI, sixteen con-
previous twenty-four hours or so, these eultigens would tained meat. The type of meat in Cop. 36 is identified
have shown up plentifully in the coprolites also. It is by the presence of mouse vertebrae. Bone was present
clear, therefore, that the individualswho had used in two other coprolites, but could not be identified.
Coxcatlan Cave had been away from home for forty- Four of the five coprolites from Zone V contained meat,
eight hours and were living off the land, using the same the source of which was identifiable in two cases. Cop.
plants as their ancestors. 74 contained hairs of Audubon cottontail, as did Cop.
The chopped maguey or yucca buds showing the 94. In Zone IV six of the eight coprolites contained
meat. Of these, only Cop. 99 had hair and bone as well:
° It may he assumed, however, that little recognizable evi-
analysis revealed plentiful hair of peccary, a few of
dence of maize would remain if corn were eaten in the form of
tortillas, atole, or pozole. ed. Audubon cottontail, and a human hair.
276
To sum up Coxcatlan Cave during the Palo Blanco seeds, all of which would indicate a meal of the wet or
phase: of twenty-onehuman coprolites from Zone VI, early dry season.
nineteen were from the same square and level; twelve Maguey, then, was the dominant food plant of
of these had pochote root tissue as the dominant ma- which we have evidence; some form of cactus was sec-
terial. Other dominant or co-dominant plants were ond in importance. This combination is in agreement
setaria, maguey, black sapote, and cactus. The remain- with the diet from this same cave in the Ajalpan phase.
ing two coprolites contained no seeds and no meat. The exceptions to this are the wet-season meals con-
Seasonal indicators pointed to both wet- and dry-season sisting of various seeds.
meals. With regard to food preparation, the maguey tissues
In Zone V five coprolites represented late wet-season, in Cops. 6, 12, and 90 were definitely roasted and
early dry-season, and dry-season meals. Another cop- charred. There are indications that the cactus tissue
rolite contained what is believed to be a city dweller’s was 6. There was no indication
roasted, at least in Cop.
meal, including chopped lily buds and pineapple. that meat had been roasted.
Still another coprolite contained two cultigens, maize Bones the size of those of a young rabbit occurred in
and squash, among remains of wild plants and probably Cop. 12. There was no hair by which the animal eaten
represents a peasant meal eaten some thirty-six hours could be identified. A few very small pieces of egg
after leaving home. Both coprolites probably represent shell recovered from Cop. 6 were too small to be fur-
dry-season meals. ther identified.
InZone IV pochote root and cactus were the dom- Zone D contained no human coprolites, but Zone C
inant plants, but the presence of a few setaria seeds in- contained two. Lcmaireocereus was the dominant ma-
dicated late wet-season or early dry-season meals. The terial in Cop. 61, which also contained a little meat.
presence of three kinds of cacti together in two copro- One piece of pochote root and only half a glume of
lites suggests a party, perhaps hunters, living off the setaria suggest that this coprolite may represent an
land in the dry season. Two further coprolites with early dry-season meal. Cop. 64 is a meat coprolite with
maguey and pochote root as co-dominant materials disintegrating rabbit bones. The vegetal materials are
have indicator materials that suggest they represent chili fruit and seeds and part of a tree-legume seed.
early dry-season meals. These items, if eaten fresh, would fix the meal as a late
Pochote root, Opuntia, and maguey tissues were wet-season or early dry-season one.
roasted or even charred. There was no evidence of Zone B contained five human coprolites, three of
boiling. them from the same square and level. Maguey and
The meat consumed has been identified bv
J
the hairs meat were present in all five, along with seeds of two or
as peccaryand Audubon cottontail. more types. They probably represent late wet-season
Insects recovered whole or in part included sucking or early dry-season meals.
lice, fleas, ticks, filth flies, fruit flies, beetle larvae, ants, Zone A contained two human coprolites. Cop. 66 had
and pseudoscorpions. Some of the ants resemble roasted and charred Lemaireocereus tissue as the dom-
honey pot ants and may have been consumed.
San Marcos Cave

Only seven of the eighteen coprolites recovered from
Zone B of San Marcos Cave are of human origin. The
principal plant materials of these seven proved to be
maguey tissue, Opuntia tissue and seeds, perhaps tis-
sue of other cacti, mesquite seeds and leaves, and var-

ious other seeds.
In six of the coprolites maguey was dominant. Three
types of seeds in two of these (Cops. 6, 92), in addition
to cactus tissue, suggest wet-season meals, as does the
combination of maguey, cactus tissue and seeds, and
meat in Cop. 12. The others (Cops. 88, 90, 91), contain-
ing only maguey tissue and meat, are almost certainly
dry-season meals. Cop. 87 contained both Opuntia tis- Fig. 182. Capsicum (chili pepper) testa shows an equally
sue and seeds, seeds of another cactus, and mesquite distinctive pattern.
277
Fig. 184. Hair of ring-tailed cat.
inant material. Roasted and charred maguey tissues El Riego Cave — East Niche
were also present. There was no meat. Cop. 67 con-
tained maguey and Opuntia tissues as co-dominant ma- Ten coprolites, of which seven were of human origin,
terials, along with meat and what very likely is cassava. were recovered from the Palo Blanco phase of the
These coprolites probably represent dry-season meals. East Niche of El Riego Cave. The human coprolites all
The coprolites from Purron Cave show the same came from Zone D. The principal plant materials of
combination of maguey and meat supplemented by these were maguey and cactus tissues and various seeds.
other materials that were seen in the coprolites from Maguey tissue was present in five of the human cop-
other caves of this and earlier phases. The dominance rolites and was dominant in four of them (Table 10).
of Lemaireocereus in three coprolites emphasizes that Most of the other plant materials were in the form of
local wild plants were used. seeds, making it evident that all seven coprolites rep-
Charring or over-roasting of plant materials is def- resent wet-season meals.
inite in three coprolites only. Lemaireocereus tissue is These coprolites contained, in fact, a number of
charred in Cop. 54 (Zone E) and Cop. 66 (Zone A), seeds which had not been seen before. There were
maguey tissue in Cops. 66 and 67 of Zone A, and cas- composite seeds which are not those of sunflower and
sava tissue in Cop. 67. It has proved impossible to dis- solanaceous seeds that are not chili but could be Phij-
tinguish between the residual bile stain and light roast- salis. There were two different types of cactus seeds in
ing of materials from the other coprolites. Cop. 49 that are neither Opuntia, Lemaireocereus, nor
Meat was present in all but one coprolite. It has been jiotilla ( Escontria ). Cop. 49 also contained loments and
possible to identify hairs of Audubon and Mexican tissue remains of mesquite pods.
cottontail, ring-tailed cat, bobcat, peccary, and deer. Cops. 53 and 86 contained maguey tissue that was
A few human hairs were also present in Cop. 55. definitely charred, as was the cactus tissue in Cop. 50.
In summary, in Zones E and C of Purron Cave, there However, there were no signs that any of the seeds had
were two coprolites with Lemaireocereus as dominant been subjected to roasting.
material (Cops. 54 and 61) and another two with meat Meat had formed part of six of the seven meals rep-
as the dominant material. In Zone B there were five resented in Zone D. It was the co-dominant material in
coprolites with maguey, cactus, and meat as dominant Cops. .50 and 53. The meat in Cop. 50 was accompanied
or co-dominant materials, Lemaireocereus being pres- by a number of hairs and the bones and teeth of what
ent in two of them. In Zone A there was roasted Le- seem to be mouse. The hairs were mainly those of
maireocereus and roasted maguey in one coprolite, and Audubon and Mexican cottontail and bobcat. A few
roasted maguey, roasted Opuntia, and roasted cassava human hairs were also present in Cops. 50 and 51.
in another. Thus our evidence suggests that Lemaireo- Adult beetles occurred in quantity in Cop. 45 of
cereus and maguey were the principal plants used in Zone D. They were broken up sufficiently so that it is
this cave. Seeds and other indicator plants seem to certain they were eaten. In fact, they formed the co-
point to the late wet season or early dry season and the dominant material with maguey tissue. Sylvilagus
dry season as the most likely times for these meals. (cottontail) hair was also present.The shape and size
278
of the coprolite and the color of the trisodium phos- Table 35. Contents of Venta Salada Phase
phate in which it was soaked all point to a human donor Coprolites from Tc 50
for this coprolite.
Other insects that could be identified in human cop-
(Phaseolus)
rolites included a flea from Cop. 51 and the remains of cactus
Lemaireocereus
beetles from Cop. 52. no.
Pochote
Setaria
Maguey Other Beans
In the East Niche of El Riego Cave during the Palo

Zone
Cop.
Meat
Blanco phase maguey was the dominant material in the

III 26 4 t* 4
human coprolites. The presence of several types of
46 4 4 t
seeds indicates wet-season meals. Some of these seeds 84 t t
had not been found in earlier cultures. Meat sources 87 4 t
were chiefly Mexican and Audubon cottontails and 86 4 t
even bobcat. II 22 4 4 4 t
24 4* 4
Sum mart/ 80 4
According to the coprolites, meat sources in the Palo 1 77 t t 4
—
Blanco phase as, of course, part of a cave diet, not a tPresent.
city diet —were mainly small animals, the kind that tDominant or co-dominant.
‘Tissue and seeds.

could be trapped. However, deer hair has been identi-
fied, so that some hunting was still practiced. At least
two coprolites (Cops. 74 and 94) apparently repre- the sixty-six animal coprolites, one came from Tecorral
sented a city diet. Cave and the others from Coxcatlan and El Riego
At some stage of their cultural development, the caves.
Tehuacan Valley inhabitants must have learned to re-
move the developing buds from maguey plants, in Coxcatlan Cave
order to hold the accumulated food in the crown of the The principal plant materials in the human coprolites
plant and lengthen the season during which the leaves from Zones I-III of Coxcatlan Cave proved to be
could be eaten. As we have pointed out, we believe maguey tissue, setaria seeds, Opuntia tissue and seeds,
that maguey was probably a cultivated as well as a Lemaireocereus tissue and seeds, and common beans.
wild plant in the Tehuacan Valley. The flower buds The Zone III, contained
earliest of the three zones,
may have been yucca or squash, however. The pine- five coprolites of human origin. Cop. 26 represents a
apple eaten with the buds was almost certainly brought meat meal, with Lemaireocereus tissue and some cactus
from a distance and represented an even greater fruit as sub-dominant materials, plus spores and pollen
luxury. grains. A few coyote hairs were present. This very
Leaves and stems of maguey and Opuntia and root likely represents a late dry-season or early wet-season
of pochote were still eaten roasted. Squash and chili meal. Cop. 46, on the other hand, contained the inter-
had been roasted as well. Setaria also showed signs of esting combination of common beans, pineapple, and
roasting. At this period, however, the grains and glumes maguey tissue as co-dominant materials, and in addi-
were shattered. Preliminary observations indicate that stamens and pollen grains, and
tion a grass flower with
this must have been done with a rolling motion, im- meat. Study of the hairs reveals kangaroo rat as the
plying the use of grinding implements resembling the source of the meat. The flower with stamens would
modern mono and metate. This is in contrast to the indicate a wet-season meal. The bean-pineapple-ma-
preparation of setaria in El Riego times, when a stir- guey combination was also seen in the Palo Blanco
ring orpounding motion fractured the grains in a man- phase, in Cop. 94.
ner suggesting the use of mortar and pestle. Cops. 84 and 86 both came from the same square and
level in Zone III. Both are composed principally of
Venta Salada Phase setaria, with traces of a number of other materials, but
Coxcatlan Cave and the East Niche of El Riego Cave no meat. They probably represent wet-season meals,
each has three occupation zones of the Venta Salada though the grain could have been stored. Lemaireo-
phase. Only fourteen of the eighty coprolites from these cereus stem tissue was the dominant material in Cop.
sixzones were of human origin, making the sample 86, which lacked meat and is likelv to represent a dry-
inadequate for any but very general conclusions. Of season meal.
279
Zone II contained three human coprolites. Cop. 22, eating of honey comb. This coprolite consists prin-
composed of the remains of beans, maguey, and a little cipally of meat, but it also includes Lemaireocereus tis-
Opuntia tissue, quite likely represents a dry-season sue and fruits. Cop. 46 contained beetle elytra (wing
meal. A single deer hair was present. Cop. 24 consisted covers), legs, and other parts; the leg of a louse; and the
of Lemaireocereus fruits and disintegrating bone and remains of a dermestid beetle larva (parasite on animal
probably came from a late dry-season or early wet- remains). The coprolites without meat did not contain
season meal. Cop. 80 consisted of Lemaireocereus stem insect remains, with the exception of one which con-
tissue exclusively and probably represents a dry-season tained a few scraps of unidentifiable insect cuticle.
meal.
Zone I contained a single human coprolite, Cop. 77.
Maguey is the dominant material. There are traces of El Riego Cave — East Niche
setaria seed and pochote root, probably from a previous The principal plant materials in the human coprolites
meal, and no meat. This probably represents an early from Zones A-C of the East Niche of El Riego Cave
dry-season meal. proved to be maguey tissue; chili fruits and seeds;
The most notable plant in the coprolites of these cactus tissue, fruits, and seeds; common beans; and
zones is the organ cactus, Lemaireocereus, eaten in maize kernels.
three of the nine meals as stem tissue, and once (Cop. Zone C contained a single human coprolite consist-
24) as fruit tissueand seeds. Lemaireocereus occurred ing mainly of unidentified seeds and maguey tissue.
as the dominant plant in the Purron Cave coprolites of Beans and chili seeds were also present, but since both
Palo Blanco times and seemed typical of the diet of that could be stored, are no longer useful as seasonal indi-
cave. Two other such coprolites appeared in the Santa cators. On the basis of the unidentified seeds, this was
Maria level of Coxcatlan Cave. The Venta Salada Le- probably a late wet-season or early dry-season meal.
maireocereus coprolites from Coxcatlan Cave are, Zone B also contained a single coprolite (Cop. 73) in
therefore, not unusual. The almost complete absence of which maguey tissue, tissue and seeds of chili fruit, and
pochote root, which in other phases is a regular part of meat were co-dominant. Beans and cactus seeds were
the diet, is noteworthy. also present. This probably represents a wet-season
The two setaria coprolites in Zone III follow the pat- meal.
tern established thousands of years before, in late El Zone A contained three coprolites. Cop. 60 was
Riego times. However, the pattern of bean, pineapple, chiefly cactus tissue and fruit, including young seeds,
and maguey tissue (Cop. 46) has been seen in the Palo with some maguey tissue and meat. Cop. 61 consisted
Blanco phase and is a “city" meal in contrast to a “cave” mainly of maguey and meat, and Cop. 62 of maize and
meal.A meal of beans and maguey (Cop. 22), this time meat, including a small piece of cob. Cop. 60 repre-
accompanied bv Opuntia, is a combination also seen in sents a meal of the early dry or late wet season, and
the Palo Blanco phase. the others probably represent dry-season meals.
Lemaireocereus remains, which occurred in four cop- Points to be noted about the plants identified in these
rolites, displayed no evidence of having been roasted coprolites are the continued use of maguey tissue as a
or charred. If they were roasted at all, they were main constituent of the cave diet in all three zones, and
roasted only lightly. This appears to be true of the ma- the combination of beans and chili pepper, so frequent
terial in all coprolites except Cop. 77 of Zone I, which in the present-day diet. The chili pepper in Cop. 73
contained maguey tissue that was definitely charred. had been eaten whole. Cactus tissue and fruits were
In two coprolites, consisting of a combination of bean also being used. Neither the maguey nor the cactus
still
and maguey, there was no evidence that either food had tissues showed evidence of roasting.
been roasted. A number of animals served as sources of meat. The
Meat was present in only four of the nine coprolites. Cop. 73 reveal that bobcat had been eaten, plus
hairs in
In Cop. 24 (Zone II) the disintegrating bone seems to a very coyote and harvest mouse; meat was the
little
have come from an animal larger than a rabbit. In co-dominant material in this coprolite. In Cop. 60 the
Cop. 46 (Zone III) hairs were identified as those of meat, identified by hairs, consisted of quite a lot of
kangaroo rat. This coprolite, as we have noted, is the Mexican cottontail and bobcat, as well as a little kanga-
second meal containing pineapple to be found among roo rat. There were no bones. The hairs in Cop. 61
the Tehuacan coprolites. show good deal of kangaroo rat and Audubon
that a
Among the insect remains in the human coprolites, cottontail had been consumed. Cop. 62 contained bone
bee and wasp were identified in Cop. 26, suggesting the chips, and the hairs indicated that a lot of Audubon
280
cottontail had been eaten. A young feather was found soaked a minimum of 72 hours, and some colored the
in Cop. 60. solution a pale straw yellow instead of a dark brownish
or blackish shade. The pale coloration was frequently
Summary a determining factor in separating human from non-
Lemaireocereus, one of the organ cacti, formed an human feces.
important part of the diet of the Venta Salada phase, Through hairs in the coprolites was possible to
it
more so than ever before in Coxcatlan Cave. It had identify the meat in the human and the more
coprolites
appeared in the two coprolites from San Marcos Cave frequent victims of the various carnivores. There was
representing the Ajalpan phase and was used fairly little or no evidence, however, to identify the donors of
extensively in Santa Maria and even Palo Blanco times. the coprolites. Many were probably coyote feces, which
Like the pochote tree this cactus can only be described physically resemble human coprolites, and some could
as a wild plant. The next most important plant in Cox- perhaps be attributed to dogs, an animal introduced
catlan Cave was maguey, which was frequently as- into the valley in Abejas times. Others may represent
sociated with beans. In the East Niche of El Riego bobcats, skunks, raccoons, and opossums. Some speci-
Cave maguey was the principal plant food and was mens from the most recent levels turned out to date
sometimes accompanied by beans and chili peppers. from post-Conquest times, being, quite obviously,
The combination of beans and chili is a pattern which burro droppings.
has survived in the dietary habits to the present day. The only animal coprolite from the El Riego phase
Pochote root and, with one exception, setaria no came from Zone XIX of Coxcatlan Cave. It contained
longer appear to play a significant part in the Venta a stone in the center, plus what appeared to be clay,
Salada diet, as they had done from late El Riego times and around that disintegrating bone. There were meat
right through the Palo Blanco period. Squash and debris and spores of some kind, but no hairs and no
maize have appeared mainly in animal coprolites, other vegetal material.
which may indicate that these two cultivars did not The single animal coprolite from the Coxcatlan
form a part of the human cave diet, but only of the citv phase was from Zone F of San Marcos Cave. Roughly
diet. Roasting of maguey and cactus still appeared to 40 percent by volume consisted of bone, which in-
be practiced, though there is not so much evidence of cluded vertebrae, ribs, and an atlas bone. Their size
it in this phase as in previous ones. Squash tissue had suggested mouse, but the few hairs in the coprolite
been roasted as well. were not identifiable. Meat debris and cartilage were
The meat eaten was mainly from animals that could also present. Some of the cartilage looked rather like
be snared or trapped, such as mouse, cottontail, ring- insect chitin but actual insect chitin and larvae were
tailed cat,and bobcat. Small animals such as mice also present. A little plant debris was present, but was
could only be identified by hairs, since bones were not not recognizable.
usually preserved. This suggests some form of cooking Two animal coprolites represent the Abejas phase.
other than roasting which would permit bones be re-
to One was recovered from San Marcos Cave, Zone D,
moved easily after cooking. In this phase remains of and consisted mainly of meat, with some bone but no
meat were absent from the coprolites more frequently hair. The bones suggest rabbit or something a little
than in any former phase, but it should be emphasized larger. The second coprolite came from Zone B of Abe-
again that this was a cave diet, probably of peasants or jas Cave. Recovered from itwere patches of iguana
slaves. skin with scales, as well as bones and teeth of a small
Tecorral and El Riego caves contained animal cop- animal. Study of the hairs present indicated that har-
rolites almost exclusively and probably were only inter- vest mouse, Audubon cottontail, and traces of deer may
mittently used for human habitation. Coxcatlan Cave, have been eaten. Roasted plant material was present,
on the other hand, was certainly inhabited by human but could not be identified. Beetle elytra (wing covers)
beings during Venta Salada times. were recovered, plus several insect abdomens that
could not be identified.
Animal Coprolites The three animal coprolites from the Ajalpan phase
Of the coprolites from the Telmacan caves, 121 were were found in San Marcos Cave, Zone C. The plant re-
judged to be of animal origin on the basis of size, shape, mains in two were meager and unidentifiable. Cops. 1
and the color of the trisodium phos-
smell, contents, and 4 contained hairs identified as those of Mexican
phate solution which the individual coprolites were
in and Audubon cottontails, along with rabbit-like bones.
soaked. As pointed out earlier, each coprolite was What appeared to be human hair was also present in
281
Cop. 4, but is probably a whisker, which closely re- of Mexican and Audubon cottontail and ring-tailed cat.
sembles human hair. The third coprolite (no. 5) con- Thirty-seven coprolites represent the Palo Blanco
tained maguey tissue and some maize pericarp, along phase. They usually contain meat debris and hair, and
with meat debris and Audubon cottontail hairs, but no often bone and cartilage. The victims from the Palo
bones. This content could have come from a human Blanco animal coprolites, as identified by hairs, include
meal, but the pale yellow color of the trisodium phos- deer, ring-tailed cat, raccoon, cottontail, and bobcat.
phate solution indicated an animal origin. The vegetal material from these coprolites consisted
Most of the insect remains in the Ajalpan phase ani- largely of leaves, such as would have been (and still
mal coprolites were unidentifiable scraps. In Cop. 5, are) lying —
upon the valley floor mesquite, acacia,
however, were the remains of ants, probably acciden- cholulo, S ideroxi/Jon (cosahuico), grass or reed frag-
tally ingested with grass and other plant material. ments —and even maize leaves and slivers of wood.
The Santa Maria phase is represented by eleven On occasion, maguey, chili pepper, and apparently
animal coprolites, of which nine were found in Zone cassava tissues were found.
VII of Coxcatlan Cave. Cop. 12 colored the trisodium Ten of the animal coprolites came from Zone VI of
phosphate a greenish brown, a color rarely seen, and Coxcatlan Cave. The vegetal matter consisted of leaves
had a strong ammonia smell. Among the meat debris of many kinds, sometimes a little maguey or cactus
in this coprolite was a complete knuckle held together tissue, a few grass seeds, and evenpepper tissue.
chili
by tendons (?), plus some other bones containing mar- The hairs in the coprolites represented mainly deer.
row. The vegetal material consisted of leaves and some The deer hairs in Cops. 38 and 39 were singed and those
charred material. The contents of Cop. 13 were mainly in Cop. 78 were well-charred. Some of these coprolites
meat with some leaves and maguey debris. Cop. 23 also might be dog in origin.
contained mainly meat, with grass and sedge remains Zone V contained five coprolites. One consisted
among other leaves. There were chips from large bones merely of remains of meat with no bones or hairs by
and hair badly —
damaged bv chewing one of onlv two which to identify it. The others contained meat debris,
instances of such damage in the Tehuacan coprolites. leaves and twigs, scraps of possible eultigens, and hairs
The hairs were those of deer, coyote, and rock squirrel. of ring-tailed cat and Audubon and Mexican cottontail.
Animal Cops. 24-29 all contained meat as the dom- One also contained chips of bone, seemingly from a
inant or co-dominant material, along with various kinds large animal.
of fragmentary leaves, some of them grass or maize. Zone IV contained only two animal coprolites —one
The leaves were not fresh, but had already partially including jaws and teeth of a small animal, along with
decayed on the ground. In some coprolites of this group cactus tissue and seeds, and unrecognizable plant re-
were found maguey, cactus, what seems to be cassava, mains. Study of the hairs revealed that a fair amount of
—
and even bean tissues of some type materials which ring-tailed cat had been eaten, as well as the two cot-
might indicate the donors were domesticated or semi- tontail species and bobcat. An odd whisker of unknown
domesticated animals. The hairs in Cop. 25 reveal that origin was also present. One of the Mexican laborers
substantial quantities of ring-tailed cat had been con- explained that this coprolite looked like coyote feces,
sumed. Cop. 26 also contained hairs of ring-tailed cat. which it probably was. The other coprolite contained
Cop. 27 contained hairs of Audubon cottontail, as did no bones or hair, though it did contain cartilage. Plant
Cop. 28. Cop. 29 contained hairs of Audubon cotton- remains were unrecognizable.
tail, a few ring-tailed cat hairs, and one unidentified Several sucking lice (Mallophaga), a flea, and some
whisker. ticks were found in these coprolites from Coxcatlan
The other two animal coprolites, Cops. 15 and 16, Cave. Fannio larvae, fruit fly larvae, several dermestid
were obtained from Purron Cave. The most striking beetle larvae (probably of the genera Anthrenus and
plants recovered from Cop. 15 were selaginella, a fila- Trogoderma ), as well as ants and the remains of
mentous green alga, and the epidermis of a small 777- pseudoscorpions appeared in both human and animal
landsia, an epiphyte that often grows on the Zizijphus coprolites. These had obviously entered the coprolites
pedunculatus (cholulo) shrubs. This coprolite also when they were fresh.
contained disintegrating bone chips and must have Eleven animal coprolites came from Zone B of San
come from a digestive system used to handling bone, Marcos Cave. Cops. 1-5 came from the same square
rather than plant tissues. Therewere no hairs to suggest and level. Besides meat debris, bone, cartilage, and
the possible origin of the meat and bone. Cop. 16 con- hair, they contain leaves of mesquite, cholulo, and other
tained meat remains, maguey tissue, grass, and hairs plants and the pods of a tree legume. Deer hair was
present in all. Mexican cottontail and possibly raccoon among which were several groups, each recovered
hair appeared in Cops. 2 and 3, and ring-tailed cat hair from a single square and level of the zone. In one group
in Cops. 1 and 5. (Cops. 1, 2, and 3) meat and maize were the co-dom-
Cops. 7-11 from Zone B were meat coprolites. Hairs inant materials, with seeds, leaves, and other plant
in Cops. 7 and 8 were identified as those of ring-tailed tissues present in smaller amounts. Maize silk was pres-
cat. Cop. 9 contained numerous hairs of Audubon and ent in Cop. 2. The hairs in these three coprolites con-
Mexican cottontail. Cop. 10 contained unidentified sisted of a good deal of ring-tailed cat and Mexican
hair. Cop. 89 contained squash and maize remains cottontail, as well as a little raccoon in Cop. 2 and bob-
along with meat debris, bone, and hairs of ring-tailed cat in Cop. 3. Portions of adult beetles, ants, pseudo-
cat. scorpions, nematodes, and beetle larvae were also re-
Six nonhuman came from four Palo Blanco

coprolites covered.
zones of Purron Cave. One of them (Cop. 70) consisted Some coprolites (nos. 15, 17, 19-27, 29-32, 42, and 45)
only of cactus ( Escontria tissue and seeds. The rest
)
were apparently burro droppings. They are not rel-
contained meat and unidentifiable plant tissue. Mouse evant to this study, since they date from post-Conquest
bones and hair of Mexican cottontail were found in times.
Cop. 59, which also contained insect eggs, as yet un- Other animal coprolites from Zone II included three
identified. Audubon cottontail hairs appeared in Cop. probably of coyote origin, all of them with meat as the
60. Bing-tailed cat hairs were present in Cop. 63. Mexi- dominant material accompanied by cactus tissue, mes-
can cottontail and ring-tailed cat hairs were identified quite, or a leaf or two. The
hairs in Cops. 17 and 25
inCop. 69. Cop. 62 contained mouse teeth, meat which and some raccoon had been
indicate that a lot of deer
had been charred, and numerous hairs of Audubon eaten. Cop. 28 contained numerous deer and pocket
cottontail. gopher hairs and a few Mexican cottontail and ring-
Three animal coprolites came from El Biego Cave. tailed cat hairs.
Cop. 47 from Zone E contained a number of hairs Cops. 41, 43, and 44 contained mainly meat and some
identified as those of Audubon and Mexican cottontail. grass or leaves. The hairs in Cops. 43 and 44 identified
There were also remains of maize, beans, cassava, and the sources of meat as deer and Audubon cottontail.
unidentified fruit tissue, and possible maguey. The Cop. 41 were those of ring-tailed cat. Cop.
hairs in
Zone D contained two animal coprolites. Cop. 46 was 79 contained remains of Audubon cottontail and rac-
fossilized, but had apparently consisted mainly of meat. coon, together with grass flowers and Tragacanthus
Cop. 48 contained the remains of common millipede, seeds. Cop. 81, probably turkey droppings, contained
beetle, and hairs of deer and Audubon cottontail. Some maguey, maize, squash, seeds of various kinds, and a
cactus tissue, mesquite leaves, and bits of wood were little meat, which was probably from larvae.
the plant remains. Among the animal coprolites from Zone I is a group
Of the Venta Salada phase animal coprolites, seven (Cops. 4-11 and 74) from the same square and level, all
were recovered from Zone III of Coxcatlan Cave. Cop. of which turned out to be of burro origin. Cops. 12, 14,
16, perhaps dog in origin, was a meat and squash cop- 23, and 52 were also burro droppings and consequently
rolite. The meat was identified by hairs as deer and are not relevant to this study.
raccoon. Other plants included cactus, maize, and Cop. 13, probably turkey droppings, contained
beans. Cop. 27 was an almost purely meat (deer and maize, cactus, chili pepper, grass, the remains of a moss
ring-tailed cat) coprolite. Cop. 82 contained maize silk plant, and a little meat. The hairs in Cop. 76 indicated
as well as grains, a lizard bone, an unidentified mam- that a large amount of raccoon and a fair amount of
mal bone, and hairs of coyote and ring-tailed cat. Cop. ring-tailed cat had been eaten. A single coyote hair
83 consisted mainly of meat, along with cactus and probably indicates the donor.
maguey tissues. The good
hairs present revealed that a Because of the greater meat content, the insect popu-
deal of ring-tailed cat and some Mexican cottontail had lation in the animal coprolites from the Venta Salada
been eaten. Cop. 85 contained maize, Opuntia tissue, zones was high. Filth flies (
Fannia ), beetles (including
Lemaireocereus fruits, and no meat. Cop. 89 was a se- dermestid larvae), ants, lice, mites, and spiders were
taria and maize coprolite, with no meat. Cop. 88 con- frequently found. Pseudoscorpion and other arthropod
tained millipede and adult beetle remains, a few Audu- remains also were present.
bon cottontail hairs, a rib (snake?), maguey tissue, and The majority of the coprolites from the Venta Salada
some seeds. zones of the East Niche of El Biego Cave were of
In Zone II there were twenty-one animal coprolites, animal origin. In Zone C there were nine, of which two,
283
bean squash maize amaranth pochote maguey setaria LEMAIREOCEREUS
AD
1500
VENTA SALADA
700
0 PALO BLANCO
200
SANTA MARIA
900
AJALPAN
1500
PURRON
2300
ABEJAS
3500
COXCATLAN
5000 -
BC
Fig. 185. Polygons showing the importance of various foods in the prehistoric diet. The number of coprolites in which a
plant occurs is expressed as a percentage of the total human coprolites for each culture.
Cops. 67 and 68, were probably rabbit droppings. too close to the cultivated fields and town or village
These contained cactus tissue and seeds, pollen grains, homes to be much used by hunting parties or agricul-
and unidentified tissue; there was no meat debris. Cop. tural laborers.
68 contained a few Audubon cottontail hairs probably A single animal coprolite was recovered from the
indicating a member of this species as donor. Cops. Venta Salada phase of Tecorral Cave. Cop. 75 was a
70-72 and 91-93 all contained meat and fair amounts meat coprolite containing skull and jaw remains of a
of vegetal material including maize, chili pepper, small animal and hairs identified as those of kangaroo
maguey, setaria, and various seeds. These items, re- rat and bobcat. There were also a few fragments of
sembling a human diet, may indicate that the copro- insect cuticle and a trace of grass.
were dog in origin. Rabbit and mouse skull bones
lites
were present, and the hairs represent a variety of ani- Conclusions

mals: Mexican and Audubon cottontail, kangaroo rat, In the polygons (Fig. 185), the number of human
bobcat, deer, and raccoon. coprolites in which a particular plant occurred is ex-
Zone B contained nine animal coprolites. Cops. 56- pressed as a percentage of the total of human copro-
59 consist mainly of beetle and millipede remains sug- lites for each culture. Since meat debris occurred in
gesting a raccoon origin. Cop. 63, with the remains of every human coprolite of the Coxcatlan phase, the
adult and larval stages of beetles, may also be raccoon figure for meat debris is 100 percent. If meat had oc-
in origin. Cop. 64 is very similar. curred in only four of the twelve coprolites, the figure
Cop. 65 consists of cactus fruit and meat; the meat would have been 33 percent.
can be identified by the presence of rodent teeth and It must be borne in mind that the data assembled
the hairs of bobcat, Mexican and Audubon cottontails, here come from several caves, each of which is rep-
and kangaroo rat. Cop. 66 appears to have been two resented by a slightly different dietary pattern. Most
coprolites squeezed together; one of these seems to be of the data for the Abejas and Ajalpan phases, for in-
burro in origin and the other contains adult and larval stance, comes from San Marcos Cave (Tc 254), while
beetles. Cop. 69 had maguey and meat as co-dominant much of the rest of the data comes from Coxcatlan
materials. The hairs present were identified as Mexican Cave (Tc 50). Thus, maguey tissue and meat emerge
and Audubon cottontail. Cop. 90 represented a meal of as the dominant materials in the two phases just men-
meat, cactus tissue, and numerous seeds of different tioned, and setaria and pochote root, commonly the
kinds. Among the hairs identified were a good many of dominant plants in the diet of Coxcatlan Cave, are
Audubon cottontail and a fair number of ring-tailed cat. practically absent. No coprolites were recovered from
The fact that most of the coprolites from the East the Purron phase.
Niche of El Riego Cave during the Venta Salada phase On looking at the eighteen columns of Fig. 185 one
were of non-human origin suggests that the cave was first notes three main shapes: (1) the broad-headed
284
black
OPUNT1A CACTUS MESQUITE SAPOTE GRASS CHILI CASSAVA BONE MEAT HAIR
figures with pinched necks, (2) the fat ones, and (3) the absence of maize, beans, and squash leads to the sup-
thin ones. The broad-headed figures are those of po- position that travelers possibly,and hunters and work-
chote, setaria, and mesquite, which were absent from ers almost certainly, did not carry with them much in
the two human coprolites representing the Ajalpan the way of food when they left the towns and appar-
phase, but were back again as important food plants in entlv depended on wild plants growing near the caves,
the much larger sample from Santa Maria times. It was and on the trapping of small mammals, to supply their
very noticeable that the dietary habits represented in immediate needs. This is my own interpretation, how-
late El Riego and early Coxcatlan coprolites and con- ever, and differs from that set forth in the concluding
tinuing into the Abejas phase disappeared from the chapter by MacNeish.
two Ajalpan coprolites (which did not come from Cox- The thick figures in Fig. 185 represent meat, grass,
catlan Cave), only to reappear in the Santa Maria phase and cactus, three categories that do not represent single
and continue into the Palo Blanco cave diet. Since the species. Even so, as described earlier in the explana-
El Riego, Coxcatlan, and Abejas cultures are classed as tion of how data was assembled, Opuntia and Le-
the incipient agriculture phases of the Tehuacan se- maireocereus have been separated from the other cacti;
quence, this would mean that there was probably a and Z ea and Setaria from the other grasses.
shift in diet with the development of agriculture and The really thin figures are those of Cucurbita, Ama-
urban communities. By the late Formative period rep- ranthus, and hair. Cucurbita, the squashes, have been
resented bv the Santa Maria phase, there must have mentioned above as apparently being a constituent of
been a very definite city diet based on maize, bean, and the city diet but not usually of the cave diet, even in
squash agriculture. However, the cave diet of former early times. Amaranth is another surprising absentee,
periods was adhered to by those occupying the caves, or near absentee. Although practically absent from the
who utilized the wild plants growing nearby, just as coprolites,it was found on various floors of Coxcatlan
their ancestors had done. This cave diet was then car- Cave (see Table 26 in Chapter 12).
ried on into the Palo Blanco phase, and to a small ex- The presence of hair in human coprolites has been
tent into Venta Salada times. taken as an indication of the actual meat that was
The reference above to the main plants of an agri- eaten. The vast majority of the hairs identified turned
cultural economy — maize, beans, and squash — draws out to be from small mammals such as mice, cottontail
attention to the fact that these are represented by inter- rabbits, ring-tailed cats, and so forth, and occasionally
mediate which should have
to thin figures of Fig. 185, from deer and peccary. Therefore, the hair graph fig-
blossomed forth into broad heads, but have not done so. ure, compared to the meat one, is indicative of skinning
There are traces of these plants in a few of the copro- before eating, or perhaps before cooking, even with
lites of the Palo Blanco and Venta Salada phases, which such small mammals as mice. The hairs that are found
has led to the suggestion that they were deposited by are probably owing to careless skinning, or may be
travelers or hunters who had been eating a city diet not “ambient hair”in general circulation around the area
more than twenty-four hours earlier. The comparative where the meat was prepared. Evidence of careless
285
skinning first appeared in the Santa Maria phase and among the small seeds and vegetal remains on the floors
increased through Palo and Venta Salada
Blanco of Coxcatlan Cave, chili fruit wall with epidermis was
times. At the same time the amount of meat consumed found from even Chapter 12). The
earlier layers (see
in the cave diet decreased. These probably are copro- other figure, that of cassava (Manihot) first appeared
,
lites from peasants or slaves. in Santa Maria times (about 900-200 b.c.). The material
Several other polygons do not immediately attract matched stem tissue of Manihot esculenta Crantz sup-
attention by their shape, and one of these is for bone. plied by the Montreal Botanical Garden. According to
Roughly it increases and decreases as the meat profile David Rogers (personal communication), formerly of the
—
does in fact, it is a thinner image of the meat figure. New York Botanical Garden, this is the oldest record for
This means, therefore, that the amount of bone swal- the introduction of cassava to Mesoamerica.
lowed, either accidentally or on purpose, has remained The remaining figure is the polygon representing
constant for some 7,000 years. Some of the bones were Diospi/ros cligyna, the black sapote. This tree was ap-
mouse bones, but always just an odd one or two, and parently not native to the valley and was introduced
therefore probably accidentally swallowed. Others are during the Coxcatlan phase (see Chapter 12). It re-
chips and slivers which by their size suggest animals quires irrigation in the Tehuacan climate, and as can
somewhat larger than rabbit — in fact, deer, according be seen, its use increased steadily into Santa Maria
to the hairs. Bone marrow was discovered in one or two times and then declined in Palo Blanco and Venta
coprolites, but the fragments are not susceptible to Salada times. This decline may have been due to a
identification. dietary change —
perhaps the fruit became part of the
Another figure, with broad shoulders and a narrow city diet. The black sapote, however, was almost cer-
neck, represents maguey or Agave. This plant was a tainly cultivated near Purron Cave in Santa Maria and
staple which increased in use during the Ajalpan phase early Palo Blanco times, when a large dam supplied
and continued to be used in fair quantity into Post- water for irrigation in the area.
classic times. The plant had many uses, not just for The frequency polygons show clearly which plants
the food value of its leaves, but also for the fiber derived formed the staples of the diet. In El Riego times, they
from the leaves, and for an alcoholic drink derived were undoubtedly setaria and pochote root, with cacti,
from the sap. The leaves are in the best condition for grasses, and some maguey. The chief difference in Cox-
eating late in the dry season, just before the plant catlan times was an increased use of meat. In Abejas
flowers, when a great deal of food material has been times beans and Lemaireocereus were absent from
stored in the plant in preparation for flowering. By the diet, while several other plants showed definite
cutting out the developing flower stalk, the food can increases: Opuntia squash, maize, and mesquite. Com-
,
be held in the leaves for a period of several weeks or ing to Ajalpan times maguey shows maximum use, also
even months. Probably the inhabitants of the Tehua- beans, Lemaireocereus, and black sapote. There is a
can caves discovered this at some stage in their agri- definite fading away of most plants in the cave diet of
cultural development. Palo Blanco and Venta Salada times. This fading away
The cactus genera Lemaireocereus and Opuntia applies even to maize and squash, though not to beans.
show different periods of use. The former, one of the As mentioned earlier, in these Classic and Postclassic
organ cacti, at the present time occurs frequently times there was a peasant and probably a slave class
around Coxcatlan Cave and less so around San Marcos that worked and harvested the fields, but apparently
Cave. Opuntia, the prickly pear cactus, also is more did not themselves eat the crops.
common around Coxcatlan Cave than around San The absence most plants from the animal copro-
of
Marcos Cave. Logically, therefore, Lemaireocereus lites Tehua-
of the incipient agriculture period of the
and Opuntia should be eaten more frequently in Cox- can sequence shows that the wild animals that used
catlan Cave. The polygons show that Opuntia was the caves were meat eaters. It is almost into Santa
eaten to a greater extent in the earlier phases, but that Maria times before plant material other than leaves
Lemaireocereus was more popular from Ajalpan times could be found. The exception was cactus tissue. The
on. leaves found in the meat coprolites were those that nor-
Two further figures should be mentioned, that of mally fall to the ground beneath xerophytic thorn
chili and that of cassava (
Manihot ). As can be seen, scrub. Leaflets of mesquite (Prosof)is sp.), or sometimes
chili was used in El Riego times, the oldest record from Acacia or Mimosa, along with leaves of cholulo
the coprolites being seeds from Zone XIV. However, (Z izyphus pedunculatus) and possibly grass were al-
286
most certain indication of an animal coprolite. evidence, while the old diet of pochote root and setaria
Mesquite pods were, of course, eaten when they fell continued in favor.
off the trees. These two diets probably represent mainly wet-sea-
In the Santa Maria and more recent phases, the son or early dry-season meals. A dry-season meal con-
presence of bean, squash, maize, and other plants in sisting of maguey and meat along with cactus or
tissue
the animal coprolites suggests more extensive cultiva- perhaps pochote root appeared in the
tissue,
first
tion of these crops, with consequent raids by such Coxcatlan phase coprolites and is most frequent in the
animals as raccoons, coyotes, and dogs. The bone, meat, Ajalpan phase. It seems not to have been used much
and hair figures need no explanation, but it is notice- beyond that period, possibly because it was replaced
able that they narrow in the Palo Blanco and Venta by other wild plant materials.
Salada cultures, probably as a result of the increased Perhaps one further diet should be mentioned here
amounts of vegetal material available. the diet which I suspect is that of city dwellers. Two
coprolites,one from the Palo Blanco phase (Cop. 94),
and the other from the Venta Salada phase (Cop. 46),
Changes and Continuity in Diet were almost identical, though only a few years or per-
It should be clearly understood at the outset that in haps as many as a thousand may have separated them
the Tehuacan sequence the earlier phases represent in time. They represent a diet of beans, maguey tissue,
cave and rock-shelter dwellers, but later, when towns and pineapple fruit, supplemented with meat. Chopped
and cities had come into existence, the caves were flower buds and fair amounts of “beer” were additions
probablv inhabited intermittently, perhaps by peas- to the diet represented by the Palo Blanco coprolite.
ants or slaves who planted and harvested the fields, or The oldest coprolites recovered from the Tehuacan
by hunters and trappers. sequence demonstrate that pochote root and setaria
In the El Riego period, as far back as coprolites were seed were dominant plant materials in the diet of the
recovered, a diet based on pochote root and setaria was people of El Riego times. Ceiba parvifolia is a tree that
in use and had probably been in use for some consider- bears pods containing edible seeds among a silky cot-
able time in Coxcatlan Cave. This combination of ton, known locally as “pochote.” This tree is native to
starchy root and grass seed, including wild maize, was the thorn-scrub-cactus vegetation of the east side of the
accompanied by other wild plants such as maguey and valley. Tissues from the starchy root of this tree were
cactus, and some meat. This dietary pattern continued common only in coprolites from Coxcatlan and Purron
through the Coxcatlan phase but disappeared in the caves. No evidence from the coprolites shows ingestion
Abejas phase, when pochote root was practically absent of the black-coated seeds, which are the commonly
from the diet as reflected in the coprolites, perhaps eaten products of the tree today. However, on occa-
because of the smallness of the sample from Coxcatlan sions the roots are eaten by hunting parties of the
Cave. No coprolites assignable to the Purron or Ajal- present day (C. Earle Smith, Jr., personal communica-
pan phases were recovered from that cave, and the tion).
two coprolites from another cave do not show this diet Setaria, however, may have been one of the early
combination either. In coprolites of the Santa Maria cultivars, since it may have been difficult to obtain wild
phase, obtained from Coxcatlan Cave again, the in the quantities eaten in the El Riego phase. True,
pochote root and setaria diet is represented as strongly seeds of other grasses were collected at the same time
as ever. It continued into Palo Blanco times, but prac- —judging by fragments, these included wild maize
leaf
tically disappeared in the Venta Salada phase. —but, nevertheless, the quantities of eaten setaria offer
The pochote root and setaria dietaccompanied in
is mute evidence that it may have been one of the earliest
the Santa Maria level of Coxcatlan Cave by another cultivated plants in the Tehuacan Valley. Setaria was
dietary pattern of almost equal importance: Lemaireo- also eaten in northern Mexico, as the coprolites of the
cereus and black sapote. These materials had been Tamaulipas sequence demonstrate (Callen 1963).
present in the Coxcatlan phase of Coxcatlan Cave, but There, the oldest material containing setaria came
as relatively insignificant items in the diet. However, from the lowest level of the Ocampo phase of Tine 248,
Lemaireocereus was one of the principal plants in the placing it at about 4000 b.c. Setaria was not obtained
coprolites from San Marcos Cave during the Ajalpan from the older Infiernillo phase of the Tamaulipas se-
phase. When Coxcatlan Cave was again occupied in quence. In the Tehuacan Valley, setaria was recovered
the Santa Maria phase, Lemaireocereus was much in from coprolites of Zone XVI (El Riego phase) of Cox-
287
catlan Cave, dating from before 5000 b.c. Among the tlan Cave. Its first appearance in the coprolites, how-
small seeds from the floors of Coxcatlan Cave, setaria ever, is from the Ajalpan phase.
was obtained from every zone back to XXIII (before some members of which are known today
Physalis,
6500 b.c.), showing that the seeds were consistently as ground cherry or cape gooseberry, is known in the
collected and brought to the cave. Tehuacan Valley as “tomate.” The epidermis of some
Perhaps during the Coxcatlan phase, and probably variety of these fruits appeared in the Santa Maria
not later than halfway through the incipient agricul- phase, and would on occasion form the sub-dominant
ture stage of the —
Tehuacan sequence defined by Mac- material of a coprolite.
Neish (1964a) as covering the El Riego, Coxcatlan, and Another addition to the plant list in the Tehuacan
—
Abejas phases maguey and Opuntia may have be- Valley was Manihot or cassava. As mentioned earlier,
come semidomesticates. When the maguey plant has the identification of this plant in a coprolite of the
flowered (which in some species climaxes about ten Santa Maria phase is the oldest record for this starchy
years of growth), the plant dies, but new shoots grow plant in Mesoamerica.
out from the base of the stem, so that the plant
propagates itself vegetatively. Since they were using Changes in Food Preparation
the plant for food, the inhabitants of the valley must Examination of the setaria material from El Riego
have observed and may have started to stick these
this and then from coprolites of Santa Maria and
coprolites
shoots into the ground at places convenient to them- subsequent levels revealed that the grains and their sur-
selves. This might account for the increasing use made rounding glumes, the hallmark of the tribe Paniceae of
of the plant for food, not to mention its other uses. the grass family, had been broken or shattered differ-
Sections of the stem of Opuntia will readily root ently in those two phases. As an experiment, setaria
when broken off, and this too must have been observed grains were broken down with a stone pestle in a stone
by some of the valley’s inhabitants. As a result, pieces mortar found in an El Riego level of Coxcatlan Cave.
of Opuntia stem were probably stuck into the ground Antoinette Nelken of the Tehuacan Project carried out
at convenient spots. The increased occurrence of the experiment by using both a pounding and a stirring
maguey and Opuntia in coprolites of the Abejas phase motion, each of which gave almost identical results.
suggests this semidomestication. A comparison of these results with setaria from the
The first remains of a maize kernel appeared in a El Riego coprolites suggests that either or both mo-
Coxcatlan phase coprolite, and a whole squash seed tions had been applied. Neither type of breaking
appeared for the first time in an Abejas phase copro- agreed with setaria from Santa Maria and Palo Blanco
lite. Chewed seeds of squash were also recovered from coprolites. Miss Nelken then used a grinding stone or
one coprolite of the Abejas phase. This does not, of metate, such as can still be bought in the markets of
course, exclude the possibility that normally only the the Tehuacan Valley, rolling a stone mano or “rolling
cotyledons were eaten and the seed coat discarded, pin” in the approved fashion. The resultant material
though no such material could be identified. Chili, as matched the setaria recovered from the Santa Maria
recorded earlier in this summary, was known in El and Palo Blanco coprolites. Exactly when the change
Riego times, as shown by seeds from the refuse of Cox- from one technique and set of tools to the other took
catlanCave and also from the coprolites. Beans of the place is not known, but on the basis of artifacts found,
genus Phaseolus, which are known from the dried plant it could have been during the Abejas and Ajalpan
remains in the caves as early as the Coxcatlan phase, periods. Since the setaria and other grass grains were
could be identified as the genus Phaseolus only in pounded, they may have been used to make some kind
coprolites of the Palo Blanco and Venta Salada phases. of coarse dough, as there is evidence that the setaria
One plant which was brought into the valley appar- was roasted. It seems unlikely that the grains were
ently under cultivation is Diospyros digijna, the black first pounded before being roasted dry. The seeds
sapote, which first appeared in coprolites of the Cox- were definitely not “popped,” like popcorn. Judging
catlan phase and remained popular into the Palo Blanco by the ground-stone implements recovered, by the time
phase. maize was domesticated the manner of grinding the
New plants that appear in coprolites after the in- grains and preparing the dough had been perfected
cipient agriculture stage was concluded, include by using setaria grains (MacNeish 1964b).
PJujsalis, Manihot, and Anmranthus. Amaranth does The pochote root appears generally to have been
not really belong here, as there are records of its oc- roasted, although the center of the root must have re-
currence among the early plant remains from Coxca- mained uncooked. The starch grains had not swollen
288
and were still contained within the plant cells. It would Although bones were frequently absent from copro-
appear that when the roots were dug up they were lites containing other indications that meat had been
thrown on the fire in order to burn off the soil and corky eaten, their absence cannot be taken as evidence that
bark. This outer layer became charred, and some of it some form of cooking other than roasting had been
was swallowed along with the raw center. This practice used. Tbe animal might have been roasted in its skin,
appears to have continued from El Riego times up to and when finally cooked, the skin was broken open so
the Spanish conquest. that the meat and bones could be separated with rela-
Maguey tissue was frequently eaten raw, or at least tive ease. In animal coprolites bone, hair, and meat all
there was little evidence of roasting or cooking. How- occur together, and quite often pieces of skull or jaw
ever, in about one quarter of the cases, there is evidence or other bones were recovered. However, it is the hairs
of roasting. Opuntia also was evidently eaten raw quite which enable the meat to be identified.
frequently, but occasionally there is definite evidence The animal hairs obtained from human coprolites
of roasting,and in fact, on at least two occasions, the are summarized in the following list, arranged accord-
stem piece had been roasted and then the contents ing to the chronological order of Tehuacan sequence:
scraped out and eaten, the charred epidermis evidently El Riego phase, none; Coxcatlan phase, Audubon cot-
being discarded. This type of preparation was also used tontail; Abejas, Ajalpan, and Santa Maria phases, none;
for other cacti, including Lemaireocereus, especially in Palo Blanco phase, Audubon cottontail, Mexican cot-
the more recent phases of the sequence. In the later tontail, ring-tailed cat, bobcat, collared peccary, and
cultures roasting appears to have been more frequent white-tailed deer; Venta Salada phase, Audubon and
than in the earlier incipient-agriculture period. Mexican cottontails, kangaroo rat, bobcat, and white-
Squash was roasted, as pieces of epidermis and of tailed deer.
mesocarp were found which were unmistakably Human hair was also present in the human coprolites
browned by roasting. There is also a hint that Plu/saJis and seems to suggest the eating of scalp parasites.
fruits may have been subjected to some form of heat There is a remarkable absence of animal hairs of any
treatment, perhaps to burn off the outer husk that en- kind in the human coprolites of the middle phases of
closes the berry. There was no evidence that the cas- the Tehuacan sequence. It is only in the two most re-
sava tissue had been roasted, in fact the starch grains cent, Palo Blanco and Venta Salada, that they occur
appeared completely normal, just as with the pochote in quantity. This would suggest care in skinning ani-
root. mals in the earlier periods, and a lack of it in later ones.
Meat appears to have been roasted only occasionally, Perhaps the peasants or slaves who were responsible
but evidence for this statement is scanty. Sometimes it for the coprolites in the caves did not prepare their
merely on the presence of hairs which have be-
rests meat as carefully as their ancestors had done.
come very dark brown.
REFERENCES
Bird, 1943. MacNeish, 1958, 1962, 1964a, 1964h.

Callen, 1963, 1965, 1967. Marsh, 1965.
Callen and Cameron, 1955, 1960. Marsh and Callen, 1964.
Fonner, R. L., 1957. Snead, 1962.
Hall and Kelson, 1959. Van Cleave and Ross, 1947.
Harshberger, 1896. Wakefield and Dellinger, 1936.
Heizer, 1960. Webb and Babv, 1957.
Jennings, 1953, 1957. Yarnell, 1965.
Jones, V. H., 1936. Young, 1910.
Loud and Harrington, 1929.
289
CHAPTER 15
A Summary of the Subsistence
Richard S. MacNeish
t is my fortune to draw what conclusions are possible We found wild com (Z ea mays), as well as twelve
from the excavated remains relating to the subsis- races of domesticated corn. Five species of beans were
I tence of the prehistoric inhabitants of the Tehuacan uncovered, including three varieties of common beans
Valley. This chapter is in part a resume of the data ( Phaseolus vulgaris), two varieties of runner bean (P.
previously described by the botanists and zoologist and coccineus), and a single variety each of teparv (P. acuti-
in part an anthropologist’s interpretation of these
is folitis), lima (P. lunatus), and jack beans Canavalia
(
sp.).
data. As the reader will notice, some of my conclusions We found fragments of several kinds of cultivated
and arrangements of the basic data — and Callen’s ma- squashes or pumpkins and of gourds Cucurbita mix-
terials are justifiably called basic — differ from those of ta, C. moschata, C. pepo, Lagenaria siceraria, Cres-
my colleagues. These variations arise in no small part centia cujete, and Apodanthera sp. Numerous remains
from the fact that my colleagues have studied the of fruits include remains of wild and domesticated avo-
plants and animals in terms of botanical and zoological cado ( Persea americana ), black sapote ( Diospyros dig-
relationships and behavior, while I view these data yna), white sapote ( Casimiroa edulis ), guava (Psidium
from the standpoint of man’s adjustment to and ex- guajava), eiruela (Spondias mombin), and in the feces a
ploitation of these resources and his eventual manipula- few tiny remains of what may possibly be pineapple.
tion of them for his own greatest benefit. Somewhat Fragments of both ground cherry or “tomate” (Physalis
consciously, we have not attempted to settle the dif- sp.) and peanuts Arachis hypogaea) were dug up. Cot-
(
ferences between our biological and anthropological ton Gossypium hirsutum) boll fragments and cotton
(i
interpretations because we feel that there is merit in string or yarn were fairly frequent in our excavations.
both views and that their presentation gives this vol- Bones of dogs ( Canis familiaris) and turkeys ( Meleagris
ume a broader outlook. gallopavo) were evidence of other kinds of domesti-
The Tehuacan Project was in large part concerned cates. In the feces, fragments of cassava Manihot escu-
(
with reconstructing the ancient subsistence and tracing lenta) may be evidence of another cultigen, but definite
the development of agriculture in the Tehuacan Valley. proof that they are such is lacking.
We uncovered vast amounts of material that had con- Throughout almost the entire sequence, remains of
siderable bearing upon these problems —
almost 11,000 the chili pepper Capsicum annuum) and amaranth ap-
(
zoological specimens, about 100,000 plant remains, and peared. Certainly some of these specimens, particularly
over 100 human Although most of these remains
feces. in later levels, are from domesticated plants, but the
came from only they are of fundamental im-
five caves, earlier remains are more fragmentary and one cannot
portance in reaching a solution of the problem of the assuredly say they were planted. Their almost continu-
development of agriculture in the Tehuacan Valley. ous occurrence from Zones XIX and XVIII through the
They also relate directly to the larger problem of the latest levels of Coxcatlan Cave —
plus the knowledge
origin and spread of plant cultivation and domestica- that they were domesticated in later time raises the —
tion in all of the New World. possibility that they may have been domesticated from
290
A SUMMARY OF THE SUBSISTENCE
El Riego time on. Admittedly, the evidence for culti- in quantity and are easily propagated, but we do not
vation at such an early period from overwhelming.
is far have firm evidence that they were.
However, the finding of a complete chili pod of def- The Bumelia specimens occur spasmodically in late
initely domesticated size in Zone XI (Coxcatlan phase), levels and are little different from modem specimens.
tends to support our belief that chili was domesticated This plant grows better under well-watered conditions,
in the previous phase. yet at Purron Cave, it does not appear until Zone A, well
C. Earle Smith in Chapter 12 has suggested the stimu- after the water held by the dam was Thus I see
gone.
lating hypothesis that a number of other plants were little evidence for its cultivation and some evidence
cultivars, such as mesquite ( Prosopis juli fiord), maguey against it. Leucaena esculenta, although cultivated on
(Agave spp.), prickly pear (
Opuntia spp.), Leucaena a small scale today, is still collected from the wild. The
esculenta, Bumelia latevirens, the grass Set aria, and the archaeological specimens are identical to modem wild
fruits, ciruela
(
Spondias mombin), cosahuico ( Sider - examples. Again, the archaeological specimens are not
oxi/lon sp.), coyol ( Acrocomia mexicana ), and chupan- popular until late Formative times, and even then this
dilla ( Cyrtocarpa procera ). I am least convinced by the tree,which grows best under well-watered conditions,
evidence for the grass setaria, as there seems to be con- is absent from most of the Purron levels when the dam
siderable indication that it was a wild collected plant. was in use. Of the plants mentioned so far, this tree
First of all, there is no proof of its being cultivated now seems most likely to have been cultivated, but the
or in the recent past. In the second place, its early in- evidence in hand does not prove that such did happen.
crease among the remains before the advent of com The evidence is slightly better for the palm fruit
agriculture, followed by a rapid diminution in popu- coyol and three other popular fruits of the valley, cirue-
almost duplicates the trend of other seeds (grass,
larity, la orhog plum, cosahuico, and chupandilla. All four
Opuntia, Lemaireocereus, Acacia) that were certainly of these plants flourish under well-watered conditions,
collected from wild plants. Third, the seeds from later and all show noticeable increases in popularity in Zones
levels show no increase in size as do seeds of many I-D of Purron Cave, when the dam was functioning. It
cutivated plants. In addition, setaria grows most pro- must be added that there is not a one-to-one correla-
fusely near present water sources, such as the Rio tion, for specimens of coyol and ciruela do appear in
Salado less than three miles from either Purron or Cox- Zones A-C, after the dam ceased functioning. These
catlan caves. However, during the time a large dam must have been gathered from some other watered
provided water near Purron Cave (Zones I-D), as well area. It seems reasonable to assume that originally
as during the period represented by Zones C 1 and R of ciruela was either collected in the barranca forests or
San Marcos Cave, when that area was irrigated, setaria grown in the valley in watered areas. I suspect it was
seed is almost totally absent from the deposits. the latter from Formative times on, as pits of this plant
Also unconvincing to me is the evidence for the culti- occur in rather large amounts in most components.
vation of mesquite. This species is, however, presently However, before this time it occurs so spasmodically
grown under cultivation in small amounts, although that it was probably the former and cannot be con-
most of the mesquite eaten locally today comes from sidered a cultivated plant. Coyol seeds are present in
wild plants. The evidence against its being a cultivar is some quantity in levels contemporaneous with the
There is no increase in seed size in our
fairly strong. Purron dam but also appeared in Zones J and K before
excavated specimens either within the sequence or the dam and after it in Zone C and A again. They are
when the archaeological seeds are compared with so rare before the time of the dam that I assume they
modern wild specimens. Mesquite grows slightly better were brought down from the mountains or were har-
near water, and yet in the Purron Cave deposits con- vested wild in the wetter parts of the valley.
temporaneous with the dam it is noticeably absent. In Chupandilla, solely on the basis of large amounts of
fact, the only occurrence of mesquite in Purron Cave is fruit dam functioned, seems
used during the period the
in Zone B, after the dam had been destroyed. The case have been cultivated by Formative times.
definitely to
for the domestication of maguey, Opuntia, Lcucacna, However, an observable increase in seed size among
and Bumelia is much Maguey and the prickly
the same. the unusually large sample in Zone XI of Tc 50 and the
pears grow wild nearly everywhere in the valley at small sample from Zone F of Tc 254 suggests that culti-
present and are easily collected. The archaeological vation may have begun as early as the Coxcatlan
specimens are no different from present-day wild ones. period.
They might have been cultivated, since they were used Cosahuico also shows a slightly increased seed size
291
from the Formative period on and increases in fre- tinuous appearance of the larger pits, I believe is evi-
quency in the Formative and Classic levels of Purron dence of early selection and planting. Therefore, avo-
Cave when the dam was in use. Probably it was under cados may have been planted as early as the El Riego
cultivation at that time. The rather large number of phase, about 8000 years ago, and they were almost cer-
seeds in Zone XI of Coxcatlan Cave hint that it may tainly under cultivation by 4000 b.c. They have con-
have been first domesticated in the Coxcatlan phase; tinued being planted and improved until the present
however, its relative infrequency in zones of the subse- time in Tehuacan. The plant, moreover, is not now
quent phase in Coxcatlan Cave, as well as its complete native to the Tehuacan Valley, nor can it grow under
absence from Abejas or Coxcatlan zones of Purron normal conditions in this arid region, nor is it likely to
and San Marcos caves argues against such an earlv have grown in the valley even during slightly wetter
beginning. periods. Therefore, it seems probable that our first
In summary,I see little or no evidence of the culiva- domesticate was an import from wetter regions to the
tion of maguey, Opuntia, setaria, or mesquite; nor do east or south. The only other archaeological specimens
I see any evidence that these were among the first of avocado occur in Peru during Formative times from
domesticates which led to the domestication of such 750 to 100 b.c. Thus the avocado may have slowly
plants as com, beans, and squash. Also, I believe the spread from N'lexico southward, but there also remains
evidence indicates that coyol, cosahuieo. and ciruela the possibility that it was redomesticated in the low-
were possibly cultivated in Formative times, whereas lands of South America near Peru. Only further archae-
chupandilla may have come under cultivation in late ological and botanical studies can give us a definite
Coxcatlan times. The domestication of these plants answer.
seems to me to be the result of the earlier planting of Perhaps even earlier than avocado, and equally
such trees as avocado and the black and white sapotes difficult to determine when it was first domesticated,
rather than the converse. As for Lcucaena and Bumelia, is the chili pepper (
Capsicut7i annuum). The complete
I see no real evidence of their cultivation at all. How- specimen from Zone XI of Coxcatlan Cave is almost
ever, neither the evidence for the verv early domestica- identical tomodern domesticated specimens and differs
tion of many of the plants suggested bv Smith nor mv from wild ones, so that the plant was certainly domesti-
opinion to the contrary is conclusive, and the case re- cated bv the Coxcatlan phase. This fact, together with
mains very much open. the relatively consistent (though not numerous) occur-
Besides these ten disputed plants, however, we did rence of chili remains back to Zone XIX and the fact
find thousands of specimens of other domesticates that that it is now an intruder into the valley, leads me to
few can or will dispute. Further, these species begin to suspect that it was planted from early El Riego times.
be cultivated early in the sequence and extend over a Of course, it is possible that during earlier wetter
long time range. This material not only reveals the de- periods it may have grown Tamaulipas
in the valley. In
velopment of agriculture in the Tehuacan Valley, but chili remains occur almost phase
as early, in Infiernillo
when considered in conjunction with other archaeologi- feces, while chili occurs in Huaca Prieta, Peru, from
cal plant sequences and botanical data, it throws much 3000 to 2000 b.c. Wild chili species are widespread in
light upon the origin and dispersal of many New World the New World from Mexico to at least as far south as
domesticates. Let us consider, in the order in which Bolivia. For these reasons I suspect the plant was do-
they first appear among the plant remains, each of the mesticated independently in a number of different
domesticated plants in terms of these wider implica- areas.
tions. The have been domesticated
third plant that might
One of the earliest domesticated plants in the Te- very early is grows wild in the Te-
amaranth, but it
huacan Valley was apparently the avocado ( Persea huacan Valley and we were unable to distinguish wild
americana ). This fruit was present in Zone XXIV of examples from domesticated ones among our speci-
Coxcatlan Cave, one of the earliest components with mens. Thus we have no clear evidence when this plant
preservation. from this level, as well as manv
The pit first came under cultivation, but its relatively continu-
of the pitsfrom components of the El Riego phase, are ous distribution throughout our sequence starting in —
obviously from wild trees. Some of the seeds from El Zone XX of Coxcatlan Cave —
that makes me suspect
Riego phase zones, however, are noticeably larger than it was planted early. Archaeological examples of ama-
others. Some of the seeds from Coxcatlan time on seem ranth occur in the Southwest of the United States at
to fall within the range of size of modem cultivated about 4000 b.c. and in southeast Tamaulipas at about
avocado pits. The increased size, as well as the con- 2.500 b.c. Whether these were domesticated at all or in-
292
dependently in a number of areas or whether amaranth trude back in Tehuacan in the following phase, sug-
spread from a single center cannot at present be de- gest that the earliest cultivated maize spread out over
termined. a fairly large area from the Tehuacan environment.
The two segments from Zone XVI of Coxca-
of cotton A few rinds of bottle gourds (
Lagenaria siceraria)
tlan Cave, even though they resemble domesticated occurred in this phase. Since no wild relatives grow in
forms, I suspect were wild or intrusive because of Tehuacan, it was probably planted first elsewhere. As
the long temporal gap before we begin in the Ajalpan yet neither botanical studies nor archaeological find-
phase to have a continuous distribution of examples of ings indicate where or when this species was first culti-
this plant. The tinv fragment of com leaf in the feces vated, but remains appear in Tamaulipas in the In-
from the same zone was probably from a wild plant, fiernillophase from 7000 to 5000 b.c., in the earliest
since the few cobs from Zone XIII are predominately levels with preservation in Peru about 3300 b.c., and in
wild. The single Cucurbita pepo seed from Zone XIV the southwestern United States from 4000 to 5000 years
is considered to be a wild form by Whitaker and Cutler, ago. Shortly after the first appearance of gourds, the
but the other two squash seeds from this zone are large squash Cucurbita rnoschata is represented by a few
enough to be cultivated and are tentatively identified seeds. The modern wild
lack of either archaeological or
as C. mixta. These early specimens, perhaps over 7000 specimens Tehuacan and its present distribution and
in
years old, plus the fact that Tehuacan is near the center diversity in the Central American tropics suggest that
of diversity of mixta, indicate that mixta may have moschata was first planted in the latter area. It later
been first planted in or near the Tehuacan Valley. The spread northward, arriving in Tamaulipas by about
relatively lateappearance of mixta in Tamaulipas in 2000 b.c. and in the southwestern and eastern United
a.d. 200 and Southwest about a.d. 1000 suggests
in the States by the time of Christ. It also spread southward,
that it had slowly spread from the Tehuacan region. arriving in Peru by about 1000 b.c.
To summarize briefly, one can say that probably Late in the Coxcatlan phase white and black sapotes
Cucurbita mixta came under cultivation near the end ( Casimiroa edulis and Diospyros digyna), both native
of the El Riego phase, while domesticated or semido- to areas of heavy rainfall, spread into Tehuacan from a
mesticated chili and avocado may have been imported nearby wetter region. These fruits have not been found
into the valley at slightly earlier dates. Amaranth may in other archaeological sequences. Also appearing in
possibly have been planted about the same time, but this phase is a single pod of the common bean Phase (
examples of cotton, com, and pumpkin from El Riego olus vulgaris).Wild species of this plant do not occur
components are most likely wild. Also, the possibility in Puebla, but grow along the west coast of Mexico in
that future investigation may uncover evidence sup- Guerrero, Oaxaca, and Chiapas, as well as in Guate-
porting the case for the early cultivation of such plants mala. Common beans occur almost as earlv
J
as those in
as setaria,maguey, prickly pears, and mesquite, as sug- Tehuacan in the Ocampo phase of Tamaulipas (5000
gested by C. E. Smith, must not be forgotten. Thus by to 2200 b.c.). This distribution suggests that beans were
the latter part of the El Riego phase there may have first domesticated in west coast areas of Mesoamerica
been the most incipient kind of incipient horticulture, and early spread eastward to northeast Mexico and
with planted species composing a negligible part of central Mexico. Later they may have spread northward,
the diet of the valley’s occupants. arriving in Hopewell Kansas City about the time of
at
It is in the following Coxcatlan phase, beginning Christ. They may also have spread southward down
about 4900 b.c., that we have clear evidence among the the west coast, arriving in Peru about 500 b.c..
plant remains of a number of definite domesticates. All of these cultigens continued to be used in the
Avocado, chili, Cucurbita mixta, and chupandilla, Abejas phase in even greater amounts. The Tehua-
which we have discussed previously, were very prob- caneros or their neighbors seem to have developed a
ably planted, and amaranth continued in use. Wild new race of corn, pre-Chapalote-Nal-Tel from the early
corn, of which we had only a doubtful trace previously, cultivated race, and this spread northward over the
now becomes fairly prominent. This seemingly was of Mexican plateau to both Sonora and the Sierra Madre
local use, but more important, a few early cultivated region of Tamaulipas and perhaps even into the South-
cobs among the remains suggest that the Coxcatlan west of the United States. A few corn cobs with tripsa-
people had been selecting seed for planting. Similar coid introgression also were found in the Abejas hori-
cobs from Bat Cave and southwest Tamaulipas of a zon. Tripsacum, a grass which will hybridize with corn,
later date, probably between 2500 and 1000 b.c., as well has not been found in Tehuacan. It is possible that the
as examples of the early tripsacoid com race that in- early cultivated race may have spread into an area of
293
tripsaeum, become hybridized, and then spread back Rinds of the tree gourd ( Crescentic cujete were
to Tehuacan. found in a late Santa Maria level, as well as in more
Late in Abejas times (3000 to 2500 b.c.) three other recent ones. Examples have not been reported from
domesticates appear. One of these is the canavalia other excavations. Although wild species are found in
bean, which also may occur in Pern as early as 2000 b.c., nearby wetter parts of Mexico, they have not been
but sufficient study has not been undertaken to speak identified in Tehuacan. Little is known about the
of its origin or dispersal. However, there is little doubt “tomate” ( Physalis sp.), which was identified in feces in
that it is an intruder in Tehuacan. A second import is the same levels. Also, from a Santa Maria level tiny
the dog, which most certainly followed man across the fragments of what appears to be Manihot were found
Bering Strait from Asia and eventually southward to in feces. If this is Manihot esculenta, then we have some
Mexico. In Tehuacan we have evidence that this ani- ideas about its origin and distribution.
mal was frequently eaten. The final import is tepary Studies of modern species, along with the finding
beans ( Phaseohis acutifolius) The Tehuacan data sug-
. of archaeological griddles, suggest that this plant was
gest that the date of first domestication of this species first domesticated in northern South America. It may
must be pushed back in time and its place of origin have spread to Peru by 750 b.c. and perhaps to Ta-
moved southward down the Mexican coast. maulipas by way of the Gulf Coast as early as 1000 b.c.,
The next two phases, Purron and Ajalpan, add little and perhaps from there to Tehuacan by about 400 b.c.
to the list even though the people of
of domesticates, However, since Tamaulipas specimens have been
these early Formative periods were full-time agricul- classified as Manihot dulcis and since others say that
turists. One of the new eultigens is Cucurbit a pepo, and M. dulcis and M. esculenta are one and the same, our
considerable evidence indicates that it was first domes- tentative reconstruction is tentative indeed. One thing
ticated in Tamaulipas between 7000 and 5000 b.c. Evi- is certain: more investigation is needed on the origin
dently before it spread to Tehuacan, where it arrived and spread of this plant.
about 1000 b.c., it had spread northward to the South- The other agricultural specimens from the Santa
west of the United States. It occurred there as early Maria horizon are all various races of maize. The un-
as 2500 to 1000 b.c., and some of the evidence from raveling of the origin and spread of each
equally is
Mammoth Cave Kentucky hints that it was in that

in complex, although all of them may have originally de-
area at the time, although more secure evidence from rived from the wild and early cultivated races of Te-
more recent excavations dates its occurrence somewhat huacan. Early Nal-Tel-Chapalote varieties are found
later (Mangelsdorf et ul. 1964). earlier than those of Tehuacan both in Tamaulipas and
It is also in the Ajalpan phase that we have string Ocos, Guatemala, and their modem distribution is in
definitely made from cotton fiber. The early date the lowlands. I suspect our Tehuacan specimens came
(2500 b.c.) of cotton cloth in Huaca Prieta, Peru, to- from another area, even though the Nal-Tel-Chapalote
gether with botanical evidence, suggests this area is complex may have been derived from early cultivated
where cotton was first domesticated. It then spread corn that had spread from Tehuacan to the lowlands.
northward (perhaps along with the concepts of the The Nal-Tel-Chapalote race almost certainly derived
spindle whorl and the loom), arriving in Tamaulipas from the earlier pre-Nal-Tel-Chapalote corn of Te-
and Tehuacan between 1500 and 1000 b.c. and reach- huacan. Since we have not found the transitional steps
ing the Southwest only slightly later (Martin et al. in Tehuacan, however, it probably originated in a near-
1953). by highland region of Mexico and then spread back to
A large number of eultigens were added to the Te- Tehuacan. The tripsacoid Nal-Tel-Chapalote hybrids
huacan sequence in the Santa Maria phase. These sud- may, however, have evolved in Tehuacan. The Zapa-
den additions may in part be owing to better preserva- lote Chico race, which seems to have been in part de-
tion in the caves, but they also derive from new cultural rived from Chapalote, may have originated outside the
contacts. Runner beans (
Phaseohis coccineus) are Tehuacan region and then diffused to it. The earliest
present for the first time and date slightly earlier than South American corn specimens (1500 b.c.) seem to be
they do in the Southwest. However, wild varieties are related to the Chapalote or early Chapalote race that
absent from both regions, although wild examples oc- might have spread from Mexico to that region. The
cur as early as 6000 b.c. in Tamaulipas. These data race of slender-pod popcorn found in Santa Maria
suggest that runner beans were domesticated some- levels, however, was a derivative of Chapalote or of
where north of Tehuacan, west of Tamaulipas, and other domesticated corns of South America that de-
south of the Southwest. veloped in that region and then spread back to Puebla.
294
The hybrid may have

or tripsacoid slender popcorns noted too that there was no single unilinear develop-
developed Tehuacan. Again more investigation is
in ment of agriculture in any hearth or hearths but a series
needed to explain the complex origins and spread of the of small developments of plant domestication in many
Santa Maria com races. regions that stimulated and contributed to the evolu-
It may be in the Santa Maria phase, as well, that tion of agriculture over a wide area.
coyol, eosahuico, and eiruela came under cultivation in Having dealt with some of the wider implications of
irrigated fields or in orchards near watered areas. the agricultural remains, let us now consider how the
The Palo Blanco phase yielded several new domesti- sequence of domesticates helped establish chronology
cates. Three of these may have come from South in the Tehuacan Valley. As shall be noted in later vol-
America, where they occur in the wild state as well as in umes, the sequence of excavated zones and cultural
earlier archaeological contexts. One of these is the pea- phases suggested by the agricultural trends is in agree-
nut ( Arachis hi/pogaea), and the other is the guava ment with the sequence determined by artifact and
( Psidium g uajava). The third is less definite, because pottery studies, as well as by the radiocarbon deter-
only very small fragments in feces have been tentatively minations.
identified as pineapple. Turkey bones also were un- A study of the sequence of domesticated plants and
covered one Palo Blanco zone. Since the wild tur-
in animals from the zones of Coxcatlan Cave (Tc 50) re-
keys of northwest Mexico are closest genetically to the veals trends of domesticates as well as ever-increasing
domesticated ones, and since turkey bones appear amounts of agricultural produce. Comparisons of the
earlier in the Southwest of the United States (Martin components from other caves (Tc 254, Tc 255, Tc 35e,
et al. 1952) than in Tehuacan, we suspect that domesti- and Tc 272) containing large samples of preserved ag-
cated turkeys spread from the greater Southwest to ricultural food stuffs with those from Tc 50 allows one
Tehuacan. Also in the Palo Blanco horizon, we find one to align the former in a relative chronological position
cob of Zapalote Grande corn, along with a number of in terms of domesticated food trends established from
cobs of the early Chapalote race. Tc 50. This, of course, by no means allows us to place
Although a number of new varieties or races of older all the components we uncovered, but at least it gives
plants appear among remains of the Venta Salada a chronology of the ones with food stuffs.
phase, only one new species occurs. This is the sieva or The earliest zones of the other caves seem to be
small lima bean ( PhaseoJus lunatus), which is related to Zones E and F of Tc 254, which have examples only of
but probably was domesticated independently of the wild corn and chupandilla. These zones of San Marcos
larger South American lima bean. Its occurrence in late Cave seem to come after Tc 50, Zones XVI, XV, and
times in Tehuacan parallels its appearance in Tamauli- XIV, which contain no corn, but before Zone XI, which
pas and the Southwest. Thus there is little archaeologi- has both wild and a few early cultivated corn cobs, as
cal evidence concerning its origin. Botanical studies, well as chupandilla remains. The presence of a chu-
however, suggest it may have been domesticated from pandilla seed of fairly large dimensions in Zone F of
wild forms in Guatemala or southern Mexico and later Tc 254 hints that it may be more recent than Zones XII
spread to the three regions previously mentioned and XIII of Tc 50, which also have almost all wild corn
(which lack wild related species). Other plants intro- and no large-sized chupandilla seeds. Zone D of Tc
duced in the Venta Salada phase are two new varieties 254, with only a few more wild cobs than early culti-
of common beans
(11 and 13), another variety of run- vated ones, seems to come after Zone XI of Tc 50 but
ner bean and four races of corn: Conico, Dent,
(9), before Zone IX, which has more early cultivated cobs
Chalqueno, and Tepecintle. Exactly where the corn than wild ones. Further, like Zone X of Tc 50, it has a
races first originated is difficult to tell from present cob of early tripsacoid com as well as remains of black
archaeological and botanical information. sapote. Thus, it seems probable that Zone D of Tc 254
We have summarized the sequence of cultigens in and Zone X of Tc 50 are roughly contemporaneous.
the Tehuacan Valley and briefly pointed out their rel- Some middle zones of Purron Cave (Tc 272)
of the
evance to some of the problems of the origin and spread seem to fit Cave trends next. Zones
into the Coxcatlan
of New World plant and animal domestication. It L, J, and H contain dog remains, suggesting that they
should be obvious that considerable more research is come after Zone VIII of Tc 50 with similar remains.
needed to solve these problems, but at least we have Also, Zones H, K and I have larger proportions of
1
,
—
made a start I hope in the right direction. Even from early tripsacoid corn, while Zones G and H have cobs
our meager data one fact is apparent there were mul- — of the Nal-Tel-Chapalote complex. This again tends to
tiple origins of New World agriculture. It should be place Zones K-G after Zone VIII of Tc 50, because the
295
latter lacks any tripsacoid Nal-Tel-Chapalote cobs. sustenance, that is, trends of foods used. These esti-
However, the presence of a number of other plants in mates are derived from calculations based upon the
Te 50, Zone VII, would place Zones K-G before it. amounts of bone and plant remains uncovered in the
Zone C of Tc 254 also seems to fit into this general time excavated occupations and are supplemented by data
period, since ithas predominately early tripsacoid cobs, derived from analysis of coprolites. Such estimates not
as well as a few Nal-Tel-Chapalote, wild, and early only yield information that solve problems of chron-
cultivated cobs. Its exact relationship to the Purron ology, but they also are reconstructions of the sub-
Cave zones is difficult to determine exactly, but it might sistence at a given time by a given group of people and
fall between Zone J of Tc 272, which has more early reveal changes in the pattern through time.
tripsacoid than early cultivated cobs and none of the The bulk estimates of food derived from all dis-
Nal-Tel-Chapalote complex, and Zone I, with only early carded plant and animal remains uncovered in major
tripsacoid and Nal-Tel-Chapalote cobs. excavation are based upon calculations that are fraught
Zone F of Tc 272, with mainly Nal-Tel-Chapalote with We are here attempting to recon-
difficulties.
cobs, is very similar to Zones VII and VI of Tc 50, but structfrom the fragments in the refuse the total
it lacks the Zapalote Chico of Zone VI, hinting it was amounts of the sustenance of each floor of the exca-
earlier. Zones A-E of Tc 272, with many slender pop- vated sites in terms of liters of food represented. Ideally
corn cobs and slender popcorn hybrids, seem to come we would have liked to have this data transferred first
after Zone VI, which has only a very small proportion into liters of usable food, then into food energy units
of them, but the relatively small amounts of late tripsa- such as proteins, carbohydrates, and minerals. Un-
fats,
coid cobs and larger proportions of the Nal-Tel-Chapa- and nutritionists were
fortunately, since nutritional tools
lote complex suggest they come before Zone V of Tc 50. unavailable for this task, we were not able to proceed
The presence of Plujsalis and tree gourd remains in beyond the first step. Perhaps someone will take our data
Zones B and C of Tc 254 suggest that they are prob-
1
and complete the task, thereby providing data for re-
ably later than Zones VI of Tc 50 and c'of Tc 272. search into the problem of energy and culture.
How much later is difficult to determine, but the pres- Flannery, as explained in Chapter 8, was able to
ence in them of relatively large amounts of early tripsa- estimate the minimum number of whole animals used
coid and pre-Chapalote cobs suggests that they are or eaten during each occupancy. For example, two right
not more recent than Zone IV of Tc 50, which lacks femurs of deer from a given floor meant that at least
these races of corn. two deer had been eaten during this occupation. Since
The zones of the East Niche of El Riego Cave (Tc we estimate that a deer yielded about 20 liters of meat,
35e) all seem to be more recent than the components the people of the floor had consumed at least 40 liters
previously mentioned, but they are difficult to place of deer meat.

exactly. Zone E with a limited sample is the most diffi- Animals were classified into eight general groups on
cult to relate to the zones ofTc 50. The occurrence of the basis of size and yield: those giving about 20 liters
Variety 1 of common beans and the high proportion of of meat, such as deer; 10 liters, such as puma; 5 liters,
late tripsacoid corn cobs suggests that it is after Zone such as peccary; 2 liters, such as dog and coyote; 1.5
IV or Zone V of Tc 50. Zone D of Tc 35e, which contains liters, such as raccoon and fox; 1.0 liters, such as rabbit,
Aj)odanthera seeds, seems to come after Zone IV, and skunk, opossum, and turkey; 0.5 liters, such as cranes
the larger amounts of Variety 1 of common beans tends and other large birds, gophers, rats, and turtles; and
to confirm this placement. Zone C, with a similar corn finally small animals yielding less than 0.25 liters of
and bean picture, but with examples of Variety 9 of meat, such as mice, snakes, lizards, and small birds.
runner beans —present also in Tc 50, Zone III — sug- The tools for measuring and estimating the vegetal
gests further that perhaps Zones C, D, and E of Tc foods were a liter can (courtesy of Pemex Oil Company)
35e are after Zone VI of Tc 50 and before Zone III. This and a measuring cup marked in liquid ounces (33.8
leaves Zones A and B of Tc 35e and Zones I and II of ounces per liter). Our calculations were based upon
Tc 50 as roughly contemporaneous. The sample from fruits and vegetables bought in the Tehuacan market
Zone A of Tc 255 is so small it is difficult to place ex- place, others collected in the fields, corn kernels repre-
actly, but it has a large number of slender popcorn senting the various races from storage bins of the
cobs, as do the latest zones of Tc 35e and Tc 50. Rockefeller Foundation, and actual archaeological
specimens.
Another means of aligning the components with pre- Market examples of Cucurbita mixta, C. pepo, C.
served food stuffs is in terms of trends of estimated moschata, and bottle gourds with rinds removed were
296
mashed up in our liter can and found to average about about 120 kernels of a size to give 8000 kernels per
one liter of flesh apiece. Thus, we estimated an archae- liter. (We again used green kernels for this estimate.)
ological peduncle of one of these cucurbits as repre- Early tripsacoid cobs averaged about 160 kernels each,
senting one liter of food. It might be added that a series and about 4,000 kernels of this type (borrowed from the
of seeds and fragments of rind of the same species Rockefeller Foundation) filled our liter can. The Cha-
found in the same excavated square as the peduncle palote-Nal-Tel complex cobs each had about 310 kernels
would be estimated as representing only one squash of a size that took about 2,700 Rockefeller kernels per
and one liter of food. However, seeds of different sizes liter. Later tripsacoid com types had about 400 kernels
or seeds from widely separated squares of a floor would per cob, and it took about 2,000 kernels of this type,
be estimated as representing other specimens of squash. also borrowed from the Rockefeller Foundation, to
The flesh of avocados, guavas, and prickly pear and make a liter. Slender popcorn and later corn types had
other cactus fruits averaged about 30 examples per about 300 kernels per cob and were sufficiently large
smelly, sticky liter. Wild avocados were estimated as so that onlv 2,300 Rockefeller kernels were needed to
being about 60 fruits per liter. The flesh of 5 black fill our can.
sapotes equaled one liter, as did flesh of 8 white sa- Estimates of food plants such as amaranth, mesquite,
potes. With these multiple-seeded fruits, we were par- yucca, acacia, setaria, and grass (most of which were
ticularly careful to estimate the minimum number of seeds) were calculated by measuring in ounces the
whole specimens which our archaeological fragments actual archaeological specimens from each component.
represented by checking the spatial distribution as well Obviously estimating the bulk food represented by
as numbers of peduncles. About 400 peanuts (from the our archaeological remains on the basis of these rough
Hotel Pehafiel bar) filled our liter container. The fleshy calculations involves some degree of error. However,
interior scraped from ten Opuntia stems collected near the degree of error is consistent through time, and our
Coxcatlan Cave made one juicy liter. calculations do reveal food or sustenance trends. The
Also, the interior of 15 pochote pods from the same largest source of error is the lack of uniformity in the
locale filled the liter container. It took the meat of about preservation of plant remains from each floor. We
250 Condalia mexicana seeds (and an afternoon of therefore attempted to estimate, for those floors with
cracking seeds) to fill about a quarter of our liter can, poor or spotty preservation, what the total amounts of
so we calculated 1,000 of these seeds per liter. Our the various plants present would have been if plants
laboratory staff also spent an afternoon chewing pieces had been preserved over the whole floor. Some of the
of pochote root and the basal ends of agave leaves be- more recent floors— Zones I-IV of Tc 50, Zone C 1 of
fore reaching the conclusion that about one-half the —
Tc 254, and Zones A-D of Tc 35e did have preserv a-
food and juice (not very tasty) was removed in chew- tion throughout their whole area, so no calculations
ing. Thus, one agave quid was equal to the same were necessary.
amount of food. Since we could squeeze about sixty However, in other floors this was not the case. In our
wet quids into a liter can, we assumed that 60 quids had calculations of total amounts for floors with incomplete
once represented one liter of food. The flesh from 60 preservation, we divided the number of square meters
chili peppers from the market equaled a liter, as did the in which food stuffswere preserved into the number
flesh from 60 coyol, ciruela, cosahuico, or ehupandilla of square meters of the entire floor, and this gave us a
seeds (the latter all supplied by Augustin Tejeda). figure we called the “preservation factor” or “PF.” We
Wild seeds were estimated at 200 fruits per liter. Var- then multiplied the number of preserved plants in a
ious pods of various kinds of beans, mesquite, and floorby the preservation factor to find out the number
guaje ( Leucaeiui sp.) averaged about 20 seeds, and of plants the floor would have contained if plants had
roughly 5,000 of these seeds filled a liter can. been preserved throughout its area. For example, plant
The corn was harder to calculate. Mangelsdorf’s fragments were preserved in only two of the ten square
studies indicate that wild com averaged eight rows of meters covered by Zone A of Tc 255. Thus the preserva-
almost seven kernels each, and using kernels of small tion factor was 5. The remains of each plant species
green ears of about that size we estimated that about recovered from this floor were multiplied by 5 to cal-
Thus we could esti-
10,000 kernels filled a liter can. culate the total amount of each plant the floor may once
mate the amount of food represented by each
in kernels have held. These figures also are roughly equivalent to
archaeological cob of wild corn. Fragmentary cobs the nonperishable bone materials (and meat estimates
were thought of as being a quarter of a whole cob. A derived from them) covering the whole ten square
cob of early cultivated or early backcross corn had meters of Zone A. Thus we attempted to derive the
297
Table 36. Volume of Meat in Liters Estimated from Faunal Remains by Zone and Phase
WILD GAME DOMESTIC ANIMALS
Yielding Yielding Yielding Yielding Yielding Yielding Yielding

20 liters 5 liters 2 liters 1.5 liters 1.0 liters 0.5 liters 0.25 liters
of meat of meat of meat of meat of meat of meat of meat
Totals
(deer) (peccary, (coyote, (raccoon, (skunk, (large (mice, Turkey by Phase
puma) ring- fox, etc.) jackrabbit, birds, small (1.0
and Zone
tailed iguana, turtle, birds, Dog liters
cat, etc.) etc.) snakes, (2.0 liters of
etc.) etc.) of meat) meat)
animals animals animals animals animals animals animals

animals
meat meat meat meat meat meat meat
whole whole whole whole whole whole whole

bones
of bones
of bones
of bones
of bones
of bones
of bones
of bones whole bones
meat meat
Meat from Food

liters liters liters liters liters liters liters
of of
wild and from Percent
Identified identified estimated identified estimated identified estimated identified estimated identified identified estimated identified identified
estimated estimated estimated
Total Percent liters liters

Total Percent domestic all of meat
total total total total total total total
volume of diet volume of diet animals sources in diet
VENTA SALADA 278.0 1,591.1 17
Tc 35e, A 10 1 20 2 1 5 36 10 10 11 5 2.5 5 1 0.25 37.8 25 3 i 2 7 2 4 3 41.8 154.3
Tc 50, 1 10 1 20 4 2 2 1 1 0.5 1 1 0.25 22.8 12 1 i 2 1 1 3 1 25.8 192.5
Tc 50, II 6 2 40 3 1 1.5 12 4 4 2 2 1.0 1 1 0.25 46.8 15 3 2 4 4 1 5 1 51.8 319.1
Tc 35e, B 11 2 40 9 2 10 17 3 3 1 1 0.25 53.3 38 1 1 2 5 1 3 1 56.3 142.6
Tc 50, III 9 1 20 3 1 2 2 1 1.5 4 3 3 3 3 1.5 28.0 6 28.0 441.6

Tc 255, A 1 1 5 5 4 4 9 3 1.5 10.5 21 10.5 49.9
Tc 35e, C 15 2 40 4 1 5 26 9 9 6 4 2.0 6 3 0.75 56.8 19 24 2 4 15 3 7 3 63.8 291.1
PALO BUNCO 333.2 1,912.0 18
Tc 35e, D 16 2 40 8 1 5 41 10 10 4 2 1.0 1 1 0.25 56.3 49 23 2 4 4 3 60.3 114.6
Tc 35e, E 8 1 20 7 3 3 4 1 0.5 23.5 37 4 1 2 2 3 25.5 63.7
Tc 50, IV 25 2 40 1 1 1.5 7 2 2 4 3 1.5 45.0 15 5 2 4 4 1 49.0 298.8

Tc 254, B 6 1 20 3 1 5 36 4 4 14 5 2.5 4 2 0.50 32.0 35 3 1 2 2 2 34.0 91.9
1
Tc 254, C 1 1 1.5 2 1 1 2.5 17 2.5 15.1
Tc V
50, 10 2 40 3 1 1.5 9 3 3 1 1 0.5 1 1 0.25 45.3 12 1 1 2 2 1 47.3 364.0
Tc 272, A 1 1 1 1.0 2 1.0 44.3
Tc 272, B 3 1 1 1.0 2 1.0 41.0
Tc 272, C 2 1 20 3 1 1.5 2 1 1 2 1 0.5 2 1 0.25 23.3 25 23.3 91.4
Tc 272, D 5 1 1 1.0 1 1.0 109.7
Tc 272, E 2 1 20 9 1 1.5 15 4 4 25.5 8 25.5 323.7
Tc 50, VI 46 2 40 4 1 5 3 1 1.5 29 7 7 3 3 1.5 2 1 0.25 55.3 18 1 1 2 1 1 3 1 58.3 309.6
Tc 272, F 3 2 2 1 1 0.5 2.5 5 3 1 2 2 5 4.5 44.2
SANTA MARIA 151.5 607.0 25
Tc 50, VII 66 4 80 2 1 5 25 9 9 1 1 0.5 4 2 0.50 95.0 32 95.0 297.5
Tc 272, G 17 5 5 5.0 3 5.0 169.1
Tc 272, H 4 1 20 26 6 6 3 2 1.0 27.0 40 7 1 2 2 3 29.0 68.2
Tc 272, 1 3 1 20 7 2 2 1 1 0.5 22.5 31 22.5 72.2
AJALPAN 24.0 70.2 27
Tc 254, C 1 1 1 1.0 3 1.0 30.3
Tc 272, J 2 1 20 2 1 1 21.0 35 3 1 2 2 3 23.0 59.9
ABEJAS 202.8 678.6 30
Tc 50. VIII 21 2 40 7 2 2 1 1 0.5 42.5 26 42.5 163.1
Tc 50, IX 44 3 60 1 1 1.5 13 2 2 1 1 0.5 64.0 34 64.0 185.3
Tc 254, D 22 4 4 4 2 1.0 2 1 0.25 5.3 20 5.3 26.2
Tc 50, X 60 4 80 2 1 5 1 1 2 6 2 2 89.0 29 1 1 2 2 1 91.0 3 04.0
COXCATUN 334.6 975.2 34
Tc 50, XI 143 9 180 3 1 5 27 8 8 4 3 1.5 194.5 34 194.5 569.0
Tc 254, E-F 1 1 20 1 1 5 1 1 2 27 4 4 9 5 2.5 7 1 0.25 33.8 48 33.8 70.3
Tc 50, XII 39 3 60 2 1 1.5 8 3 3 1 1 0.25 64.8 33 64.8 196.8
Tc 50, XIII 15 2 40 3 1 1 1 1 0.5 41.5 30 41.5 139.1
ELRIEGO 487.8 902.4 60
Tc 50, XIV 91 8 160 26 6 6 1 1 0.5 2 2 0.50 167.0 55 167.0 3 04.9
Tc 50, XV 67 6 120 1 1 5 1 1 1.5 16 4 4 130.5 61 130.5 213.8
Tc 50, XVI 100 9 180 24 10 10 1 1 0.25 190.3 50 190.3 383.7
298
amount by using three sets of

of food for each floor avocado pits were sufficiently large to suggest that they
data: liters of meat estimated from minimum numbers came from planted trees, and multiplied by the PF
of whole animals derived from a study of the bones, they represented about 6.6 liters of food. Remains of
liters of plant food derived from a study of the actual various wild plants, including sizable amounts of fine
plant remains and estimates of the bulk yield of modem edible seeds and six pits of what were probably wild
plants, and a calculation of the amounts of each plant avocados from the mountains nearby, were estimated
one would have found in a given floor if preservation as representing 72.3 liters, or 34 percent of the total
had been constant over the whole area. Again let me food. Meat accounted for 61 percent and food from
state that the figures derived from these data are esti- cultivated plants about 3 percent of the total diet.
mates, and there is probably a large margin of error in Zone XIV of Tc 50, the final El Riego phase occupa-
each calculation. However, the figures do reveal trends tion,had slightly better preservation (a factor of 20).
of subsistence, and the errors are constant through The bones represented at least eighteen animals that
time. These figures are presented in Tables 36, 37, and might have yielded about 167 liters of meat or 55 per-
38. cent of the total food for the zone. Two avocado pits
The earliest floor for which we attempted to calcu- and three chilis, along with two seeds of at least one
late the bulk sustenance was Zone XVI of Coxcatlan cultivated squash plant (probably Cucurbita mixta )
Cave. Below that level plant preservation was too poor represented 22.6 food or 7 percent of the total.
liters of
to make estimates of the sustenance feasible, and even Fine seeds again were a major item among the wild
in Zone XVI the sample is barely adequate. In this plants, but stems or leaves and quids were also fairly
zone we uncovered 125 bones. Flannery (Chapter 8) prominent. The wild food plants were estimated as be-
identified 100 of them as being from at least nine deer, ing 115.3 liters, or 38 percent of the total food.
representing 180 liters of meat. Twenty-four bones Combining the three meager floors of the late El
came from ten animals that would yield one liter of Riego phase, even though the sample is not very large
meat each, and one lizard bone might represent 0.25 and our estimates may have a wide range of error, does
liters of meat. Thus about 190.3 liters of meat must have give us a set of figures to start with. The data as a whole
been consumed by the occupants of this zone. Plants from the three floors reveal that 372.6 liters or 41 per-
were preserved in only one square meter of this floor cent of the total food came from wild plants, 40.3 liters
covering about 50 square meters, so our preservation or 5 percent from cultivated plants, and 487.8 liters or
factor was Using this factor, we estimated that three
50. 54 percent from meat.
larger avocado pits were equal to about 5 liters of food A glance at Tables 36-38 shows that the preservation
from a cultivated plant. That is, the three pits were factor decreases as preservation improved in the floors
multiplied by the PF of 50, and the resulting 150 pits assigned to the Coxcatlan phase. Quids and leaf or
were divided by 30, the number of avocados per liter, stem fragments begin to outnumber small edible seeds
to give 5 liters of avocado flesh. The chili pepper was among the wild plant remains. New species of culti-
estimated as representing 1.66 liters of food from a vated plants include com, squash, beans, and black
cultivated plant. Thus food from cultivated plants, and white sapotes. The wild plant foods represented a
rounded to one decimal, was calculated as 6.7 liters. total of 500.7 liters or 52 percent of all foods combined.
The estimated amounts of food from wild plants were Food from cultivated plants (including half the amount
much larger, as Table 37 demonstrates, representing a of chupandilla represented in Zone XI) totaled 139.9
total of 185 liters of food. In terms of percentages the liters or 14 percent. Meat decreased in importance,
occupants of Zone XVI seem have obtained 48 per-

to representing 334.6 liters of food or 34 percent of the
cent of their food from wild plants, 2 percent from do- total. The trends through the El Riego and Coxcatlan
mesticated plants, and 50 percent from meat. phases are similar, with both wild and cultivated foods
The picture from Zone XV is similar, and again pres- increasing and meat diminishing.
ervation was far from perfect. The bones indicate that Preservation was poor again in the earliest occupa-
about 130.5 liters of meat might have been consumed. tion of the Abejas phase, Zone X of Coxcatlan Cave, but
Three fragments of chili pepper epidermis recovered improved to a factor of 5 in each of the remaining three
by flotation from the dust of three widely separate zones. Zone X contained the remains of one dog, but
squares were assumed to come from three different the two liters of meat it represented would have been a
chilis.Since about two of 80 squares had preservation, negligible part of the diet. The major portion of the
the preservation factor was 40 and the number of chilis food in these levels still came from wild plants, in-
became 120, representing 4.4 liters of food. Five of 11 cluding large amounts of pochote pods and roots, agave
299
quids, and Opuntia stems. Foods from wild plants were plant remains 17 percent, and meat 18 percent. Thus
estimated as 331.3 liters or 49 percent of the total diet. the trend that started in the Abejas phase, with agri-
Cultivated plants, including new species of beans, fur- produce rising sharply at the expense of both
cultural
nished an increasing amount of food, estimated at 144.5 meat and wild plant products, continues into Palo
liters or 21 percent of the total. Meat accounted for Blanco times.
202.8 liters or 30 percent of the food consumed. The trend toward increasing agricultural produce
Food remains representing the Purron phase were (75 percent) continued into the Venta Salada horizon,
so meager that we did not estimate the food used. The with wild plant foods (8 percent) decreasing much
Ajalpan phase had a slightly better sample, but it can- more rapidly than meat (17 percent). The most reliable
not be considered adequate. The only floors contain- estimates can be made for this phase because six of its
ing plant fragments were Zone J of Purron Cave and seven components (Zones I — III of Tc 50 and Zones
Zone C of San Marcos. The former had preservation A-C of Tc 35e) had food stuffs preserved over the
in only one of its 50 square meters. The latter zone entire areas occupied, and even Zone A of Tc 255 had
covered approximately four square meters, one of preservation over a fifth of its living surface.
which had preservation. These two small summer oc- In terms of over-all trends of sustenance, our rough
cupations contained only 90.2 liters of food, of which estimates reveal that agricultural food increased
55 percent is from cultivated plants, 18 percent is from throughout the Tehuacan sequence, with the late El
wild plants, and 27 percent represents meat. These Riego phase deriving from this source 5 percent of the
proportions are probably somewhat distorted by the total food represented; Coxcatlan, 14 percent; Abejas,
unrepresentative sample for this phase. The trends 21 percent; Purron, a projected 35 percent; Ajalpan, be-
from the Abejas to the Santa Maria phase would sug- tween 40 and 55 percent; Santa Maria, .58 percent; Palo
gest that the Purron phase, for which figures are Blanco, 65 percent; and Venta Salada, 75 percent. Wild
lacking, derived about 35 percent of its diet from plant foods show a gradual increase through Coxcatlan
agriculture, perhaps 31 percent from meat, and about times and then a gradual decrease through the rest
34 percent from wild plants. If we had a larger sample of the sequence. Wild plants probably accounted for 30
for the Ajalpan phase, we would probably see the percent of the diet in the Ajuereado phase, and our esti-
figures we listed above corrected to about 40 percent mates show they represented 41 percent of the diet in
from agriculture, 31 percent from wild plants, and 29 the later El Riego phase.They increased to 52 percent
percent from meat. in Coxcatlan times, and thereafter decreased over the
The sample improved for the Santa Maria phase, remainder of the sequence, to a mere 8 percent in the
largely because of well-preserved materials in Zone Venta Salada phase. Meat diminishes throughout the
VII of Tc 50 (with a preservation factor of 2). Remains sequence. We might say it represented nearly 70 per-
from three earlier zones (G-I of Tc 272) averaged 73.2 cent of the Ajuereado people’s diet, and our estimates
liters of food from agricultural plants, or 71 percent of for the El Riego phase show that accounted for 54 it
the total food for the three zones; 18.2 liters of meat percent of the diet. This percentage
34 in Coxca- fell to
from wild animals, or 18 percent; and 11.1 liters from tlan times, and thereafter diminished gradually, to a
wild plants, or 11 percent. The bones of one dog in final 17 percent in the Venta Salada phase.
Zone H represented 2 liters of food and less than one Obviously, one supplement to or check on the esti-
percent of the total food for the three zones. mate of food trends from the bones and plant remains
The much more adequate sample in Zone VII of recovered from the refuse is the sustenance trends re-
Tc 50 revealed that agricultural remains, predomi- vealed by Callen’s analysis of food particles in the
nately corn, represented 132.8 liters of food or 45 per- feces. Determination of trends from this source, how-
cent of the total diet, wild plants 69.7 liters or 23 per- ever, presents a number of difficulties. First of all,
cent, and meat 95 liters or 32 percent. Combining all certain foods digest more readily than others, and
the Santa Maria refuse material reveals wild plant re- particles may not be discernible in the fecal remains.
mains representing about 17 percent, meat 25 percent, Second, certain ways of preparing food, such as soak-
and domesticated plants 58 percent of the total food. ing corn kernels in lime before grinding them into a
The Palo Blanco components, with the exception of fine paste, make it difficult to discern the particles in
Zones A-F of Purron Cave, had the best samples of the feces. Third, there is usually not an adequate
food remains. When one combines the food estimates sample of feces from any single zone. Fourth, it is diffi-
from all Palo Blanco components, foods from agricul- cult to quantify the various proportions of food repre-
tural remains comprise 65 percent of the total, wild sented in the feces. Finally, there is no fool-proof
300
Fig. 186. Changing trends in the importance of the principal sources of food.
method of differentiating between human and non- of bones, and more important, by the factor of diges-
human feces. tion.
In terms of general categories, the trends of The trend of wild vegetal remains in the feces also
sustenance as revealed by the feces are very different bears small resemblance to the trends of wild plant
from that derived from estimating plant and animal remains from the refuse on the floors. In the feces, wild
remains in the garbage. Meat, instead of beginning as plants are always dominant, though they generally de-
the dominant food stuff and gradually decreasing, decrease throughout the sequence, but the trends based
creases in the coprolite sample from middle levels and on preserved plant remains show wild plants increas-
then increases in later times. Further, the meat as re- ing from El Riego to Coxcatlan time and decreasing,
vealed in the feces is always a small proportion of the at times sharply, thereafter. Part of this difference un-
total diet, never as large as the smallest percentage of doubtedly arises from the fact that many of the wild
meat as determined from the bones. This difference, vegetal foods are not readily digestible and therefore
I believe, can be accounted for by the fact that our are preserved in large amounts in the feces. However,
sample of feces is less representative than the sample when one looks at various proportions of the different
301
kinds of wild plants, estimates from the feces and from cado, chili peppers, and various seeds from pods were
the refuse do bear some resemblance to each other. probably eaten raw or with little preparation other
The dominant wild plant remains in both the garbage than cutting or picking them from tree or bush. The
and feces in El Riego times are fine seeds (mainly pattern of wear exhibited by the teeth of El Riego
setaria). In Coxcatlan and Abejas levels leaf or stem skeletons gives evidence of the gnawing of raw leaves.
material increases, and in the later phases fruits show Agave leaves, pochote roots,Opuntia stems, mesquite
an increase. seeds, and grass seeds in feces seem to have been
However, the big difference in proportions shown roasted. Such seeds as amaranth, Lemaireocereus,
by the two sources is in terms of the domesticated Opuntia, and setaria, seem to have been ground with
plants. Although particles of agricultural plants do in- a stirring or pounding motion (and were perhaps
crease in variety and slightly in proportion throughout roasted). The presence of mullers, milling stones, mor-
the sequence in the coprolites, they never approach the tars, and pestles in El Riego levels indicates that such
sort of dominance in the later levels that they do in a technique actually was employed.
the estimates based on preserved plant remains among The food remains and coprolites of the Coxcatlan
the refuse. This is almost entirely owing to the fact phase show that some bones were scraped, some meat
that little or no corn hasbeen identified in the feces, was eaten raw or steamed, and some meat was roasted.
whereas in the garbage estimates the tremendous in- Fruits and vegetables that were probably eaten raw
crease in corn is the main reason agricultural produce comprised about the same proportion of the diet as
increases. This lack of corn in the feces,
I believe, can fleshy plant products such as agave and grass leaves,
be accounted for by the following factors: com ker- Opuntia stems, and pochote roots, which apparently
nels were soaked in lime to remove the tough pericarp, were roasted before being eaten. Small seeds increase
and from Abejas time onward, the soaked kernels were in number in this phase and perhaps represent over 10
finely ground to make nixtamal, a highly digestible percent of the food used. Some of these separated from
product which would be very difficult to distinguish in the feces appear to have been milled, and mullers, mill-
the feces. Also, some of the corn represented by cobs ing stones, mortars, and pestles were found in increas-
in the caves may have been exported to settlements ing numbers. The presence of corn cobs and a few
and not eaten by the cave dwellers. manos and metates may mean that a tiny amount of
For the reasons I have stated above, I believe our the food was finely ground.
study of the bulk food remains is a fairly reliable indi- In the Abejas phase more meat was eaten raw or
cator of the foods actually used throughout our se- steamed than roasted, and marrow or bone stew was
quence, and a more accurate indicator than the food becoming less popular. Raw vegetables and fruit prob-
particles in the coprolites. The feces, however, do reveal ably represented about 40 percent of the diet and more
certain foods not found in the refuse,and they give a than twice the share of roasted wild plants and cu-
clear indication of what wild plants were utilized, as curbits. Only a small proportion of the total food is
well as how they and the other foods were used or represented by seeds of the type that could be milled.
prepared. Increased amounts of com cobs and metates and
Of course, ancient food-preparation techniques are manos indicate that some of the food was finely
difficult to describe precisely or quantitatively. How- ground. The presence of beans together with burned
ever, a consideration of the percentages of food used in matter inside one stone bowl suggests that the tech-
each phase knowledge gained from study
in light of the nique of boiling food might have been known.
and from analysis of the artifacts and
of the coprolites In the Ajalpan phase a fairly large number of pots
physical anthropology does indicate some trends and with burned interiors or carbon adhering to the in-
changes. evidence that some food was cooked
teriors are reliable
In the El Riego horizon 54 percent of the food was by The large number of manos and metates,
boiling.
meat. The presence of a few scraped interiors of bones the type of wear on the human teeth, and the large
may indicate thatmarrow or bone stew was one way proportion of corn cobs suggest that nearly a third of
food w'as prepared, and carbon found with the bones the food was finely ground. The lack of comales and
may mean that a little of the meat was roasted. How- the remains of food adhering to the inside of pots may
ever, evidence from the feces suggests that most of indicate that the corn was eaten as gruel. Fruits, both
the meat was eaten raw (or perhaps was steamed inside wild and domesticated, comprised about a fifth of the
leaves). The wild composed 41 percent
plant remains diet and were probably eaten raw. Analysis of the
of the diet in the El Riego phase. Chupandilla, avo- coprolites show that cactus stems, agave leaves, and
302
Fig. 187. Changing popularity of methods of preparing foods.
similar plant products were roasted. Seeds plus the a third of the total food, but a large amount of corn was
milling stones and mullers suggest that a small fraction still only finely ground and accounts for about a fifth
of the diet was milled. Probably half the meat was of the diet. A few seeds still were milled according to
roasted and the other half was eaten raw or steamed. the mortars, not to the seeds themselves. Some leaves
Only one bone was scraped. and stems were roasted and others were eaten raw.
In the Santa Maria phase meat was treated in about Most of the meat seems to have been roasted, but a
the same manner. About a quarter of the diet consisted little may have been boiled or eaten raw. The smell of
of fruits and the like which were eaten raw. Roasted one coprolite suggests brewing, a technique which may
cucurbits, leaves, stems, and roots accounted for less have become known in this phase.
than a fifth of the food, and a small fraction was still In the Venta Salada phase the small proportion of
derived from seeds that were milled. A few pots and meat used was probably roasted, although the presence
the presence of beans suggest some boiling of food. of one mouse suggests boiling. Probably beans and
Corn made up over a quarter of the diet, and the pres- some of the ground corn were also boiled. Fruits were
ence of many manos and metates, as well as the pattern still eaten raw, and leaves and stems were roasted.
of wear on the human teeth of this phase, suggests that Large numbers of ceramic molcajetes undoubtedly in-
the kernels were finely ground. The presence of clay dicate the grating of chili peppers, and a large number
comales, on which tortillas traditionally are baked, of comales suggests that perhaps two-thirds of the
suggests that about a fourth of the ground corn was ground com was baked as tortillas. Brewing was prob-
grilled. The absence of corn in the coprolites repre- ably carried over from Palo Blanco times, as I know
senting this phase is probably attributable to the prep- of no (Indian) culture that has given it up after the
aration, which made it highly digestible. first drink.
The trend of baking ground corn on comales con- The discussion of the food used, or the sustenance,
tinues in the Palo Blanco phase and accounts for about and its preparation leads to an even more speculative
303
consideration of how the people got this food — the Coxcatlan phase, 27 percent in the Abejas phase, 22
subsistence activities. Obviously our classification of percent in the Ajalpan phase, 20 percent in the Santa
the food into meat, wild plants, and cultivated plants Maria phase, 13 percent in the Palo Blanco phase, and
suggests hunting and trapping, plant collecting, and 12 percent in the Venta Salada phase. Trapping and
agriculture —or more simply, food collection and food collecting remained roughly the same, with El Riego,
production. I believe, however, that a study of the plant Coxcatlan, and Abejas each deriving about 3 percent
and animal remains in conjunction with the associated of the total diet from this source, Ajalpan, 1 percent;
— together with some general ethnographic
artifacts Santa Maria, 6 percent; Palo Blanco, 3 percent; and
knowledge— can given us concrete information about Venta Salada, 2 percent. The tools of the latter trade
specific subsistence activities. We must fall back upon —
were the snare with slip knot and probably a quick
our bulk food data, the results of the food-getting en- movement of the hand or a well-aimed rock.
deavors. Ideally, a study of this sort ought to be based Study of the projectile points hinted that the hunting
on the number of man hours needed to obtain the was probably of different kinds. In terms of function,
various kinds of food. I strongly suspect, for example, projectile points seem to have been used in at least
that the number man
hours needed to obtain a thou-
of four ways. Small points like Harrell, Starr, Tula, and
sand liters meat by hunting would be far
of deer Teotihuaean, which in Tehuacan were associated with
greater than the number of hours needed to derive arrow shafts and in Tamaulipas were sometimes found
a thousand liters of edible corn from planting. Unfor- attached to arrow shafts, are considered arrow points
tunately, when this study was made, we had no means and indicate rapid-fire, bow-and-arrow shooting of
by which we could transfer the estimates of bulk food game. Large points and points that are generally leaf-
into the man hours necessary to obtain that amount of shaped, such as Lerma, Plainview, Agate Basin, and
food. Perhaps such ethnographic information exists, El Riego, are lance points and were probably used as
but I do not know of it. spearing implements in ambush hunting.
In spite of this limitation, let us consider how meat Flannery has pointed out that the other projectile
was obtained from El Riego to Venta Salada times. The points could be divided into two groups, both used to
bone materials show that the sources of meat can be tip atlatl darts: (1) Thin, sharp, weakly barbed points
—
divided into two classes large animals, such as deer, such as Pelona, Catan, Coxcatlan, Garyito, Zacatenco,
puma, and peccary, which would probably be hunted; and Matamoros, which would slash into and deeply
and small animals, such as various rodents, fox, skunk, penetrate an animal, and if they failed to hit a vital
turtle, lizard, and certain birds, which most likely point, would cause the stricken animal to bleed so that
would be trapped or collected. Further, the artifacts we it could be trailed until it dropped. (2) Large, rela-
found seem to fit these two general categories. The tively thick, barbed points which would remain im-
spear, atlatl, and arrow points and the broken shafts of bedded in the prey and cause the animal to drop if
atlatls and arrows must have been for hunting and kill- enough were fired into it. The latter heavy dart points
ing animals, whereas the slip noose and fragments of were probably used in stalking game, perhaps by a
snares and nets must have been connected with trap- number of hunters, while the former may have been
ping. Thus at the outset we grouped our bones into more often used by individuals.
those representing hunted animals and those represent- The proportions of the various kinds of point types
ing trapped or collected beasts. change through the levels in which we found animal
In terms of phases, late El Riego (Zones XVI, XV, bones indicating hunting. In the El Riego phase, with
XIV of Tchad the remains of 24 large animals and
50) 51 percent of the total diet from meat of large animals,
25 small ones; Coxcatlan, 18 large and 28 small ani- 21 percent of the points are lance points and 79 percent
mals; Abejas, 10 large and 17 small animals; Ajalpan, atlatl stalking points, suggesting that 11 percent of the
1 large and 2 small ones; Santa Maria, 7 large and 28 meat came from ambushing or stalking with lances and
small animals; Palo Blanco, 15 large and 69 small ani- that 40 percent came from stalking with darts. Thirty-
mals; and Venta Salada, 14 large and 63 small animals. nine percent of the food of the Coxcatlan phase was
Thus there is a definite trend from more animals’ being meat from hunted animals. Twelve percent of the
hunted to more animals’ being trapped and collected. points were lance tips, 59 percent were heavy dart
Moreover, in terms of the amount of meat produced points, and 29 percent were thin dart points, suggesting
by the two techniques, hunting had decreased. Meat that 5 percent of the meat came from ambush hunting,
from big animals represented about 51 percent of the 23 percent from stalking, and 11 percent from bleeding
total diet in the late El Riego phase, 39 percent in the and trailing. The Abejas phase derived 27 percent of
304
Table 37. Volume of Food in Liters Estimated from Remains of Wild Plants by Zone and Phase
Setarlacl Prosopis
Prosopls Cyrtocarpa Slderoxylon Spondias Acrocomia Leucaena Ceiba Opunlia
Yucca mactoslachya lemaireocereus
mexicana julillora esculenta parvillola and other
porlculou jullllwa ameiicana procera cf.templsque
and other Cephaleocereus
(avocado) (chupandilla) (cosahuico) (ciruela) (coyol) (mesquite) (guaje) (pochole) cadus stems
flrolo) (mesqullo) Escontrla
grass seed
chiotilla
1 §
i ?
i I
!1 s
o
S
|
3
oof
VENTA SALAOA
Tc ~23 73 0.7
35o, A 242 242 4.0 29 29 0.5 2 2 0.2
Tc 50, 9 0.3 274 224 3.7 20 20 0.3 10 10 1.0
Tc 50, II 19 0.6 2 2 46 46 i
1,268 1.268 21 1 58 58 1.0 78 78 7,8
Tc 35o, 0 II 11 0.3 7 0.2 2 2 181 181 3.0 27 27 0.5 7 7 0.7
Tc 50, III 14 0.5 2 2 17 340 0.1 13 13 0.2 765 765 12.8 57 57 1.0 69 69 6,9
Tc 755, A 600 0.1 12 60 1.0 13 65 6.5
Tc 350, C 16 16 0.5 23 460 0.1 12 12 0.2 712 712 11.9 85 85 1.4 12 12 1.2
I'AlTTDOfTCTJ
Tc 35o, 0
Tc 350, E
i r 328 328 5.5
Tc 50, IV 22 440 0.1 155 155 2.6 93 93 1,6

Tc 254, D 3 0.1 13 13 0.4 26 520 0.1 10 10 0.2 295 295 4.9 139 139 2.3
1
Tc 254, C 24 480 0.1 57 57 1.0 123 l?3 2.1 49 49 0.8
Tc 50, V 48 1.6 6 18 0.6 37 111 1.9 679 2,037 34.0 39 117 2.0
Tc 212, A
Tc 212, 0
Tc 212, C 30 120 2.0 10 40 0.7
Tc 212, 0 201 1,005 16.8 81 405 6.0 50 5.0
Tc.272, E 709 107
3,545 59.1 535 8.9 90 9.0
Tc 50, VI 19,7 19.7 5.7 258 258 17.2 1,254 1,254 20.9 50 5.0
Tc 272. F
63 13 (M 165 330 22.0 68 136 2-3

Tc 212, G 4 80 1.3
Tc 212, II
Tc 272, 2 100 1.7
Tc 50, VIII "1 100 * 4 400 0.1 238 1,190 70 1.2

Tc 50. IX 62 310 5.2 14 70 1.2
Tc 254, 0
4 400 0.1 5 500 0.1 7 35 0.6
Tc 50, X 30
10 300 20.0 1 0.5
COXCA LATT
O-
I
Tc 50, XI 34.6 '776.8 *

32 0.2 2 16 0.5 7 56 0.9 7 56 0.3 177 1,416 94.4 1 160 342 2,736 45.6 5? 416 6.9
Tc 754, E-F
16C 1.920 32.0 1 12 0.2
Tc 50, XII 106.5 1,778 37.8
1 12 0.4 4 48 0.8 1 12 0.1 4 960 0.2 106 1.272 21.2 5 60 1.0
Tc 50, XIII
60 0.3 63 756 12.6
EFR1EGD
Tc 50, XIV
Tc 50, XV
64,1)
73.8
1 ,280
95?
37.8
78.7
"
79
13
580
53?
17.2
15.4
10
3.3
700
13?
5.9
0.7
1 60 1.0 1 20 0.1 2 40 0.2 1 20 0.7 19 380 6.3 4 80 1.3
6 240 4.0 3 120 0.6 3 120 2.0
Tc 50, XVI 45.7 7,760 66.9 10 500 14.8 5 250 7.4 4 200 3.3 37 1,850 9.3 5 250 1.3 3 150 5.0 1 50 0.8
Table 38. Volume of Food in Liters Estimated from Remains of Cultivated Plants by Zone and Phase
Cucurblta Lange narla Zea mays Cucurblta Dlospyros Casimlroa Phaseolus Cyrtocarpa Zea mays Zea mays Zea mays Canavatia Phaseolus Cucurbila Sideroxylon Acrocomia Spondias Ciescentia Zca mays Zea mays Zea mays Phascolus Zea mays Arachis Psidium Phascolus
annuum ameileana slcerarla imarzc) moschala digyna edulis vulgaris (maize) (maize) (maize) spp. acutilolius repo mexlcana mombin cujele (maize) (maize) (maize) cxcineus (maize) hypogaea guajava lunalus
Icommon bean) quids Nal-lel- (lack lepary (pumpkin) lempisque [coyol cuuela) (Wee gourd) late Wipsacoid Zapalote Slender Pop rac (runner bean) Zapalote (pcariul) (guava) (sieva bean)
(chili (avocado) (squash) boltle gourd) carlv (squash) black sapole) while sapole) (chupandilla) early Wipsacoid
cultivated race Chapalote complex beans) bean) (cosahuico) Chico race Grande race
popper)
E E E
E * 1 E E E
& i E 5,
E E E E E E
e i E ;
&
I Ji £
* 1 !jj i J «
i
1 i i % § 1 8
S §
3 C 1 1 1 ! S | -§ S3
I
i
1 I I
it 1
i § i f 1 1
1 § f i
i I | i i
E i 1 5 l 111
! 5 1 § 1 I i 1 li 1 E
&
h a
8 ill 1 s 1 1
! i 5 1 1 1
g g g
lg?
c
3
M TS
1 4 IB s*l I g g i g g g 1 i I g g I §
-g
5! ^ i t S g 1
fc •g
| | 0-
g g i i I 1 i g 1 1 ? ! 1
B !! fl B * 3 o a a V 0 0 • e 0 a a e o 0 0 0 0 0 0 0 0 0
1 g £ s d a s g i S 8 s a 8 s S S i s 8 s s 8 8 S 8 S s 8 8 8 8 s 8 S f s 8 § s S 1 s 8 8 £ s 8 S s 8 s s 2 k S § sis
VENTA SALAOA
Tc 35o, A 0 8 8 0.26 6 6 0.7 8 1 1.0 1 1 1.0 1 1 0.2 19 8 1.0 41 60 0.01 31 31 0.5 1 1 143 44,330 16.4 5 5 so 1 1 2 2 233 93,200 4 6.6 23 6,900 3.5 209 62,700 27.2 Tc 35c, A
Tc 50, 0 3 3 0.1 17 1? 170 19 2 ?.<) 1 1 1.0 1 1 0.2 1 1 0.1 4/ 47 0.6 22 3,520 0,9 11 II 0.2 636 197,160 73.0 3 3 3.0 9 9 0.2 4 4 0.1 4 215 86,000 43.0 123 36,900 16.0 Tc 50.
Tc 50, II 0 1 1 0.03 6 6 0.2 7 7 0.2 13 13 13.0 11 1 1 0 26 1 1.0 / 5 1.0 2 2 0.3 119 119 1.9 1 160 22 22 0.4 601 186,310 69.0 1 1 1.0 43 43 U./ 4 4 0 1 63 63 353 141,200 70.6 482 144.600 62 9 Tc 50. II
Tc 35o, 0 0 2 7 0.06 8 8 0.3 1 1 1.0 3 1 0.2 17 8 1.0 196 200 ow i; 17 0.3 160 '49,600 18.4 1 1 1.0 2 2 1 1 105 42,000 21.0 4 1,200 0.6 264 79,200 34.4 Tc 35o, 0
Tc 50, III 0 78 78 2.3 7 7 0.2 19 19 190 15 2 2.0 1 1 1.0 3 3 0.6 3 3 0.4 3 40 0.01 136 136 2.3 18 2,880 0.7 30 30 0.5 1,967 608,220 225.3 1.0 58 58 0.9 5 5 0.1 25 25 416 166.400 83.2 117 35,100 15.3 1 20 0.01 1 20 0.01 Tc 50. Ill
Tc A
255, 5 1 5 50 3 1,800 0.2 1 5 0.6 2 1(1 0.7 7 10,850 4.0 33 4 9.500 21.5 Tc 255, A
Tc 35o, C 0 2 2 0.06 8 8 0.2 8 8 0.3 9 1 1.0 1 1 0.2 66 7 0.9 156 180 0.03 12 12 20 0.8 1/ 1/ 0.3 466 144,460 53.5 2 2 2.0 8 8 0.1 2 2 2 2 479 191,600 95.8 418 125,400 54.5 6 20 0.01 1 20 0.01 Tc 35o, C
PALO HLANCO PALO BLANCO
Tc 35o, 0 0 2 2 0.06 8 8 0.3 2 2 7.0 2 1 1.0 1 1 1.0 1 1 0.2 53 20 2.5 109 109 0,02 1 1 140 43,400 16.1 1 0 1 i 2 1 1.0 100 40,000 20.0 1 300 0.1 Tc 3Se, 0
Tc 35o, E 3 3 9 0,3 7 6 6.0 2 3 0.4 2 120 0.02 12 36 0,6 3 1,440 0.4 1 3 0.1 12 11,160 4 :
12 3.0 15 45 0.8 13 39 26 31,200 15.6 1 900 0.4 Tc 35o, E
Tc 50, IV 0 loz.l 0.03 34 34 1.0 15 15 0.5 4 4,0 1 1 1.0 11 4 0.8 4 2 0.3 I 20 206 206 3 4 41 6,560 1.6 1 1 597 185,070 68.5 438 438 12 559 223,600 111.8 101 31,200 13.6 Tc 50, IV
*
Tc 254, O 0 5 5 0.2 7 240 3 2 0.3 1 l 226 36.160 9.0 58 58 1.0 173 38,130 14 l
1 1 1.0 1 1 S 1 1.0 93 37,200 18.6 Tc 254, 0
Tc 254, C' 0 51 6,120 0.8 53 8.480 2.1 10 3,100 1.1 15 6,000 3.0 Tc 254, C'
Tc 50, V 3 loz.3 0.02 7 6 0.2 14 4? 1.4 1 3 3.0 9 9 1.8 7 15 1.9 2 60 0.01 167 501 8.4 1 480 0.1 1 3 190 176.700 74 222 3.7 23 69 1.2 12 2 1,800 0.8 Tc 50. V
Tc 272, A 7 124 248 4.2 16 0.3 89 55.180 20.5 4 8 0.1 1 2 8 16 35 21,000 9.1 Tc 272, A
4 54 216 3.6 4 16 31 14.2 1 4 0.1 7 28 27 32,400 14.1 Tc 272, 0
Tc 272, C 4 1 4 4,0 61 244 4.1 1 4 80 18 7? 1.2 24 96 12 19,200 9.6 13 15,600 6.8 Tc 272, C
Tc 272,0 5 6 30 1.0 2 1,200 0.2 3 5 0.6 249 1,245 20/ 8 40 0.7 76 40,300 14.9 33 165 2.8 68 340 5.7 254 1,270 7 14,000 7.0 5 7,500 3.3 Tc 272, D
c?'?,t s 1 5 5.0 1 600 0.1 4 15 1.9 923 4.615 7b. 4 20 0.3 3(1 ISO 75 68 340 5 / 517 2,560 14 28,000 14.0 26 39,000 17.0 Tc 272, E
I
7 7 0.23 >44 544 16.0 85 85 7.8 3 3 3,0 3 1 1.0 291 13 2.6 13 5 0.6 1 20 79? 292 S.O 1,562 249,920 62.5 161 49,910 18 5 485 485 0.09 17/ 17/ 79 46 46 08 19 19 291 116,400 58.2 Tc 50, VI
3 5 0.6 48 240 4.0 13 20,150 7.5 1 5 5.0 92 460 l.l 3/ 185 3.1 1/ 85 Tc 272, F
SANTA MARIA SANTA MARIA
Tc 50. VII 2 5 10 0.33 1 2 0.1 35 70 2.3 7 14 14.0 1 2 2.0 35 8,400 1.0 13 24 4.8 7 8 1.0 979 1,958 32.6 282 90,240 22.6 1 2 12.4 07 114 Tc 50, VII
Tc 272, G 5 100 3,3 1 20 20.0 4 20 2.5 26 570 8.7 2 40 0.7 20 124,000 45.9 7 140 2.3 2 40 Tc 272, G
30 6.0 1 30 3.8 120 7.0 13 390 6.5 4 37,200 13.8 30 0.5 Tc 272, H
3 150 2.5 7 56,000 14.0 11 550 9.2 2 31.000 11.5 Tc 272,
4 38 18,240 2.3 86 9,920 3.7 1 4 4 0

50 1 6,000 0.8 10 80,000 20.0 6 300 5.0
ADEJAS
ABEJAS
Tc 50, VIII 5 4 70 0.66 1 5 0.2 4 20 0.6 5 5.0 2 5 5.0 7 4,200 0.5 5 5.0 6 30 6.0 9 40 5.0 100 0.02 140
1 1 28 1.2 0.7 118 590 0.12 Tc 50, VIII
2 5 5.0 13 7,800 1.0 1 5 5.0 50 160 32.0 1 5 0.6 3,100 1.2 Tc 50, IX
5 32 19,200 2.4 9 45 0.8 Tc 254.0
Tc 50, X 30 1 30 0.9 2 30 30.0 1 3,600 0.5 5 150 30.0 4,800 Tc 50, X
1 30 0.5 1 1.2
COXCATLAN
Tc 50, XI 8 3 74 0.80 32 256 8.5 2 8 8.0 3 2,880 0.4 2 8 1.6 4 8 1.0 160 600 4,800 Rh.O Tc 50, XI
1 0.03
Tc 254. E-F 12
9’ Tc 254, E-F
108 1.8
Tc 50, XII 17 7 12 0.40 1 12 0.4 1 12 0.4 1? 12 12.0 Tc 50, XII
Tc 50, XIII 12 1 17 040 12 17.0 1 12 12.0 1,440 0.2 Tc
1
50, XIII
EL RIEGO EL RIEGO
Tc 50, XIV 20 3 40 1.33 . 2 40 1.3 2 20 20.0 Tc 50. XIV
Tc 50, XV 40 3 170 4.40 5 200 6.6 Tc 50, XV
Tc 50, XVI 50 1 50 1.66 3 150 5.0 Tc XVI
50,
Nolo: Total volumes ol food Worn all souicos aio listed, by phase and zone, in Table 36.
•Loss than 0.05 litor.
Fig. 188. The relative importance of various subsistence activities.
its meat from hunted animals. Two percent lance horizon, only 12 percent of the meat was derived from
points suggest that nearly one percent of the meat large game. The projectile points were 58 percent
came from ambush hunting, 38 percent heavy dart arrow points, accounting for 7 percent of the meat; 11
came from stalking,
points suggests that 10 percent percent thin dart points (one percent of the meat), and
and 60 percent thin dart points suggests that 16 percent 31 percent heavy dart points (4 percent of the meat).
came from bleeding and trailing. In the earliest phase of the Tehuacan sequence, the
In the Ajalpan phase about 22 percent of the meat Ajuereado culture, which in terms of trends probably
came from big game. The projectile points were 5 per- derived about 60 percent of its subsistence from hunt-
cent lance points (one pereent ambush hunting), 37 ing, the lance points outnumber the barbed dart points
percent heavy dart points (8 percent stalking), and 58 by two to one (42 percent of the big game from am-
percent thin dart points (13 percent bleeding and trail- bushing and 18 percent from dart stalking). The large
ing). The Santa Maria phase, with 20 percent of the quantities of rabbit bones in Zones XXV and XXVI
meat from hunting, had 5 percent lance points, 49 per- seem to Flannery (Chapter be the probable re-
8) to
cent heavy dart points, and 46 percent thin dart points sult of rabbit drives. Perhaps 6 percent of the food
— or one percent of the meat from ambush hunting, 10 came from this activity and about 4 percent from trap-
percent from stalking, and 9 percent from bleeding and ping and collecting.
trailing. In the Palo Blanco phase hunting accounted A perusal of the wild plants we uncovered reveals
for 13 percent of the meat. Seventy percent of the at least four kinds of plant gathering that have slightly
points were large and barbed; 20 percent were thin, different periods of popularity. They are collecting
slashing dart points;and 10 percent were arrowpoints. seeds, picking fruits and pods, cutting leaves or stems,
That is, 9 percent of the meat came from stalking, 3 and digging roots. The tools for these activities were an
percent from bleeding and trailing, and one percent instrument with a sharp edge for cutting and a con-
from bow-and-arrow hunting. In the Venta Salada tainer for the stuffs collected —perhaps a basket for
305
seeds; a net bag for fruits, leaves, and roots; and a of the bulk food stuffs. The complexity of the first dam
carrying loop for leaves. In the El Riego phase the built near Purron Cave early in the following phase
major part of the plant food came from collected wild suggests that perhaps in the Ajalpan phase some irri-
seeds. Fruits represented only half the quantity of gation and orchard agriculture was undertaken. The
seeds, and cactus stems and agave leaves were even appearance of certain fruits needing supplementary
less popular. In the Coxcatlan phase, the wild seeds water as early as the Coxcatlan phase lends support to
decreased in popularity while fruits, leaves, and stems this theory.
increased and a few roots occur for the first time. These From Santa Maria times on it becomes difficult to
trends continue without much modification in the tellwhat produce came from dry farming and what
Abejas phase. However, with an increasing portion of came from wet farming. I tend to think of the squashes
the food coming from cultivated plants, subsistence from Coxcatlan Cave as being products of barranca
based on collecting wild plants begin to
activities horticulture (3 percent), the fruits as products of hydro-
diminish and diminish gradually throughout the re- horticulture (3 percent) and the com as being from
mainder of the sequence. These subsistence trends are barranca agriculture (23 percent). However, I regard
shown graphically in Fig. 188. corn from floors of Purron Cave that are contem-
Some of the activities connected with plant domesti- poraneous with the dam as being from irrigation agri-
cation are relatively short lived and rise and fall within culture (9 percent) and the fruits from orchard
our sequence, but others steadily increase. The earliest agriculture (7 percent). Obviously this is a somewhat
agriculture seems to have been practiced in the El arbitrary division, and attempting to separate products
Riego phase and even it seems to be of two kinds. One of the two kinds of activities in the last two phases is
kind, represented by cultivated squash, is what I have even more arbitrary. Certainly in both the Palo Blanco
called “barranca horticulture.” By this term I mean and Venta Salada phases products of barranca horti-
the planting of individual hardy cultivars in the bar- culture and hvdro-horticulture account for no more
rancas near the cave Whether the plants grew in
sites. than one percent of the total food —perhaps becoming,
actual gardens or not open to question. The other
is as they are today in the Tehuacan Valley, the fruit of
kind of early planting has been called “hydro-horti- the odd avocado or sapote tree in the well-watered
culture,” and means that individual domesticates patio or of a few squash or chili plants growing near
—
avocado trees and chili plants were planted beside the chicken coop, or in the Palo Blanco and Venta
springs or along the flanks of the Rio Salado where they Salada periods, the turkey coop. Most of the fruits must
received a steady year-round supply of water. These have come from well-watered orchards, which perhaps
two types of cultivation each accounted for about 2.5 produced about 7 percent of the diet in the Palo Blanco
percent of the total diet. period and only one percent in the final period. In-
In the Coxcatlan phase increased amounts of food creasing amounts of com and increasing numbers of
were derived from both barranca horticulture (5 per- irrigation features in both phases make me suspect
cent) and hydro-horticulture (4 percent). There is also that irrigation agriculture was the major subsistence
the first evidence {less than one percent) of corn planted activity, at least as far as food production is concerned.
in barranca agriculture. By “barranca agriculture” I However, some sites and caves were located along
mean the planting of such grains as corn and amaranth barrancas, so barranca agriculture must have persisted,
in fields (albeit in this phase very small), in the arroyo as it still does today. In Fig. 188 I have arbitrarily
bottoms, or on low terraces next to arroyos or bar- shown it to represent 10 percent of the food produced
rancas where they would receive a supply of moisture in Palo Blanco and Venta Salada times.
from runoff during the rainy season. In the Abejas Today slash-and-bum agriculture is practiced spar-
phase increased amounts of cultivars were produced ingly along mountain flanks of the valley. When this
from these three activities: barranca horticulture, 9 method began prehistorically is difficult to tell. The
percent; hydro-horticulture, 14 percent; and barranca only hints we have come from the site survey, which
agriculture, 1 percent (representing a few improved shows a few hillside sites attributable to the Venta
corn races). Salada phase and even fewer dating back to Palo
A radical shift in this pattern occurs in the Ajalpan Blanco times. I would guess that slash-and-burn farm-
phase, for produce from both barranca horticulture (3 ing contributed no more than 3 percent of the valley’s
percent) and hydro-horticulture (2 percent) decreases, diet in the Venta Salada phase and even less (one per-
while the proportion of improved com derived from cent) in the Palo Blanco phase. This would mean that
barranca or flood-plain agriculture rises to 35 percent about 40 percent of the food produced in the Palo
306
Blanco period and 60 percent in the Venta Salada pe- of southwest Tamaulipas, a dry mesquite desert in
riod were derived from irrigation. The increased num- deep canyons —does reveal somesimilarities. Both cul-
bers of irrigation features attributed to the final phase tures relied chiefly on ameat diet, with most of the meat
are consistent with these estimates. obtained from hunting and a small proportion from
In summary, our study of subsistence activities trapping or collecting, but even here there is a measure
shows a steady increase in food production owing to of difference. The El Riego diet consisted of over 50
more means of production. In other words,
efficient percent meat, and the Infiernillo diet of only 40 per-
people got more and more food by expending decreas- cent meat. The greater amounts of wild plant finds in
ing amounts of energy. Not only that, but the food pro- Infiernillo — about 59 percent as against 45 percent for
duced yielded larger quantities of protein per unit of El Riego — also may be The El Riego com-
significant.
measure, or more and more food energy. In terms of plex seemed about 5 percent of its food from
to derive
our present analytical techniques we have not been agricultural produce, whereas Infiernillo similarly de-
able to quantify the energy going into subsistence ac- rived about one percent. Even here the domesticated
tivity or the energy derived from the food
utilized, but plants were different, there being a few seeds of
there doubt that as time spent in food-getting
is little pumpkins and gourds in Infiernillo, whereas El Riego
activities decreased, the food energy increased. had avocado, cultivated squash, and probably chili and
Throughout our sequence more time and more energy amaranth.
became available for other activities. When one compares the El Riego and Infiernillo
materials with the reconstruction of the subsistence

For the most part other archaeological sequences pattern based only on bones and artifacts from Santa
from Mesoamerica are not so complete as the one from Marta Cave, located in a basin surrounded by tropical
Tehuacan, nor do they have sufficient recovered food become ap-
rain forest in Chiapas, further differences
stuffs to reconstruct the changes in the subsistence. parent.The Santa Marta people derived as much as 70
Nevertheless, when one looks at them from a not too percent of their diet from wild plants and only about
critical standpoint some general trends are evident. 30 percent from animals, and they no evidence of
left
Certainly there was a trend toward increased and im- agriculture. The Marta crude
similarity of the Santa
proved food by more effective means of production. artifacts to those Jose Luis Lorenzo uncovered in the
This does not mean, however, that all cultures evolv- lower levels of shell middens at Islona de Chantuto, in
ing in the varied ecological zones of the Mesoamerican swampy lowlands on the Pacific coast of Chiapas, sug-
culture area went through the same specific subsistence two complexes were perhaps contempo-
gests that the
developments or even the same stages of economic evo- raneous and that on the Chiapas coast there was still
lution revealed in the Tehuacan Valley. In other words, another tvpe of subsistence — shell-food collecting.
the specific aspects of the sustenance, food prepara- These hints of different types of subsistence in dif-
tion, and subsistence activities that changed through ferent ecological zones of our poorly documented first
time in the unique ecological zone that includes the period become clearer as evidence for the reconstruc-
Tehuacan Valley cannot be considered typical of all tion of subsistence patterns increases for the next
Mesoamerica, nor can one generalize that all other period, 5000 to 2300 b.c. In Tehuacan use of wild plants
ecological zones ofMesoamerica underwent the same decreases from the Coxcatlan to the Abejas phase (from
kind of evolution in subsistence pattern as did the 52 to 49 percent), as does use of meat (34 to 30 percent),
Tehuacan Valley. contrasting with an increase in agricultural produce in
A brief comparison of the Tehuacan subsistence se- the two phases from 14 to 21 percent. Flannery’s recent
quence with rather tentative reconstructions of sub- excavations of similar archaeological phases in caves
sistence patterns from other regions in terms of general near Mitla, Oaxaca, in a somewhat more mesic high-
time periods brings out this point rather clearly. Un- land environment, seems to reveal a similar subsistence
fortunately, information on the subsistence of the pattern, but perhaps with a slightly smaller proportion
earliest period, from before 7000 b.c., is too sparse to of the diet from meat and more from plant collecting.
be enlightening, so we shall begin with the period Both sequences representative of this period in
from 7000 to 5000 b.c. Tamaulipas, one in the Sierra de Tamaulipas and the
A comparison of the two Mesoamerican archae- other in Infiernillo Canyon, show similar decreases in
ological complexes of this period with the most sub- meat and wild plant remains with concomitant in-
sistence information —
El Riego of Tehuacan, a high- creases in agricultural produce. Again, however,
land desert valley, and Infiernillo of the Ocampo region amounts of wild plant remains are greater than those in
307
Tehuacan and agricultural remains do not become so culture and from fruit trees planted in watered or-
abundant. It should be noted that whereas by the end chards. Recent studies of materials from dry caves in
of this period most of the agricultural produce of all the Mitla area indicate that a similar situation pre-
these sequences is composed of corn, beans, squash, vailed in that region, although the growing of beans
chili,and amaranth, these plants were domesticated in seems to have been far was in
more important than it
different regions and arrived in the respective se- Tehuacan (Flannery, personal communication). Vague
quences at different times. hints from the Valley of Mexico, with still another
Nevertheless, there are some similarities in the kinds environment, suggest similar means of production to
of plant remains for these varied areas which contrast those in the Puebla and Oaxaca areas. In Tamaulipas,
with inadequate estimates that can be made for coastal where the mesquite desert environment is very differ-
cultural phases of 5000 to 2300 b.c. The pre-ceramic ent from southern Puebla and Oaxaca, agricultural
levels found by Charles Brush at Puerto Marquez, food seems to have been produced almost entirely
Guerrero, had many shellfish and fish and animal by dry fanning, along with some barranca agriculture
bones, and no mortars, manos, or metates. The late and gardening. The subsistence pattern from the Vera-
preceramic phase of Islona de Chantuto in coastal cruz tropical lowlands is difficult to reconstruct, but
Chiapas and the huge preceramic sites in the fossil there are hints that dry fanning, flood-plain agricul-
sand dunes near Viejon in central Veracruz seem to ture, and perhaps fruit growing were supplementing
have about the same kinds of materials. The bones and plant collection and the rich harvest from the sea. The
shells suggest that the people of these cultures were Pacific coastal area of Guatemala, with a similar en-
predominately hunters, fishermen, and gatherers of vironment, may have had a similar subsistence pattern.
shell-food, and perhaps did some plant collecting. The People of the central Chiapas basin may have evolved
lack of manos and metates indicates that they did not a similar agricultural pattern, but deriving less from
grind com and hints that they did not practice com food collection than did the coastal regions. The Maya
agriculture. lowland cultures of this period are usually thought of
This suggestion that some of the lowland cultures as having an agriculture based on slash-and-burn farm-
of the period before 2300 b.c. did not have corn agri- ing, and there are hints that highland Guatemala (and
culture is bolstered by the absence of corn in the pollen perhaps the mountain-ridge areas of central Mexico)
profile of Santa Marta Cave, with a slightly different may have used slash-and-burn techniques in a different
environment. It must be added, however, that the late way. Thus it seems possible that the varieties of food
Santa Marta remains reveal a still different subsistence production used in the previous period became greater
pattern, one of plant collecting and animal hunting and from 2300 to 200 b.c.
collecting. Thus again the period from 5000 to 2300 By the final period, from 200 b.c. to the Spanish con-
b.c. shows varying subsistence patterns in different quest, not only had the corn-squash-bean-chili-ama-
ecological zones. ranth complex spread throughout Mesoameriea, but so
In the next period, from 2300 to 200 b.c.,we have had many other Mesoamerican domesticates and even
large amounts of archaeological remains from many a few of South American origin. In Tehuacan, the
ecological zones of Mesoameriea. Unfortunately, pre- majority of the food was produced by irrigation, but
served food stuffs are absent from most of these ar- within the valley, often in micro-environments, a vari-
chaeological complexes, so subsistence estimates are for ety of other agricultural techniques were employed.
the most part verv speculative. Nevertheless, it seems In the Mitla area irrigation was also used, but because
safe to assume that agriculture based on corn, beans, of a slightly differentand wetter environment other
squash, chili peppers, and amaranth was practiced techniques, such as dry-farming and flood-plain farm-
throughout Mesoameriea, although some of these ing, were of considerable importance. Although the
plants were grown in different proportions in different subsistence data are very inadequate, a similar situa-
ecological zones. Other domesticated plants may have tion may have existed in the Valley of Mexico, but
varied from one area to another. here the chinampa tvpe of agriculture played an im-
The biggest difference, however, seems to have been portant part.
in the means of production of the agriculture. In Te- In Morelos, there was more emphasis on irrigation
huacan some of the agricultural food resulted from agriculture, particularly in the area near Xochicalco,
dry farming and the planting of garden plots and ter- but there may have been considerable variation in
raced fields in the barrancas, but increasing amounts agricultural practices in terms of adaptation to micro-
of food through the period came from irrigation agri- environments in this region also. In better-watered
308
areas, such as mountain tops or flanks throughout the One gets the impression that the development of
altiplano, irrigation may have been relied on less and civilization and more effective food production in
other agricultural techniques more than in the broader Mesoameriea is not due to a single evolution of develop-
and drier valleys. Nevertheless, in the entire altiplano, mental stages of culture and subsistence, but that a series
irrigation seems to have been used to some extent, in of concomitant developments of rather different ecologi-
contrast to the lowland regions, where irrigation was cal zones are interacting with and interstimulating one
rarely employed. In central Veracruz, near Cempoala, another in such a manner as to bring about cultural de-
rainfall was supplemented by the use of irrigation. In velopment and increasingly effective food production.
Postclassic highland Guatemala one finds terraced In fact, is not this sort of symbiotic development of
hillsides. In Maya lowlands and coastal regions slash- agriculture and culture one of the causative processes
and-bum or flood-plain agriculture seem for the most and civilization in
that leads to effective food production
part to have been used, but I would suspect that a Mesoameriea? Or to put it another way, was not the de-
careful study might reveal considerable variation of velopment of effective food production and the con-
agricultural practices according to environmental dif- comitant cultural development in Mesoameriea in no
ferences within the lowland zone. In the dry northland small part owing to the fact that there were contiguous
of Tamaulipas, dry-farming seems to have continued in environments or ecological zones that were exploited
use; near the end of this period, perhaps due to agriculturally in different ways, and evolved through
“drought” conditions, was not very successful. Dry
it different cultural stages, and their geographical close-
northwest Mexico may have met this changing envi- ness allowed the varied subsistence developments to
ronmental situation by a gradual shift from dry-farming interact and interstimulate one another? Further, is not
to increased reliance on irrigation agriculture. this symbiotic process which developed effective food
From our superficial survey of Mesoamerican ar- production, and civilization, in Mesoameriea the same
chaeological sequences the impression emerges of sort of process that must have occurred in other areas of
more effective food production of an increasing variety primary civilizations, such as the Near East, Peru, or
of plants, accomplished by a wide variety of agricul- even China? If so, can we make universal or unilinear
tural practices that may represent local adaptive re- generalizations about civilization, other than the broad
sponse to different ecological zones. It appears that all kind of statement that food-gathering preceded food-
zones of Mesoameriea did not undergo exactly the production, or that cultures developed from simple to
same subsistence development as Tehuacan. In fact, it complex? In fact, must not our generalizations about the
seems almost impossible to classify the subsistence development of agriculture and the rise of civilization
data of all of Mesoameriea into one set of neat unilinear be of the variety that Stewart has called “multilinear
stages. One can, of course, classify the data into two evolutions”? Is it not time, even with our meager data, to
very broad stages such as food-gathering and food- think in these terms? We hope that this volume with its
production, but finer divisions for
all of Mesoameriea ecological data and estimates of the changing subsis-
on the basis of the present meager subsistence infor- tence pattern over nearly ten thousand years in the
mation just cannot be done. Tehuacan Valley is a step in this direction.
REFERENCES
Mangelsdorf, MacNeish, and Willey, 1964.

MacNeish, 1964a.
309
ABBREVIATIONS USED IN THE BIBLIOGRAPHY
AIB Anales del Iirstituto de Biologia, Universidad

Nacional Autonoma, Mexico.
Am. Ant. American Antiquity, Menasha and Salt Lake
Citv.
AMNH-AP American Museum of Natural History, An-
thropological Papers, New York.
APS-P American Philosophical Society, Proceed-
ings, Philadelphia.
APS-T American Philosophical Society, Transac-
tions, Philadelphia.
BAE-B Bureau of American Ethnology, Bulletin,
Washington, D. C.
BML Botanical Museum Leaflets, Harvard Univer-
sity, Cambridge, Mass.
CASS- California Avocado Society Yearbook.

IA Ibero-Americana, University of California
Press, Berkeley and Los Angeles.
IN AH Instituto Nacional de Antropologia e His-
toria, Mexico.
HMAI Handbook of Middle American Indians, ed.
by Robert Wauchope, University of Texas
Press, Austin.
MBG-A Annals of the Missouri Botanical Garden,
St. Louis.
Mex. Ant. El Mexico Antiguo, Sociedad Alemana Mexi-
canista, Mexico.
VFPA Viking Fund Publications in Anthropology,
New York.
BIBLIOGRAPHY
Aguilera, J. G. ology, University of California Archaeological

1906 De Mexico a Tehuacan and Excursion de Tehua- Survey Report, no. 42:23-40.
can a San Juan Raya. “Guide des Excursiones, Ashley Montague, M. F.
Congr. Geol. Int., Xeme Ses.” Nos. 4 and 7. 1960 An Introduction
Physical Anthropology
to
Mexico. Thomas.
Springfield: Charles C.
Albanese, A. A. Atlas Climatologico e Hidrologico de la Cuenca
1959 Protein and Amino Acid Nutrition. New York: del Papaloapan
Academic Press. 1956 Comision del Papaloapan. Secretaria de Recursos
Alcedo, D. Antonio de Hidraulicos, Estados Unidos Mexicanos. Mexico.
1786 Diccionario Geografco-Historico de las lndias Oc- Aveleyra Arroyo de Anda, Luis
cident ales O America. Madrid: Benito Cano. 1955 El Segundo Mamut Fosil de Santa Isabel Iztapan
Alexander, H. L., Jr. y Artefactos Asociados. 1NAH, 1.
1963 “The Levi Site: A Paleo-Indian Campsite in Cen- Bailey, L. H.

tral Texas.” Am. Ant., 28:510-28. 1943 “Species of Cucurbita.” Gentes Herbarium,
Altschul, A. M. 6:266-322. Ithaca.
1962 “Seed Proteins and World Food Problems.” Eco- 1948 “Jottings in the Cucurbitas.” Ibid., 7:448—77.
nomic Botany, 16 (1):2-13. Baines, E.
Alvarez del Toro, Miguel 1835 History of the Cotton Manufacture in Great Bri-
1952 Los Animales Silvestres de Chiapas. Tuxtla Gu- tain.London.
tierrez: Ediciones del Gobierno del Estado. Barchoorn, E. S., M. K. Wolfe, and K. H. Clisby
Anderson, Edgar 1954 “Fossil Maize from the Valiev of Mexico.” BAIL,
1950 “Variation in Avocados at the Rodiles Plantation.” 16:229-40.
Ceiba, 1:50-55. Barrera, Tomas
and R. O. Erickson 1946 Guia Geologica de Oaxaca. Institute de Geologia,
1941 “Antithetical Dominance in North American Universidad Nacional Autonoma de Mexico,
Maize.” Proceedings of the National Academy of Mexico.
Science, 27:436—40. Bernal, Ignacio
Anderson, J. E. 1965a “Archaeological Svnthesis of Oaxaca.” HMAI, vol.
'
1963 The People of Fairty. Ottawa: National Museum 3, pt. 2:788-814

of Canada. 1965b “Brasero Ceremonial.” IN AH Boletin 22, Diciem-
Antevs, Ernst bre 1965:1-2.
1955 “Geologic-climatic Dating in the West.” Am. Ant., Beyer, Hermann
20:317-35. 1965 “Mito y Simbolismo del Mexico Antiguo.” Mex.
1962 “Late Quaternary Climates in Arizona.” Ibid., Ant., vol. 10.
28:193-98. Bird, J.
B.
Applin, E. R. 1943 Excavations in Northern Chile. AMNH-AP, vol.
1955 “Report of the Studv of Some Microfossils Col- 38, no. 4:171-316.
lected by A. Calderon.” Unpublished report, U. S. 1948 “America’s Oldest Farmers.” Natural History, 57:
Geological Survey. 296-303, 334-35.
Arber, A. and J.
Mahler
1934 The Gramineae. Cambridge University Press. 1951-52 “America’s Oldest Cotton Fabrics.” American Fab-
Aschmann, H. H. rics, 73.
1958 “Great Basin Climates in Relation to Human Oc- Blake, S. F.
cupance.” Current Views on Great Basin Archae- 1920 “A Preliminary Revision of the North American
311
and West Indian Avocados ( Persea spp. ).” Jour- Cardenas, M.

nal of the Washington Academy of Science, 10:9- 1948 “Plantas Alimenticias Nativas de los Andes de
21 . Bolivia.” Eol. Universit., 2:36-51.
Blasquez, L. L. Carlson, R. L.
1957 “Hidrogeologia de una Parte del Valle de Tehua- 1963 “Basketmaker III Sites near Durango, Colorado.”
can, Estado de Puebla.” Anales del Instituto de The Earl Morris Papers, no. 1. University of Colo-
Geologic, 12:9-55. Mexico. rado Studies, Series in Anthropology, no. 8:1-82.

Block, R. J., and K. W. Weiss Carter, G. F.
1956 Amino Acid Handbook. Springfield: Charles C. 1945a Plant Geography and Culture History in the
Thomas. American Southwest. VFPA, no. 5.
Boletin Hidrologico or Hidrometrico 1945b “Some Archaeologic Cucurbit Seed from Peru.”
1955-60 Comision del Papaloapan. Secretaria de Recursos Acta Amer., 3:163-72.
1962 Hidraulieos, Estados Unidos Mexicanos. Mexico. 1951 “An Early American Description Probably Refer-
Borah, Woodrow, and S. F. Cook ring to Phaseolus lunatus.” Biologia II, Chronica
1960 The Population of Central Mexico in 1548: An Botanica, 12, 4/6.
Analysis of the “Suma de Visitas de Pueblos.” IA, Castetter, E. F., W. H. Bell, and A. R. Grove
no. 43. 1938 “The Earlv Utilization and the Distribution of
1963 The Aboriginal Population of Central Mexico on Agave in the American Southwest.” University of
theEve of the Spanish Conquest. IA, no. 45. New Mexico Bulletin, Biological Series, 5:1-92.
Brooks, C. E. P.
Caso, Alfonso
1949 Climate through the Ages. 2nd ed. London: Ern-
1927 “Las ruinas de Tizatlan, Tlaxcala.” Revista Mexi-
est Benn.
cana de Estudios Historicos, 1(4) 139—72.
Brooks, R. H., et al.
1938 Exploraciones en Oaxaca, Quinta y Sexta Tem-
1962 “Plant Material from a Cave on the Rio Zape, poradas, 1936-37. Instituto Panamericano de Ge-
Durango, Mexico.” Am. Ant., 27:356-69.
ografia e Historia, Pub. 34.
Bryan, Alan L., and Ruth Gruhn
Chatfield, Charlotte, and Georgian Adams
1964 “Problems Relating to the Neothermal Climatic
1940 Proximate Compositions of American Food Mate-
Sequence.” Am. Ant., 29:307-15.
rials. U.S.D.A. Circular, no. 549.
Bukasov, S. M.
Codex Bodley
1930 “The Cultivated Plants of Mexico, Guatemala,
1960 Facsimile edition of a Mexican painting preserved
and Colombia.” Bulletin of Applied Botany, Ge-
in the collection of Sir Thomas Bodley, Bodleian
netics, and Plant Breeding, Supplement 47:151-
bv Alfonso Caso,
Library, Oxford. Interpretation
76. Leningrad.
bv Ruth Morales and John Paddock. Socie-
trans.
Byers, H. R.
dad Mexicana de Antropologia, Mexico.
1959 General Meteorology. New York: McGraw Hill.
Calderon Garcia, A.
Codex Borgia
1904- Eine altmexikanische Bilderschrift der Bibliothek
1956 “Bosquejo Geologico de la Region de San Juan
1909 de Congregatio de Propaganda Fide Herausgege-
Rava, Estado de Puebla” and “Estratigrafia del
ben auf Kosten Seiner Excellenz des Herzogs von
Mezozoico v Tectonica del Sur del Estado de
Loubat, erlautert von Eduard Seler. 3 vols. Berlin.
Puebla; presa de Valsequillo.” Excursion A—11:
Congreso Geologico Internacional, XXa Ses.
1963 Codice Borgia. Facsimile edition of a pre-Colum-
bian codex preserved in the Ethnographical Mu-
Callen, E. O.
1963 “Diet as Revealed bv Coprolites.” Science in Ar-
seum of the Vatican, Rome; With commentary by
chaeology, 19:186-94. London: Thames and Hud- Eduard Seler. 3 vols. Fondo de Cultura Eco-
nomica, Mexico.
son.
1965 “Food Habits of some pre-Columbian Indians.” Codex Cospiano (Bologna)
Economic Botany, 19 (4):335—43. 1898 A pre-Columbian codex in the Library of the Uni-
and T. W. M. Cameron versity of Bologna. Published bv the Due de
1955 “The Diet and Parasites of Prehistoric Huaca Loubat, Rome.
Prieta Indians as Determined bv Dried Copro- Codex Fejfrvary-Mayer
lites.” Proceedings of the Boyal Society of Canada, 1901- An old Mexican picture manuscript in the Liver-
1955:51. 1902 pool Free Public Museum. Comments by Eduard

1960 “A Prehistoric Diet Revealed in Coprolites.” The Seler, trans. by A. H. Keane. Berlin and London.
New Scientist, 8 (190) :35-40. Codex Laud
“The First New World Cereal.” Am. Ant. In press. 1937 Manuscripto Mexicano que existe en la Biblioteca
312
BIBLIOGRAPHY
Bodleiana, Universidad de Oxford. Mexico: G. M. Cook de Leonard, Carmen

Echaniz. 1953 “Los Popolocas de Puebla,” in “Huastecos, Toto-
Codex Mendoza nacQs y sus Vecinos.” Revista Mexicana de Es-
1938 Codex Mendoza. Ed. and trans. by James Cooper tudios Antropologicos, 13:423-45. Mexico.
Clark. 3 vols. London: Waterlow. 1963 “Seasons for Plants at Tehuacan.” Unpublished
Codex Nuttall MS.
1902 Facsimile edition of an ancient Mexican codex , ED.
belonging to Lord Zouche of Harynworth, Eng- 1959 Esplendor del Mexico Antiguo. 2 vols. Mexico.
land, with an introduction by Zelia Nuttall. Cam- Cooper Clark, James, ed. and trans.
bridge, Mass.: Peabodv Museum. 1938 Codex Mendoza. 3 vols. London: Waterlow.
Codex Selden II (3135; A. 2) Crane, H. R., and J. Griffin
B.
1964 Interpretation of the Codex by Alfonso Caso, 1962 “University Michigan Radiocarbon Dates, VII.”
of
trans. by Jacinto Quirate and John Paddock, with Radiocarbon, American Journal of Science, 4:
a facsimile of the codex. Sociedad Mexicana de 183-203.
Antropologia, Mexico. Cutler, H. C.
Codex Vaticanus 3773 (Vaticanus B) 1960 “Cultivated Plant Remains from Waterfall Cave,
1902- A pre-Columbian codex preserved in the Library Chihuahua.” Am. Ant., 26:277-79.
1903 Rome. Commentary by Eduard
of the Vatican, and Martin Cardenas
Seler, trans. by A. H. Keane. Berlin and London. 1947 “Chicha, a Native South American Beer.” BML,
Codex Vindobonensis 13:33-60.
1929 Codex Vindobonensis Mexicanus I, facsimile edi- and T. W. Whitaker
tion, ed. by Walter Lehmann and Ottaker Smital, 1956 “Cucurbita mixta Pang.; Its Classification and Re-
Vienna. lationships.” Torrey Botanical Club Bulletin, 83:
Coe, M. D., and K. V. Flannery 253-60. New York.
1964 “Microenvironments and Mesoamerican Prehis- 1961 “History and Distribution of the Cultivated Cu-
tory.” Science, 143:650-54. curbits in the Americas.” Am. Ant., 26:469-85.
Coit, J.
E. Dewey, L. H.
1948 “Mexican Explorations of 1948.” CASY, 1948:59- 1943 Fiber Production in the Western Hemisphere.
62. U.S.D.A. Misc. Pub. No. 518.
Collier, Albert Ditmars, R. L.
1964 “The American Mediterranean.” HMAI, 1:122- 1940 The Reptiles of North America. New York.
42. Dobyns, H. F.
Collins, G. N. 1966 “Estimating Aboriginal American Population: An
1914 “A Drought-resisting Adaptation in Seedlings of Appraisal of Techniques for a New Hemispheric
Hopi Maize.” Journal of Agricultural Research, Estimate.” Current Anthropology, 7:395-416.
1:293-302. Drucker, Philip
Comas, Juan 1943 Ceramic Sequences at Tres Zapotes, Veracruz,
1958
1960 Manual of Physical Anthropology. Springfield: Mexico and Ceramic Stratigraphy at Cerro de las
Charles C. Thomas. Mesas, Veracruz, Mexico. BAE-B, nos. 140 and
Cook, O. F. 141.
1925 “Peru Center of Domestication.” Journal of
as a Ekholm, G. F.
Heredity, 16:32-46, 93-110. 1944 Excavations at Tampico and Panuco in the Huas-
Cook, S. F. teca, Mexico. AMNH-AP, vol. 38, pt. 2.
1936-39 “Dwelling Construction in the Mixteca.” Mex. Ellis, G. P.
Ant., 4:375-83. 1959 “The Maillard Reaction.” Advances in Carbohy-
1949 Soil Erosion and Population in Central Mexico. drate Chemistry, 14:63-134.
IA, no. 34. Emerson, R. A., G. W. Beadle, and A. C. Fraser
Santa Maria lxcatlan: Habitat, Population, Sub- 1935 “A Summary of Linkage Studies in Maize.” Cor-
sistence. IA, no. 41. nell University Agricultural Experiment Station
and Woodrow Borah Memoir, no. 180.
1960 The Indian Population of Central Mexico, 1531- Engel, F. E.
1610. I A, no. 44. 1963 A Preceramic Settlement on the Central Coast of
and L. B. Simpson Peru. APS-T,
n.s. 53 (3).
1948 The Population of Central Mexico in the Sixteenth Evans, Clifford, and B. J. Meggers
Century. IA, no. 31. 1960 “A New Dating Method Using Obsidian: Part II,
313
An Archaeological Evaluation of the Method.” chaeology. Ed. by R. F. Heizer and S. F. Cook.

Am. Ant., 25:523—37. VFPA, 28:93-157.
Fonner, R. L. Hess, S. L.
1957 Appendix B in J. D. Jennings, Danger Cave. Soci- 1959 Introduction to Theoretical Meteorology. New
ety for American Archaeology, Memoir no. 14. York: Henry Holt.
Menasha and Salt Lake City. Hester, J. J.
Freeman, G. F. 1960 “Late Pleistocene Extinction and Radiocarbon
1912 “Southwestern Beans and Teparies.” University Dating.” Am. Ant., 26:58-77
1960 of Arizona Agricultural Experimental Station, Bul-
Hill, A. F.
letin 68. 1952 Economic Botany. New York.
Friedman, Irving, and R. L. Smith Hitchcock, A. S.

“A New Dating Method Using Obsidian: Part I, 1913 “Mexican Grasses in the U. S. National Her-
The Development of the Method.” Am. Ant., 25: barium.” U. S. National Museum Contribution
U. S. National Herbarium, 17:181-390.
476-522.
Fries, C., C
.W. Hirbard, and D. H. Dunkle
jr.,
Howard, R. A.
1961 “The Correct Names for ‘Diospyros ebenaster ”
1955 “Early Cenezoic Vertebrates in the Red Conglom-
Journal of the Arnold Arboretum, 42:430—35.
erates at Guanajuato, Mexico.” Smithsonian Mis-
cellaneous Collections, vol. 123, no. 7:1-25.
and Tycho Norlindh
1962
1962 “The Tvpifieation of Diospyros ebenum and Di-
Gerstel, D. U.
ospyros ebenaster.” Ibid., 43:94-107.
1953a “Chromosomal Translocations in Inter-specific
Hubbs, Carl, and Gunner Roden
Hybrids of the Genus Gossypium.” Evolution,
1964 “Oceanography and Marine Life along the Pacific-
7:234-44.
Coast of Middle America.” HMAI, 1:143-86.
1953b “Genetic Segregation of Allopolyploids in the
Hutchinson, J. B.
Genus Gossypium.” Genetics, 38:664.
“The History and Relationships of the World’s
Genoves, Santiago
Cottons.” Endeavour, 21:5-15.
1960 “Reevaluation of Age, Stature and Sex of the
R. A. Silow,and S. G. Stephens
Tepexpan Remains, Mexico.” American Journal
1947 The Evolution of Gossypium. London: Oxford
of Physical Anthropology, 18:205.
University Press.
1962 Diferencias Sexuales en el Hueso Coxal. Univer-
Irwin, Henry, and E. S. Barghoorn
sidad Nacional Autonoma, Mexico.
1965 “Identification of the Pollen of Maize, Teosinte
1964 La Antropologia Fisica en Mexico. Universidad
and Tripsacum bv Phase Contrast Microscopy.”
Nacional Autonoma, Mexico.
BML, 21:37-57.
Griswold, H. B.
Jenks, A. E.
1946 “Primitive Avocados of Central America and Mex-
1937 Minnesota’s Browns Valley Man. American An-
ico.” CAST, 1946:103-105.
thropological Association, Memoir no. 49. Me-
Grobman, A., et al.
nasha and Salt Lake City.
1961 Races of Maize in Peru. Nat. Acad. Sci. Nat. Res.
AND L. A. WlLFORD
Council Pub., no. 915.
1938 “The Sauk Valley Skeleton.” Texas Archaeological
Gulati, A. N., and A. J. Turner and Paleontological Society Bulletin, 10:136-69.
1928 A Note on the Early Historij of Cotton. Ind. Cent. D.
Jennings, J.
Cot. Comm. Tech. Lab., Bulletin 17, Tech. Ser.
1953 “Danger Cave: Progress Summary.” El Palacio,
12 .
60:179-214.
Hall, E. R., and K. R. Kelson 1957 Danger Cave. Society for American Archaeology,
1959 The Mammals of North America. New York: Memoir no. 14. Menasha and Salt Lake City.
Ronald Press. Johnston, M. C.
Harshberger, J. W. 1962 “Revision of Condalia including Microrhamnus
1896 “The Purpose of Ethnobotanv.” American Anti- (Rhamnaceae).” Brittonia, 14:332-68.
quarian, 17 2 ) 73— 81
( Jones, D. B., C. E. F. Gersdorff, and Sammie Phillips
Heiser, C. B., jr., and P. G. Smith 1938 Journal of Biological Chemistry, 122:745-55.
1953 “The Cultivated Capsicum Peppers.” Economic Jones, V. H.
Botany, 7:214-27. 1936 “The Vegetal Remains of Newt Kash Hollow
1958 “New Species of Capsicum from South America.” Shelter.” University of Kentucky Reports in Ar-
Brittonia, 10:194-201. chaeology and Anthropology, 3 (4): 147-65.
Heizer, R. F. Kaplan, Lawrence
1960 “Physical Analysis of Habitation Residues,” in 1956 “The Cultivated Beans of the Prehistoric South-
The Application of Quantitative Methods in Ar- west.” MBG-A, 43:189-251.
314
BIBLIOGRAPHY
1960 “Historical and Ethnobotanical Aspects of Do- 1961b “Recent Finds Concerned with the Incipient Ag-
mestication in Tagetes.” Economic Botany, 14: riculture Stage in Prehistoric Mesoamerica.”
200- 202 . Homenaje a Pablo Martinez del Rio, 91-101.
AND R. S. MacNeish 1962 Second Annual Report of the Tehuacan Archaeo-
1960 “Prehistoric Bean Remains from Caves in the logical-Botanical Project. R. S.Peabodv Founda-
Ocampo Region of Tamaulipas, Mexico.” BML, tion for Archaeology. Andover, Mass.
19:33-56. 1964a “Food-gathering and Incipient Agriculture Stage
Kelly, Isabel, and Angel Palerm in Prehistoric Middle America." HMAI, 1:413—26.
1952 The Tajin Totonac. “Part I: History, Subsistence, 1964b “Ancient Mesoamerican Civilization.” Science,
Shelter and Technology. ” Smithsonian Institution, 143:531-37.
Institute of Social Anthropology, Pub. no. 13. AND A. NELKEN
Washington, D.C.
1961 “Le Mexique et les Debuts de l’Agriculture au
Kidder, A. V., and S. J. Guernsey
Nouveau Monde.” L’Anthropol., 65:349—53.
1919 Archaeological Explorations in Northeastern Ari-
Maldonado-Koerdell, Manuel
zona. BAE-B, 65.
1964 “Geohistory and Paleogeography of Middle
D. Jennings, and E. M. Shook
J.
America.” HMAI, 1:3—32.
1946 Excavations at Kaminaljuyu, Guatemala. Carnegie
Mangelsdorf, P. C.
Institute of Washington, Pub. 561.
1958a “Reconstructing the Ancestor of Corn.” APS-P,
Kirchhoff, Paul
102:454-63.
1952 “Mesoamerica: Its Geographic Limits, Ethnic
1958b “The Mutegenic Effect of Hybridizing Maize and
Composition and Cultural Characteristics.” Heri-
Teosinte.” Cold Spring Harbor Symposium Quant.
tage of Conquest, ed. bv Sol Tax. Glencoe, Illinois.
Biol., 23:409-21.
Krickeberg, Walter
1961 “Introgression in Maize.” Euphytica, 10:157—68.
1956 Altmexikanische Kulturen. Safari-Verlag: Berlin.
and R. H. Lister
La Farge, Oliver ii, and D. S. Byers
1956 “Archaeological Evidence on the Diffusion and
1931 The Year Bearer’s People. Tulane University of
Evolution of Maize in Northwestern Mexico.”
Louisiana Middle American Research Series, Pub.
no. 3. New Orleans.
BML 17:151-78.,
R. S. MacNeish, and W. C. Galinat

Lehmann, Walter
1956 “Archaeological Evidence on the Diffusion and
1905 “Les Peintures Mixteco-zapoteques et quelques
Evolution of Maize in Northeastern Mexico.”
Documents Apparentes.” Journal de la Societe de
Americanistes de Paris, n.s., 2:241-80. Paris.
BML ,
17:125-50.
1964 “Domestication of Corn.” Science, 143:538-45.
Leopold, A. S.
R. S. MacNeish, and G. R. Willey
1959 Wild Life of Mexico: The Game Birds and Mam-
1964 “Origins of Agriculture in Middle America.”
mals. Berkeley: University of California Press.
Loud, L. L., and M. R. Harrington
HMAI, L427-44.
and R. G. Reeves
1929 Lovelock Cave. University of California Publica-
tions in American Archaeology and Ethnology,
1939 “The Origin of Indian Corn and Its Relatives."
vol. 25, no. 1. Berkeley.
Texas Agricultural Experiment Station Bulletin,
Lowe, R. T. 574.
1868 A Manual: Flora of Madeira. London. Marsh, D. C.

Mackie, W. W. 1965 “Some Insects and Vertebrates Recovered from
1943 “Origin, Dispersal and Variability of the Lima the Coprolites of Prehistoric Indians of South-
Bean, Phaseolus lunatus.” Hilgardia, 15:1-29. western Tamaulipas, Mexico.” Unpublished MS.,
MacNeish, R. S. McGill University.
1954 An Early Archaeological Site Near Panuco, Vera and E. O. Callen
Cruz. APS-T, n.s., vol. 44, pt. 5. 1964 “Pre-Columbian Insects from Tamaulipas.”
1958 Preliminary Archaeological Investigations in the Phytoprotection, 45 (3): 134.
Sierra de Tamaulipas, Mexico. APS-T, n.s., vol. Martin, Paul Schultz
48, pt. 6. 1963 “Earlv Man in Arizona: The Pollen Evidence.”
1959 “Origin and Spread of Some Domesticated Plants Am. Anf., 29:67-73.
as Seen from Tamaulipas, Mexico.” Paper pre- 1964 “Pollen Analysis and the Full Glacial Landscape.’
sented to annual meeting, American Anthropolo- The Reconstruction of Past Environments, as-
gical Association. sembled bv J. J. Hester and James Schoenwetter.
1961a First Annual Report of the Tehuacan Archaeo- Fort Burgwin Research Center, no. 3.
logical-Botanical Project. R. S. Peabodv Founda- James Schoenwetter, and B. C. Arms
tion for Archaeology. Andover, Mass. 1963 The Last 10,000 Years: A Fossil Pollen Record of
315
the American Southwest. Tucson: The University Oschinsky, Laurence

of Arizona Press. 1964 The Most Ancient Eskimos. Ottawa: Canadian
Martin, Paul Sidney, et al. Research Center for Anthropology.
1952 Mogollon Cultural Continuity and Change. Paddock, John
Fieldiana: Anthropology, vol. 40. Chicago: 1966a Comment on Dobvns, “An Appraisal of Tech-
Natural History Museum. niques for a New Hemispheric Estimate.” Current
Martinez, Maximino Anthropology, 7:437-38.
1928 Plantas Utiles de la Republica Mexicana. Mexico. 1966b “Oaxaca in Ancient Mesoamerica.” Ancient
1937 Catalogo de Nombres Vulgares y Cientificos de Oaxaca: Discoveries inMexican Archaeology and
Plantas Mexicanas. Mexico. History, ed. by John Paddock. Stanford: Stanford
Martin del Campo, R. University Press.
1944 “Huevos Subfosiles de Grulla en el Valle de Pangalo, K. I.
Mexico.” AIB, 15:313-18. 1930 “A New Species of Cultivated Pumpkin.” Bulletin

M ASSEY, W. C., AND C. M. OSBORNE and Plant Breeding,
of Applied Botany, Genetics,
1961 A Burial Cave in Baja California. University of 23 (3): 253-65. Leningrad.
California Anthropological Record, no. 16. Peterson, F. A.
Mayer-Oakes, W. J. 1959 Ancient Mexico: An Introduction to the Pre-
1959 “A Stratigraphic Excavation at El Risco, Mexico.” Hispanic Cultures. London: Allen and Unwin.
1963 Piper, C. V.
APS-P 103:332-73.
Medellin Zenil, Alfonso 1926 “Studies of American Phaseolinae.” U. S. National
1960 Ceramicas del Totonacupan: Exploraciones Ar- Museum Contribution U. S. National Herbarium,
queologicas en el Centro de Veracruz. Mexico. 22:663-701.
Meggers, R. and Clifford Evans, Editors. Popenoe, Wilson
J.,
Aboriginal Cultural Development in Latin

1941 “The Avocado —A Horticultural Problem.” CASY,
1941:79-85.
America: An Interpretive Review. Smithsonian
Miscellaneous Collections, 146.
and L. O. Williams
1947 “The Expedition to Mexico of October, 1947.”
MeHRINGER, P. J., JR., AND C. V. HAYNES, JR.
1965 “The Pollen Evidence for the Environment of
CASY, 1947:22-28.
1948 “Mexican Explorations of 1948.” CASY, 1948:54-
Early Man and Extinct Mammals at the Lehner
58.
Mommoth Site, Southeastern Arizona.” Am. Ant.,
Riehl, Herbert
31:17-24.
Merrill, E. D.
1965 Introduction to the Atmosphere. New York: Mc-
1950 “Observations on Cultivated Plants with Refer-
Graw Hill.
Roberts, L. M., et al.
ence to Certain American Problems.” Ceiba,
1957 Races of Maize in Colombia. Nat. Acad. Sci. Nat.
1:3-36.
Res. Council Pub., no. 510.
Miller, G. S., jr., and Remington Kellogg
Robertson, Donald
1955 List of North American Recent Mammals. U. S.
1963 “The Stvie of the Borgia Group of Mexican Pre-
National Museum Bulletin, 205.
Conquest Manuscripts.” Studies in Western Art:
Miranda, F. Acts of the Twentieth International Congress of
1942 “Estudios sobre la Vegetacion de Mexico, III.”
the History of Art, 3: 148-64.
AIB, 13:417-59.
1964 “Los Manuscritos Religiosos Mixtecos.” Actas y
1948 “Datos sobre la Vegetacion en la Cuenca Alta del
Memorias del XXXV
Corigreso Internacional de
Papaloapan.” AIB. 19:333-64.
Americanistas, Mexico, 1962, 1:425—35.
Motolinia (Fray Toribio de Benavente) 1966 “The Mixtec Religious Manuscripts.” In Ancient
1903 Memoriales. Paris. Oaxaca: Discoveries in Mexican Archaeology and
Mullerried, F. K. G. History, ed. bv John Paddock. Stanford: Stan-
1931 “Estudios Paleontologicos v Estratigraficos de la ford University Press.
Region de Tehuacan, Puebla.” AIB, 4-1:33-46. Roman y Zamora, Fray Jeronimo
Nicholson, H. B. 1897 Republicas de Indias; Idolatries y Gobierno en
1955 “The Temalacatl of Tehuacan.” Mex. Ant., vol. 8. Mexico y Peru antes de la Conquista. Coleccion
“The Problem of the Provenience of the Members de Libros Rares o Curiosos que traten de America,
of the ‘Codex Borgia Group’: A Summary.” In vols. 14 and 15. Madrid.
press. Romano, Arturo
Noguera, Eduardo 1955 “Nota Preliminar sobre los Restos Humanos Sub-
1940 “Excavaciones en Calipan, Estado de Puebla.” fosiles de Santa Maria Astahuacan.” Anales de
Mex. Ant., vol. 5, no. 3—5:63-124. IN AH, 7:65-74.
316
BIBLIOGRAPHY
Rzedowski, J. ticesnear Tehuacan, Mexico.” Economic Botany,

1962 “Contribuciones a la Fitogeografia Floristica e 17:200-10.
Historica de Mexico: I. Algunas consideraciones 1965a “Agriculture, Tehuacan Valley.” Fieldiana:
acerca del Elemento Endemico en la Flora Botany, 31:49-100. Chicago: Natural History
Mexicana.” Boletin de la Sociedad Botanica Museum.
1964
Mexicana, 27:52-65. 1965b “Flora, Tehuacan Valley. Ibid., 31:107—42.
Salas, G. P. and R. S. MacNeish
1949 “Bosquejo Geologico de la Cuenca Sedimentaria “Antiquity of American Polyploid Cottons.”
de Oaxaca.” Bol. Asoc. Mex. Geol. Petroleros, Science, 143:675-76.
1-2:79-156. Mexico. Smith, H. M.
Sauer, J. D. 1946 Handbook of Lizards. Ithaca: Comstock Publish-
1950 “The Grain Amaranths: A Survey of Their His- ing Co.
tory and Classification.” MBG-A, 37:561-632. Snead, P. H. A.
Schmidt, K. P., and R. F. Inger 1962 “Longevity of Microorganisms.” Nature, 195
1957 Living Reptiles of the World. New York: Double- (4842): 643-46.
day. Standley, P. C.
Schmidt-Nielsen, Knut 1920—26 “Trees and Shrubs of Mexico.” U. S. National
1964 Desert Animals: Physiological Problems of Heat Museum Contribution U. S. National Herbarium,
and Water. Oxford: Clarendon Press. 23.
ScHROEDER, C. A. Steward, J. H.
1947 “The Expedition to Mexico of May, 1947.” CAST, 1955 Theory of Cultural Change. Urbana: University
1947:29-33. of Illinois Press.
SCHUCHERT, C., AND C. O. DUNBAR Stonor, C. R., and Edgar Anderson
1947 A Textbook of Geology; II: Historical Geology. 1949 “Maize among the Hill Peoples of Assam.”
4th ed. MBG-A, 36:355-404.
Sears, P. B. Strong, W. D., and Clifford Evans
1953 “An Ecological View of Land Use in Middle 1952 Cultural Stratigraphy in the Viru Valley, Northern
America.” Ceiba, 3:157-65. Peru. New York.
and K. H. Clisby Sullivan, L. R., and M. Hellman
1955 “Palynology in Southern North America; IV: 1925 The Punin Calvarium. AMNH-AP, 23.
Pleistocene Climate in Mexico.” Geological So- SwADESH, MaURICIO
ciety of America Bulletin, 66:521-30. 1959 Indian Linguistic Groups of Mexico. INAH.
Seler, Eduard Tamayo, Jorge
1904- Codex Borgia. Eine altmexikanische Bilderschrift 1962 Geografia General de Mexico. 2nd ed. Institute
1909 der Congregatio de Propaganda Fide Herausgege- Mexicano de Investigaciones Economicas, Mexico.
ben auf Kosten Seiner Excellenz de Herzogs von and R. C. West
Loubat, erlautert von Eduard Seler. 3 vols. Berlin. 1964 “The Hvdrographv of Middle America. HMAI.
1963 Comentarios al Codice Borgia. Mexico: Fondo de 1:84-121.
Cultura Economica. Taylor, W. W.
Shamel, A. D. 1956 “Some Implications of the Carbon-14 Dates from
1938 “Avocado Studies in Mexico in 1938.” CASY, a Cave in Coahuila.” Texas Archaeological So-
1938:67-85. ciety Bulletin, 27:215-34.
Sharp, A. J. Terra, Helmut de
1953 “Notes on the Flora of Mexico: World Distribu- 1949 “Earlv Man in Mexico.” Tepexpan Man. VFPA,
tion of the Woody Dicotyledonous Families and no. 11.
the Origin of the Modern Vegetation.” Journal of Towle, M. A.
Ecology, 41:374-80. 1961 The Ethnobotany of Pre-Colombian Peru. VFPA,
Skovsted, A. no. 30.
1934 “Cytological Studies in Cotton; II: Two Inter- Trotter, M., and G. Gleser
specific Hybrids between Asiatic and New World 1958 “A Reevaluation of Estimation of Stature.” Ameri-
Cottons.” Journal of Genetics, 28:407. can Journal of Physical Anthropology, n.s. 16:79.
1937 Ibid. “IV: Chromosome Conjugation in Inter- Vaillant, G. C.
specific Hvbrids.” 34:97-134. 1966 Aztecs of Mexico. Revised by S. B. Vaillant.
Smith, C. Earle Baltimore: Penguin Books.
1950 “Prehistoric Plant Remains from Bat Cave.” BML, Van Cleave, H. J.,
and J.
A. Ross
14 (7). 1947 “A Method of Reclaiming Dried Zoological Speci-
1963 “Something Old, Something New; Farm Prac- mens.” Science, 105 (2725) :318.
317
Vavilov, N. I. 1957 “Archeological Cucurbitaceae from a Cave in

1931 “Mexico and Central America as the Principal Southern Baja California.” Southwestern Journal
Center of Origin of Cultivated Plants of the New of Anthropology, 13:144-48. Albuquerque, New
World.” Bulletin of Applied Botany, Genetics, and Mexico.
Plant Breeding, 26 (3): 179-99. Leningrad. 1961 “Biosystematics of the Cultivated Cucurbita.”
Villa R., Bernardo Recent Advances in Botany, 859-62.
1953 “Mamiferos Silvestres del Valle de Mexico.” AIB, and B. Bird
J.
23:269-492. 1949 “Identification and Significance of the Cucurbit
1954 “Contribucion al Conocimiento de las Epocas de Materials from Huaca Prieta, Peru.” American
Caida y Nacimiento de la Cornamenta y de su Museum Novitates, 1426:1-15.
Terciopelo en Venados Cola Blanca ( Odocoileus
and G. F. Carter
virginianus de San Cayetano, Estado de Mexico,
1961 “A Note on the Longevity of Seed of Lagenaria
Mexico.” AIB, 25:451-61.
siceraria (Mol.) Floating in Sea
Standi, after
Vivo Escoto, J. A.
Water.” Torrey Botanical Club Bulletin, 88:104-
1964 “Weather and Climate of Mexico and Central
106.
America.” HMA1, 1:187-215.
and H. C. Cutler
Wagner, P. L.
1966 “Food Plants in a Mexican Market.” Economic
1964 “The Natural Vegetation of Middle America.”
Botany, 20:6-16.
HMAI, 1:216-64.
Wakefield, E. F., and S. C. Dellinger
H. C. Cutler, and R. S. MacNeish
1936 “Diet of the Bluff Dwellers of the Ozark Moun- 1957 “Cucurbit Materials from Three Caves near
tains and Its Skeletal Effects.” American Internal Ocampo, Tamaulipas.” Am. Ant., 22:352-58.
Medicine, 9:1412-18. and G. N. Davis

Walter, H. V., A. Cathoud, and A. Mattos 1962 Cucurbits: Botany, Cultivation, and Utilization.
1937 “The Confins Man: A Contribution to the Study London: Leonard Hill.
of Early Man
South America.” Early Man, ed.
in White, Theodore
by G. G. MacCurdv, 345-48. 1953 “A Method of Calculating the Dietarv Percentage
Watson, P.
J.
and R. A. Yarnell Food Animals Utilized by Aboriginal
of Various
1966 “Archeological and Paleoethnobotanical Investi- Peoples.” Am. Ant., 18:396-98.
gations in Salts Cave, Mammoth Cave National Whitford, A. C.
Park, Kentucky.” Am. Ant. 31:842-49. 1943 “Fiber Plants of the North American Aborigines.”
Weather wax, Paul Journal the New York Botanical Garden,
of
1954 Indian Corn in Old America. New York: Mac- 44:25-34.
millan.
Willey, G. R., and J. M. Corbett
Webb, W. and R. S. Baby
S.,
1954 Early Ancon and Early Supe Culture. Columbia
195 1 The Adena People, No.2. Ohio Historical Society.
Studies in Archeology and Ethnology, III.
Columbus: Ohio State University Press.
Williams, L. O.
Webber, J.
M.
1952 “Beans, Maize and Civilization.” Ceiba, 3:77-85.
1935 “Interspecific Hybridization in Gossypium and
Willis, J.
C.
the Meiotic Behavior of F, Plants.” Journal of
Agricultural Research, 51:1047-70.
1951 Dictionary of the Flowering Plants and Ferns.
6th ed. Cambridge: Cambridge University Press.
1939 “Relationships in the Genus Gossypium as Indi-
cated bv Cytological Data.” Ibid., 58:237-61. Woodbury’, R. B.
Wellhausen, E. et al.
1961a “Climatic Changes and Prehistoric Agriculture in
J.,
1952 Races of Maize in Mexico. Bussev Institution, the Southwestern United States.” Annals of the
Harvard University. New York Academy of Science, 95:70.5—709.

West, R. C. 1961b Prehistoric Agriculture at Point of Pines, Arizona.
1964 “Surface Configuration and Associated Geologv of Society for American Archaeology, Memoir no.
Middle America.” HMAI, 1:33-83. 17. Menasha and Salt Lake Citv.
Wheat, B. 1961c “A Reappraisal of Hohokam Irrigation.” Ameri-

J.
1954 “Southwestern Cultural Interrelationships and the can Anthropologist, 63:550—60. Lancaster and
Question of Area Contradiction.” American Menasha.
Anthropologist, 56 (4):576-91. Lancaster and Yarnell, R. A.
Menasha. 1964 Aboriginal Relationships between Cultures and
Whitaker, T. W. Plant Fife in the Upper Great Lakes Region.
1956 “The Origin of the Cultivated Cucurbita.” Amer- Anthropological Papers, University of Michigan,
ican Naturalist, 90:171-76. no. 23.
318
BIBLIOGRAPHY
Young, B. H.
1910 The Prehistoric Men of Kentucky. Filson Club
Publications, no. 25. Louisville: J.
P. Morton.
Zhiteneva, N. E.
1930a “Survey of the Principal Literature on the
Systematics of Pumpkins and Squashes.” Bulletin
of Applied Botany, Genetics, and Plant Breed-
23ing,343-56. Leningrad.
(3):
1930b “The World’s Assortment of Pumpkins.” Ibid.,
157-207.
INDEX
Note: Pages on which definitions or descriptions occur, or on which major material is concentrated, are indicated by bold type.
Abad y Queipo, M., 42 Altepexi, Puebla, 41-46 passim, 52, 58, 59

Abejas Cave (Tc 307): stratigraphy, 16; location, 37, 45, 228; Amaranthus spp.: remains in caves, 225, 227, 231—235 passim;
climate, 55, 64; flora, 64, 227-228; fauna, 139, 163-165, 172, described, 239; remains in feces, 272, 274, 285, 288; cultiva-
173 tion and distribution, 290-293, 308; preparation, 302
Abejas phase: and Archaic period, 19; described, 23—24; human Anderson, Edgar, 213, 221
skeletons, 97; fauna, 162, 167, 175; hunting patterns, 173, Andrews, G. Willys, 205, 208
304—305; cultivated plants, 205, 208, 213, 217; diet and food Antelope: and climate, 64, 140, 144; described, 152; in earliest
preparation, 233, 285-288 passim, 299-300, 302; agriculture, phase, 158, 166; and hunting, 171, 172
293, 294, 306; plant collecting, 306; mentioned, 18, 163, 165 Antevs, Ernst, 64
Acacia spp., 134, 231, 232, 240-241, 291 Antlers, 157, 159, 168, 169
Acatepec, Puebla, 25, 43, 74 Apala, Puebla, 39
Acatlan, Puebla, 67 Apoclanthera spp., 216, 217, 219, 290
Acorns, 239 Aqueducts, 11, 38, 81-82. See also Irrigation
Acrocomia mexicana (coyol): remains in caves, 225, 228, 229, Archaic period, 19
233, 234; question of cultivation, 230, 252, 254, 291; de- Architectural structures, 41, 45, 123-129 passim, 130
scribed, 236 Argentina, 212
Cumbres de, 74
Acultzingo, Veracruz, 59; Armenta, Juan, 152
Agate Basin points, 304 Arrocillo Amarillo (maize), 195, 197
Agave spp. (maguey): modern plantations, 35; effect on teeth, Arrows, 127, 171, 174, 304
106; in Codex Borgia, 126; ecological zones, 137, 158, 224, Arroyo de Atexcala, 38
250; remains in caves, 225-234 passim; modern spp., 231; Arroyo Coxcatlan, 37
described, 238; uses and question of cultivation, 250-251, Arroyo Lencho Diego, 8, 24, 37, 45, 173; site of dam, 11, 74;
254, 288, 291, 292, 293; remains in feces, 265—288 passim fauna, 139, 147, 149, 163-166; flora, 227-228
preparation, 270, 302, 303 Arthritis, 96, 99, 100. See also Osteoarthritis
Agriculture: and rise of civilization, 14, 309; post-Conquest, 44, Artifacts, 27-28, 32; and subsistence, 21, 304, 305, 306
221, 222; animal domestication, 150, 155; origins and develop- Atenayuca, Puebla, 74
ment, 220-221, 222, 226, 248—255, 290-295; evidence from Atlatls, 127, 128, 171, 304
plant remains, 225-234 passim New World distribution, 253, Atoyac Valley, 80
254, 293-295; role in subsistence, 253, 299-300; evidence Atzinsintla, Puebla, 74
from feces, 271-281 passim, 285; techniques, 306-307, 308. Avocados (Persea americana ): cultivation, 224, 230, 240, 254,
See also Beans; Fruits; Irrigation; Maize; Squash; Subsistence 292, 293; remains in caves, 225, 231-234 passim described,
Aguilera, J.
G., 68, 70, 72 239—240; role in diet, 251; in Peru, 253; mentioned, 36, 64,
Ajalpan, Puebla, 43, 44, 66, 82, 83 181, 222, 290, 302
Ajalpan phase: and early Formative, 19; described, 24; human Awls, 127, 168-169
skeletons, 97-98, 106, 107; fauna, 132, 165, 167, 168; sub- Axusco (Axochitlan), Puebla, 41, 46, 71, 73
sistence, 173, 294, 300, 304, 305, 306; diet and food prepara- Aztecs, 14, 41, 114, 126, 203
tion, 266, 271-272, 300, 302-303 Azumbilla, Puebla, 245
Ajalpan plain, 44, 132, 146, 147, 148, 154, 168
Ajalpan sites (Ts 204, A-E), 9, 17, 44 Bailey, L. H„ 213
Ajuereado phase: and Early Man period, 19; described, 23; Ball courts, 24, 127
climate and fauna, 64, 140-145; human skeletons, 92-93; diet, Balzaguillo, Valley of, 43
105, 300; fauna and hunting patterns, 158-159, 167, 172; Barghoorn, Elso S., 6, 19
subsistence trends, 305 Bark beaters, 129
Albanese, A. A., 202 Bark cloth, 24, 129
Alcedo, Antonio de, 42-43, 44 Barley, 14, 38, 43
Alexander, H. L., }r., 171 Barlow, Robert H., 114
38
Alfalfa, 35, Barranca de Acultzingo, 34
Aloapan, Oaxaca, 74 Barranca de Alterango, 37
321
Barranca de los Mangos, 36—37, 138, 150, 151 Calipan, Puebla, 24, 45, 58, 80, 119
Barranca de San Antonio, 36 Camps and campsites, 23, 24, 97, 157-174 passim
Barranca de Soyolapa, 37 Canada Oaxaquena-Poblana, see Oaxaca-Puebla Trough; Te-
Barrancas: farming in, 179, 224-225, 306, 308; flora, 222, 224- huacan-Cuicatlan Basin
225, 240, 248 Canada de San Antonio, 82
73
Basalt, Canals, see Irrigation
Basketmaker III, 208 Canavalia sp. (jack beans), 208, 233, 290, 294
Baskets, 23, 24, 38, 129 Cane, river, 44, 134, 158, 236
Bat Cave, New Mexico, 3, 4, 252, 293 Canyons, 139. See also Arroyo Lencho Diego; Tecorral Canyon
Bats, 138, 139, 145-146, 170 Cardenas, Martin, 213
Beads, bone, 167, 169 Casimiroa edulis (white sapote): remains in caves, 225, 227, 228,
Beans ( Phaseolus spp.): remains in caves, 201—210, 250, 290; 233, 234; and cultivation, 230, 252, 293; described, 241—242;
remains in feces, 272, 276, 280, 285—288 passim preparation, mentioned, 221, 290
302, 303; in Mesoamerican diet, 308; P. acutifolius (tepary), Caso, Alfonso, 114, 115
201-210 passim, 233, 234, 294; P. coccineus (runner), 202, Cassava, see Manihot
204, 205, 233, 234, 294, 295; P. lunatus (lima, sieva), 201-210 Cassia pringlei, 231
passim, 234, 295; P. vulgaris (common), 201-210 passim, 225, Castela tortuosa, 132, 231
232-234 passim, 275, 279, 281, 293, 295 Cataloguing, 27-28, 32, 33
Beaucarnea gracilis (sotolin), 225, 227, 228, 237 Cataluna, Veracruz, 51
Bees, 280 Catan points, 304
Beetles, 265, 273-280 passim Caterpillars, 268, 269
Bernal, Ignacio, 12, 119 Cattle, 43
Birds: local distribution, 134-139 passim, 159-167 passim de- Caves (sites), 4, 27, 171. See also names of caves
cribed, 154—156; collecting of, 172, 173, 174 Ceiba parvifolia, see Pochote
Bison, 171 Celts, 128
Blades, 127, 129, 162, 163, 167, 168, 169 Cempoala pottery, 117
Blasquez, L., 73 Cenozoic, of Tehuacan Valley: Middle and Upper, 66, 68, 72,
Bobcats: description and distribution, 134, 139, 151—152, 159, 73, 80-81; Lower, 84, 88, 89
163, 171; hunted or trapped, 172, 174; remains in feces, 264, Central America, see Mesoamerica
280, 289 Cephalocereus hoppenstedtii (tetecho): local distribution, 138,
Bobwhites, see Quail 139, 224; remains in caves, 225, 227, 229, 232; described, 245
Boiling, 202, 204, 302 Ceramics, see Pottery
Bolivia, 214, 292 Ceremonialism, 23, 24, 115
Bone artifacts, 127, 167, 168 Cerro Agujereado, 37
Bone marrow, 159, 163, 165, 169, 172, 173, 286 Cerro Chichiltepec, 37, 84
Bones, see Fauna; Skeletons, human Cerro Colorado de Ajalpan, 71, 77
Borah, Woodrow, 22, 41, 42 Cerro Colorado de Tehuacan, 36, 71
Borgia Codex, see Codex Borgia Cerro Ellotzihuatl, 37
Borgia, Stephano Cardinal, 115 Cerro de la Mesa de Tehuacan, 35, 71-77 passim, 117, 135, 225
Bracelets, 128, 130 Cerro de las Mesas pottery, 115
Brahea dulcis (palm), 63, 225, 227, 231, 236 Cerro Paredones, 36
Brasero, 119 Cerro Pelon, 46
Brazil, 112, 259 Cerro Prieto, 45
Brewing, 275, 286, 287, 303 Cerro Tepetroje, 90
Brown’s Valley Man, 111-112 Cerro el Zapote, 149—150
Brush, Charles, 308 Chalma, Puebla, 36
Bumelia laetevirens (tempixquistle), 138, 227, 246, 252, 254, 291 Chalqueno (maize), 197, 295
Burgoa, Francisco de, 114 Chapalote (maize), 178, 180, 194, 195, 197; described, 188-189;
Burials, 23, 32, 92—103 passim distribution, 294, 295
Burins, 169 Chapulco, Puebla, 41, 71
Burning and charring, evidence in human skeletons, 94—103 Cherokees, 202
passim Chiapas, 4, 12, 15, 202, 205, 210, 293
Bursera arida, 231 Chicozapotes, 63, 64
Butchering, 43, 159-169 passim, 173 Chihuahua, 2, 4, 14
Butterlin, Jacques, 71 Chilac, Puebla, 39-44 passim, 55, 58, 73, 82
Byers, Mrs. Douglas S., 9, 10 Chilca, Peru, 202
Chili peppers ( Capsicum annuum ): early remains, 181, 222, 225,
Cacaloapan, Puebla, 66 232, 233; described, 247-248; cultivation and distribution,
Cacti, 106, 137, 138, 224, 245—246, 302; remains in feces, 266— 250, 254, 290—291, 292, 308; remains in feces, 266—270 pas-
278 passim, 285, 286, 287. See also names of species sim, 274-281 passim, 286, 288; preparation, 302, 303
Caesalpinia velutina, 241 Chinanteca, 39
Calapilla, Oaxaca, 58, 59, 60 Chipmunks, 158
Caliche, 77 Chipped-stone artifacts, 21-22, 23, 24
322
INDEX
Chocha language, 43 Conico (maize), 295

Chocho-Popoloca, roof type, 125—126 Conjomeapa, Puebla, 43
Cholula, Puebla, 115, 126, 204 Cook, S. F., 22, 41, 42, 62-63
Cholulteca I-III pottery, 115 Cook de Leonard, Carmen, 39, 114, 125, 126
Choppers, stone, 163 Cooking methods, 302-303. See also Food preparation
Chronology, in Tehuacan Valley, 15, 17, 18, 27, 33; on basis of Cooper Clark, James, 203
agricultural remains, 295-296; on basis of subsistence trends, Cooper Cole, Fay, 2
296-300 Coprolites: animal, 263, 266, 281—284, 286—287
Chupandilla, see Cijrtocarpa human: methods of analysis, 261-264; literature on, 264—
Cipiapa, hacienda, 67, 70 266; described by phase, 266—281; dietary evidence, 267-274
Cipiapa formation, 70 passim, 279, 281, 284—288; and food preparation, 288-289
Ciruela, see Spondius Cord and cordage, 38, 129
Cissus, sp., 225, 227, 244 Corn, see Maize
Cities, prehistoric, 24, 100 Cosahuico, see Sideroxijlon
Citlaltepetl, 34, 35, 64 Costa Rica, 202
Citrus trees, 36, 221 Cotton, 24, 42, 63, 290; Gossypium barbadense, 257, 258, 259
Ciudad Serdan, Puebla, 74 G. hirsutum: remains in caves, 225, 227, 228, 232, 233, 234;
Clapulco, Puebla, 43 described, 244; origin and distribution, 252, 254, 259, 260,
Clarion Fracture Zone, 34 293, 294; earliest boll segments analyzed, 256—260
Clark, Donovan, 18 Coxcatlan, Puebla, 41, 42, 43, 45, 221; ancient town, 37, 114,
Classic period, 19, 25, 27, 175, 286, 292 115; climate, 51, 54, 55, 58, 59
Clay artifacts, 17 Coxcatlan Black-on-Orange pottery, 119, 120, 123
Climate, in Tehuacan Valley: present-day, 48-62; early post- Coxcatlan Cave (Tc 50): discovery, 5; stratigraphy, 15; location,
Conquest, 62-64 37, 45; burials, 93—96, 100-102; faunal remains, 158—163;
late Pleistocene to post-Pleistocene: evidence from plant maize, bean, squash remains, 179, 194—195, 210, 217, 232,
remains, 19, 222, 248—249; meteorological evidence, 64—65; 233, 234; other plant remains, 231-234. See also Coxcatlan
geological evidence, 75; faunal evidence, 139, 140-145; and subarea
wild maize, 178 Coxcatlan Coarse pottery, 117, 119, 120, 123
Clisby, K. H„ 4 Coxcatlan Gray pottery, 119, 120
Clothing, 113, 129 Coxcatlan phase: and Archaic period, 19; described, 23; human
Clubs, wooden, 129 skeletons, 96—97, 105, 107, 109; fauna and hunting, 162, 165,
Coapan, Puebla, 39, 44, 70 167, 172-173, 304; maize, bean, squash remains, 180, 205,
Coatepec site (Ts 368e, 368w), 10, 12; stratigraphy, 16; burials, 210, 214, 217, 232, 234; agriculture, 233, 291-293, 306; diet,
97-103 passim fauna, 168, 169 233, 268-269, 287-288, 299; food preparation, 302
Codex Bodley, 126 Coxcatlan points, 304
Codex Borgia: and other Mixtec codices, 114; and Venta Salada Coxcatlan Polychrome pottery, 119, 123
pottery, 114, 115, 117-123; distinctive glyphs, 115-117, 126; Coxcatlan Red pottery, 117, 120
buildings, 123—126; tools and other cultural manifestations, Coxcatlan Red-on-Cream pottery, 119, 120, 123
126-130 Coxcatlan subarea: climate, 55, 64; geology, 70, 82; described,
Codex Cospi, 114 138-139, 230-231; faunal sequence and hunting, 158—163,
Codex Fejervary- Mayer, 114 172, 173, 174
Codex Laud, 114, 117 Coxcatlan Terrace (Ts 51), 16, 37, 158-163 passim, 172, 173
Codex Mendozo, 203 Coyol, see Acrocomia
Codex Nuttall, 126, 127 Coyomeapan, Puebla, 74
Codex Porfirio Diaz, 114, 126 Covotes ( Canis latrans ): local distribution, 132, 138, 139, 158;
Codex Sanchez Solis, 126 described, 149—150; hunted or trapped, 171; hair identified,
Codex Selden, 126 264, 279, 280; meat yield, 296; mentioned, 162, 166, 167, 168
Codex Vaticanus 3773, 114, 127 Cranes, sandhill, 135, 157, 162, 296
Codex Vienna, 126 Cremation, 96, 103. See also Burning and charring
Codices, 23, 24, 114, 115, 117, 126 Crescentia cujete (tree gourd), 212, 218, 219, 225, 232, 290;
Coixtlahuaca, Oaxaca, 63, 74 description and distribution, 216—217, 294
Collecting, plant, see Gathering Cretaceous, 68, 70, 84, 88
Colombia, 190, 195, 202 Cross-dating, 18
Comales, 21, 303 Cucurbita, see Squash
Commelina erecta, 231 Cues (San Juan de los), Oaxaca, 45, 82
Commerce, see Trade Cuicateca, 39, 42
Components (occupations), of Tehuacan sites, 17, 18, 21, 22; Cuicatlan, Oaxaca, 39, 41, 61
chronology of, 295-300 Culhua Mexica, 41
Compositae, 248 Cultivation, see Agriculture
Concepcion Papalo, Oaxaca, 74 Cyclones, 50
Condalia mexicana, 232, 243 Cijrtocarpa procera (chupandilla): native tree, 138, 221; remains
Confins Man, 112 in caves, 225—234 passim question of cultivation, 230, 251,
Confite Morocho (maize), 180 254, 291; described, 242—243; use, 302
323
Daggers, 127, 128 El Risco site, Valley of Mexico, 112
Dam, remains of (Arroyo Lencho Diego): described, 84—87; and Engel, Frederic, 202
agriculture, 224, 228, 229, 230, 291, 292, 306; mentioned, 11, Environment, see Ecological zones
37, 165, 286 Eocene-Oligocene, 72
Danger Cave, Utah, 252, 265 Erosion, 61-62, 63, 74, 75, 77, 82, 83, 221
Davalos Hurtado, Eusabio, 7 Escontria chiotilla (jiotilla), 224, 231, 245
Davis, G. N., 213 Eskimo, 113
Deer, white-tailed ( Odocoileus virginianus) local distribution, Evapotranspiration, 54
134, 135, 138, 139, 152; and climate, 140, 144; at Levi Rock Excavation techniques, 28-32
shelter, 145, 171; described, 152; hunting of and butchering
techniques, 157-169 passim, 171-175 passim; hair identified, Farming, see Agriculture
278, 280, 285, 289; meat yield, 296 Fauna: in documents, 63, 126, 127; in Tehuacan Valley, 132—
Dent maize, 197, 295 139, 156-170; as climatic indicators, 140-145; species de-
Dentition, see Teeth scribed, 145—156; hunting and trapping of, 171-174, 304,
Dick, Herbert, 3 305; hairs identified, 264; diet of, 286-287; meat yield, 296.
Dickerman, Robert, 162 See also names of animals
Diet: effect on teeth, 105-107; of animals, 175-176, 281-284; Feathers, 273, 274, 281
corn-bean relationship, 202-203; modern Mexican, 223, 232; Feces, see Coprolites
wild plants in, 228-234 passim, 266-281 passim; staples of, Fertilizer, 38, 44, 221
249-250; fruits in, 251-252; evidence from feces, 266-281, Field techniques, 2.5—33
285, 286; trends, by phase, 300; early, compared, 307. See Figurines, 23, 24
also Subsistence; Sustenance Fire, 126; goddess of, 129
Digging, see Excavation techniques Fire-making, 129
Digging stick, 129 Fish, 159, 162, 171
Dioon edule, 235 Fishing, 43, 127, 162, 172, 173
Dioscorea sp., 238 Flannery, Kent V., 307
Diospyros digyna (black sapote): remains in caves, 225, 228, 233, Fleas, 277, 278
234; described, 247; and cultivation, 252, 253; remains in Flies, 265, 269, 273-274, 277
feces, 268-277 passim, 286, 287, 288 Flint Knife glyphs, 115-117
Diseases, 42, 94-97 passim, 99, 108-110 by agriculture,
Flora: effect of climate on, 51, 54, 63; disturbance
Dogs, domestic: described, 150; introduced to Tehuacan Valley, 61, 62, 220-222; of Tehuacan Valley subareas, 132-139 pas-
162, 165, 175, 294; meat yield, 296; mentioned, 163, 167, 168, sim, 225-232 passim; composition, 220; useful to man, 221,
169, 173, 174, 290 223; and climate, 248-249. See also names of plants; Plant
Domestication, see Agriculture remains
Doves, 159, 162, 163, 166, 172, 173; white-winged (Z enaida Florida, 212
asiatica ),134-139 passim, 155, 167; ground {Columbia gallina Flowers, 38, 43; as food, 238, 275, 277, 287
passerina), 138, 139, 155 Food, see Diet; Fauna; Flora; Subsistence; Sustenance
Drills, wooden, 129 Food preparation: reflected by artifacts, 21, 302-303; in Borgia
Ducks (Anas cyanoptera), 134, 135, 154 Codex, 126; meat, 159, 162-169 passim; beans, 202, 204, 209;
Durango site, Colorado, 208 squash or pumpkins, 214, 216, 271, 279; evidence from feces,
Dzibilchaltun, Yucatan, 205, 207 267-278 passim, 288; changing techniques, 302-303
Ford, James A., 12
Early Man period, 18, 19 Forests: oak-and-pine, 27, 35, 61, 151, 158, 202, 224; thorn,
Earrings, 128 150-158 passim, 172, 173, 174, 224, 287
Echinocactus grandis (barrel cactus), 135, 245 Formative period, 19
Ecological zones: of Tehuacan Valley, 132-139, 178-179, 224- Fossils, 70, 88, 89, 90
231 passim; of Mesoamerica, 307-308 Foxes: gray ( Urocyon cinereoargenteus), 132-139 passim, 150,
Ecuador, 112 159-168 passim, 171-174 passim; late Pleistocene, 140, 150,
Egret ( Casmerodius albus), 135, 154 158, 172
Ejidos, 44, 46 Fractures, bone, 94, 95, 96, 108-109
Eloxuhitlan (Eloxohitlan), Puebla, 43 Freeman, G. F., 208, 209, 210
El Riego, Puebla, 66, 72, 81, 82 Frightful Cave, Coahuila, 144
El Riego bottling plant, 137 251-252, 307; wild species in
Fruit: cultivation, 43, 63, 64, 240,
El Riego Cave (Tc 35e, w), 15-16, 81, 167-168, 225 Coxcatlan subarea, 138, 232, 233, 234; collecting, 173, 174,
El Riego phase: and Early Man period, 19; described, 23; human 305, 306; grown in watered fields, 230, 306, 308; role in diet,
skeletons, 93-96; fauna and hunting patterns, 159, 163—165, 252, 302; remains in feces, 274, 275, 280
166, 172, 304; food preparation, 159, 289, 302; and origins of Furniture, 129
agriculture, 232, 234, 291-293, 306; cotton boll segment, 256-
260; diet, 266-267, 299, 300, 307 Galls, edible, 251
El Riego points, 304 Game drives, 172, 174, 305
El Riego springs, 137-138, 167-168, 225 Garci-Crespo bottling plant, 137
El Riego subarea: climate and hydrology, 54, 55, 58, 81-82; Garments, 117, 129
described, 135-136, 137-138, 167-168, 225 Garters, 130
324
INDEX
Garyito points, 304 Hurricanes, 50—51, 55, 58, 59

Gathering, plant, 21, 23, 138, 173, 174, 305-306. See also Plant Hydrogeology, 66, 75—87 passim
remains Hydrography, 73-74, 81-87
Geology, of Tehuacan Valley, 66—90 passim Hydro-horticulture, 306, 307
Glyphs, 11,5-1 17, 126
Goats, 38, 43, 44, 61, 62, 170, 210, 222 Ignacio Mejia, Puebla, 45
Gods and goddesses, 23, 115, 119, 120, 126, 129, 130 Iguanas: green ( Iguana iguana), 134, 135, 153
Gophers 156, 171, 296; Heterogeomys sp., 132, 138, 147, 169, black Ctenosaura pectinate ): local distribution, 134, 138,
(
264; Cratogeomys sp., 136, 139, 147, 163—175 passim 139; 153—154; trapped or collected, 174; men-
described,
Gorgets, 128, 130 tioned, 140, 162, 163, 165, 170
Gossypium, see Cotton Incense burners, 23, 24, 119
Gourds, 129, 212, 254, 290. See also Crescentia; Lagenaria Infiernillo phase, 215, 287, 307
Grain, 38, 43 Insects, 261-280 passim
Grass and grasses: local distribution, 132, 135, 136, 138; prepara- Interdisciplinary approach, and Tehuacan project, 6, 7, 14—24
tion of seed, 204, 270, 288, 302; remains in caves, 231, 232, passim
233, 235—236; remains in feces, 267, 269, 270, 274, 275. See International Congress of Americanists, XXXV, 10
also Setaria Irrigation: structural evidence of ancient use, 11, 37, 38, 44—45,
Graves, 16. See also Burials 81-82, 84—87; and beans, 205; long-range effects, 221; meth-
Great Basin, U. S., 203 ods, 223-224, 306, 307; inferred from plant remains, 240, 242,
Great Lakes Basin, 64 247, 252; role in agriculture, 306—309 passim
Griffin, James B., 5 Irwin, Henry, 6, 19
Grinding, of seeds, 21, 204, 270, 288, 302-303 Isla de los Sacrificios, Veracruz, 115
Ground cherry, see Physalis Islona de Chantuto, Chiapas, 307, 308
Ground-stone artifacts, 22, 23, 24, 128, 288 Isotopes Incorporated, 7
Guaje (pods of tree legumes), 137, 138, 225, 227, 270-277 pas- Ixcateca, 39
sim. See also Acacia; Leucaena Ixcatlan, Oaxaca, 38, 39, 41, 62, 63; described by Alcedo, 42—43
Guanajuato, 72 Ixcatlan Valley, 64
Guatemala, 157, 218, 221, 308, 309; and beans, 201, 202, 205; Ixtaccihuatl, 64
and Cucurbita, 213, 214; and cotton, 258 Ixtlilxochitl, 114
Guava, 229, 252, 254, 290; described, 246; distribution, 253, 295
Guerrero, 4, 293 Jack rabbits, 174; late Pleistocene, 140, 144, 146, 147, 171-172;
Guianas, the, 259 white-sided ( Lepus callotis), 135-138 passim, 146, 167, 168,
Gulf Stream, 48 169, 173, 174
Gutierrez, Bicardo, 4 Jaguars, 127
Gypsum Cave, Nevada, 144 Jalisco, 201, 209
Jatropha spp.: J. urens, 135, 139; J. neopauciflora, 227, 229, 232,
Haciendas, 43, 44, 46 233, 234, 242; J. diocia, 231; unidentified sp., 242
Hair: identification, 261, 263-265; of various animals, 269, 276- Jewelry, see Ornaments
280 passim amount eaten, 285, 289 Jicotlan, Oaxaca, 74
Halophytes, 224 Johnson, Mrs. Frederick, 9, 11
Hammers, antler, 168 Juke Box Cave, Utah, 265
Handles, wood, 128-129 Jurassic, 66, 68
Harrell points, 304
Hats, palm-leaf, 38, 63 Kangaroo rat, see Mouse, kangaroo
Hawks, 167, 168, 169, 173; Buteo sp., 135, 136 Karst phenomena, 77
Headdresses, 130 Killdeer( Charadrius vociferus), 134, 135, 155
Hechtia sp., 225, 227, 237 Kingsborough, Edward King, Lord, 115
Hernandez X., E., 3 Knives, stone, 127, 128
Herrera y Tordesillas, Antonio de, 114 Koeppen classification, 54, 55
Highways, 34, 44 Krickeberg, Walter, 126
Hoes, 129
Honduras, 4 Ladle, 129
Honey comb, 280 Lagenaria siceraria (bottle gourd): described, 216; remains in
Horse, extinct, 140, 144, 152, 158, 171 caves, 217, 219, 225, 232; cultivation and distribution, 293;
Huaca Prieta, Peru, 204, 214, 219, 252, 265, 292, 294 mentioned, 181, 212, 290
Huaje trees, 63. See also Guaje Lagoa Santa, Brazil, 112
Huajuapan formation, 71, 72, 73 Lagunas, Puebla, 59
Huajuapan de Leon, Oaxaca, 25 Languages, in Tehuacan Valley, 39
Huatusco, Veracruz, 60 La Perra Cave, Tamaulipas, 3, 178, 194
Huautla de Jimenez, Oaxaca, 39, 59 Las Canoas site (Ts 367), 16, 44, 151, 168, 169
Humboldt, Alexander von, Baron, 115 Laureles, Michoacan, 259
Humidity, 49 Leaves, collecting of, 305, 306
Hunting, 138, 158-169 passim patterns, 171-174, 175, 304—305 Lechugilla, 135, 137, 139
325
Leg bands, 130 163, 165, 168, 169, 289, 302—303; remains in feces, 266-281
Lemaireocereus spp., 267—281 passim, 285-289 passim; L. passim; volume estimated from feces, 284—287 passim; volume
weberi (organ cactus, cardon) 138, 139, 224, 225, 231, 245; estimated from bones, 296—300 passim role in diet, 299—300,
L. hollianus, 232, 245; L. stellatus, 245 301, 307. See also Fauna; Hunting
Leopold, A. S„ 151, 152, 157 Medellin Zenil, Alfonso, 12
Lerma points, 304 Mendoza, Antonio de, Viceroy, 203
Leucaena spp. (guaje): L. esculenta, 221, 232, 234, 241, 251, Mequitongo dams, 84—85, 227
291; L. pueblana, 241 Merchants, 129
Levi Rock Shelter, Texas, 145, 171 Mesa Central, 34, 35, 195
Lice, 274, 277 Mesa del Sur, 35
Lime, 61 Mesoamerica: com agriculture in, 3, 14, 178-200 passim; cli-
Limestone, 70, 73, 84, 135, 146, 224 mate, 49-51; corn-bean relationship, 202, 203, 209; cotton
Linguistic groups, 39 varieties, 258; agriculture and ecology, 307-308; irrigation,
Lintels, 124,129 308, 309
Lip plugs, 100 Mesozoic, 67—68, 80
Lister, Robert H., 4 Mesquite ( Prosopis juliflora ): local distribution, 43-44, 64, 132,
Lizards, 158, 163-168 passim, 171-174 passim; homed
159, 135, 137, 221; remains in caves, 225, 227, 229, 231-234
(Phrynosoma spp.) 134; race-runner ( Ameiva undulata) 134- passim described, 241; question of cultivation, 251, 254, 291,
139 passim, 154, 157-168 passim, 171 292, 293; remains in feces, 266, 272, 274, 277, 278, 285, 286,
Loin cloth, 129 287; seed preparation, 303
Looms, 129, 294 Mestizos, 42, 43
Lorenzo, Jose Luis, 5, 7, 8, 12, 307 Metates, 21, 23, 126, 128, 288, 302
Lovelock Cave, Nevada, 264 Mexican language, 43
Lowe, Gareth, 12 Mexican Manuscript Number 20, 114
Lowland cultures, 308 Mexico, Valley of, 35, 112, 114, 115, 171, 203, 308; fossil com
pollen from, 4, 14, 180
MacNeish, Richard S., 41, 42, 256, 285 Miahuatepec formation, 68, 70
Macrobands, 172, 173 Miahuatlan, Puebla, 41, 43, 45
Macro-Mixteca languages, 39 Mice, 171, 264, 270-278 passim, 281, 285. See also Mouse
Madeira, 202 Miehoacan, 259
Maguey, see Agave Microbands, 165
Maize (Zea mays), 18-25 passim, 38, 42, 62, 136, 166, 167, 172, Milling stones, 21, 23, 302
225-234 passim; origin and distribution, 3-5, 14, 15, 178-200 Mimosa sp., 241, 286
passim, 209, 293-294, 295; and farming practices, 62, 306, Miranda, F., 61-62, 64
307; grinding and cooking, 126, 302, 303; role in diet, 173, Mites, 265
197, 202—203, 249, 254, 297, 308; remains in feces, 266, 268, Mitla, Oaxaca, 4, 307, 308
272-281 passim, 302 Mixteca Alta, 25, 35, 36, 38, 51-52, 55, 58, 126
races: wild, 178, 179-181, 193, 194, 198-200, 293; early Mixteca-Puebla culture, 24, 27
cultivated, 181, 194, 293—294; early tripsacoid, 182-189 pas- Mixtec, 38, 39, 42; codices, 114, 126, 127
sim, 194, 195, 293; wild-type segregates, 184, 189; Nal-Tel Moctezuma II, 41, 203
and Chapalote, 188-189, 194, 195, 294, 295; late tripsacoid, Modes, artifact, 17
189-190, 197; other 294—295; modem, 197-198
late, 194, 195, Molcajetes, 21, 303
botanical characteristics: 179-184 passim, 188-
of cobs, Moles, 171
190, 194—200 passim; of husks, 182-183, 184, 190; of tassels, Monte Alban, 14, 24, 27, 115, 117
183, 191, 193—194; of leaf sheaths and leaves, 185, 190-191; Moore, H. E., 213
of roots, 185-187; of kernels, 188; of stalks and shanks, 190 Morelos, 308
Malphigia sp. (nanche), 221, 242 Mortars, 21, 268, 270, 279, 288, 302
Mammillaria sp., 245 Motolinia (Fray Toribio de Benavente), 114, 127
Mammoth, 80, 171 Mountain lion ( Felis concolor), 139, 151, 159, 168, 296
Mammoth Cave, Kentucky, 264, 294 Mouse: harvest (Reithrodontomys fulvescens), 132, 134, 138,
Manihot, 272, 273, 274, 278, 286; M. esculenta, 286, 288, 290, 148, 156, 280; spiny Lionuys irroratus), 132-141 passim, 148,
(
294; M. dulcis, 294 149, 152, 156, 175; pigmy Baiomys musculus), 138, 148-149,
(
Manos, 21, 126, 128, 288, 302 156, 265

Manuel Avila Camacho Dam, 35 deer ( Peromyscus spp.), 132-139 passim, 148, 149, 156,
Markets, 36, 41, 218 171-172, 265; as climate indicators, 140—144 passim ; P. me-
Marl, 84, 89, 229 lanophrys, 139, 148, 149, 175, 176; P. leucopus, 148
Martin, Paul Schultz, 64, 265 kangaroo ( Dipodomys phillipsii): local distribution, 132,
Matamoros 304
points, 135-144 passim, 156, 169; described, 148, 149; hair identified,
Matzitzi formation, 68 264, 279, 280, 289
Maya, 14, 114, 126, 308, 309 Mullers, 21, 302
Mazapetan, Puebla, 43 Murine opossums, see Opossums
Mazatec, 39, 45 Museo Nacional de Antropologia, Mexico, 12, 33
Meat: sources, 132-139 passim, 304-305; preparation, 159, 162, Museo de la Revolution, Puebla, 12, 157
326
INDEX
Myrtillocactus geometrizans, 224, 229, 231, 245 274-279 passim, 285-289 passim; agricultural techniques,
294, 306, 307; sustenance trends, 300; food preparation, 303
Nahua, 39 Panama, 202
Nahuat language, 39 Pangalo, K. I., 213
Nahuatl language, 114 Papaloapan Commission, 25, 27, 35, 49, 55, 80
Nal-Tel (maize), 3, 178, 180, 194, 195, 197; described, 188-189; Papaloapan watershed, 55, 58-59, 61, 64, 66, 82, 83
distribution, 294 Papaya trees, 63, 64
Necklaces, shell, 128, 130 Parching, of seeds, 209, 270
Nets and netting, 22, 23, 127, 304 Paredes Colin, Joaquin, 114
Newt Kash Hollow Shelter, Kentucky, 265 Paso y Troncoso, Francisco del, 114
Nicholson, Henry B., 114, 115 Peanuts, 24, 126, 234, 241, 290, 295
Nighthawk ( Chordeiles acutipennis), 134, 135, 138, 138, 156 Peccary ( Pecari tajaeu): local distribution, 134—139 passim; de-
Nightjar ( caprimulgus ridgwayi), 134, 135, 138, 139, 156, 176- cribed, 152; remains at sites, 159, 163, 166—169 passim and
177 hunting patterns, 172, 173, 174; hair identified, 264, 276, 278,
Nitrates,43 285, 289; meat yield, 296
Noguera, Eduardo, 114, 119 Pectis canescens, 231
Nomadism, 23 Pelona points, 304
Nose plugs, 130 Penafiel bottling plant, 137
Pendants, 128, 130
Oaks, 239. See also Forests Periods, archaeological, 18—19
Oaxaca, 66—71 passim, 74, 114; climate, 49, 52, 60, 62;
4, 25, 35, Periostitis, 101, 109
pottery, 115, 117; ball courts, 127; agriculture, 213, 215, 293 Perishable artifacts, 12, 128-130
Oaxaca-Puebla Trough, 34, 35, 54, 60 Peru, 254-255, 265; beans and maize in, 202, 203, 204, 293, 294;
Oaxaca, Valley of, 39 squash in, 214, 219; other plants from, 252, 294
Obsidian artifacts, 24, 129, 162, 163 Pestles, 21, 268, 270, 279, 302
Ocampo, Tamaulipas, 203, 205, 212-219 passim, 265, 287 Petate, 129
Ocean currents, 48 Phaseolus, see Beans
Ocos, Guatemala, 294 Phases, cultural, 14, 18, 19, 22, 23—24
Olmecs, 14 Phillips, Allen, 162
Olotillo (maize), 197 Phillips, L. L., 259
Opossums, 137, 159, 163, 167, 168, 174, 296; Didelphis mar- Physalis sp. (ground cherry), 227, 252, 290; described, 248;
supialis, 134, 138, 139, 145, 175; murine (Marmosa canescens), remains in feces, 272, 273, 274, 278, 288, 294; preparation,
134, 138, 139, 145, 156, 170 289

Opuntia spp. (prickly pear cactus): local distribution, 127, 137, Pineapples (?), 275-280 passim, 287, 290, 295
138, 139, 221, 224; remains in caves, 225, 227, 229, 231-234 Pines, see Forests
passim; described, 245—246; question of cultivation, 250-251, Piper, C. V., 201
254, 288, 291, 292, 293; remains in feces, 265—270 passim, Pipes, clay, 120
276-279 passim, 285, 286, 288; preparation, 289, 302; O. Pira Naranja (maize), 186, 190, 195
hyptiacantha and O. pilifera, 231 Pit houses, 11, 23
Orchards, see Fruit Pits, 16, 17, 32
Organ cactus, see Lemaireocereus weberi Plainview points, 304
Orizaba, Peak of, 34, 35, 64 Plant remains: and chronology, 18, 295-296; and domestication,
Ornaments, 100, 128, 130, 169 19, 292-295; and subsistence, 19, 30.5—306; and diet, 21, 299-
Osteoarthritis, 94, 95, 96, 97, 102 302, 307; from Tehuacan caves, 220, 225-234; species de-
Osteophytosis, 94—101 passim, 109 scribed, 234-248
Otlaltepec, Puebla, 67 Platforms, 123
Owls, 135, 169, 170; barn ( Tyto alba), 138, 139, 140, 155-156, Pleistocene, 83; climate and rainfall, 64—65, 136, 139, 144; fauna
162, 168; elf ( Micrathene whitneyi), 138 and hunting patterns, 140—144, 152, 158, 171, 172
Oxalis spp., 231 Pliocene, 73
Plowing, 44
Pacific Ocean, 50 Plumeria rubra, 247
Paddock, John, 12, 119, 126 Pochote ( Ceiba parvifolia ): local distribution, 138; pod and root
Painter, Joseph H., 216 remains in caves, 225-234 passim; uses, 224; described, 244—
Paleo-Indian period, 158, 172. See also Ajuereado phase 245; root remains in feces, 266-280 passim, 284—287 passim;
Paleozoic, 67 preparation, 288-289, 302
Palerm, Angel, 214 Pods, collecting of, 305, 306
Palm, see Brahea dulcis Poinsettias, 36
Palmillas phase, 205 Points, projectile, 21, 32, 127, 162, 171; and types of hunting,
Palo Blanco phase: and Classic period, 19; salt industry, 23; 304-305
described, 24, 163; human skeletons, 100-109 passim; fauna, Pollen, 4, 6, 10, 14-19 passim, 265
145, 155, 163—169 passim, 175; hunting patterns and sub- Polio (maize), 180, 186, 190
sistence, 174, 304, 305; cultivated plants, 205, 208, 210, 216, Pomegranates, 43
217, 229-230, 234, 295; dietary evidence from feces, 266, Ponchos, 129
327
Popcorn, 195 passim, 162, 168, 169; described, 151; trapping of, 173; hair
Popocateptl, 64 identified, 264; meat yield, 296
Popolocas, 39, 125—126 Radiocarbon dates and dating, 3—11 passim, 18
Population, in Tehuacan Valley, 12, 22, 23, 24, 41-42, 43 Rainfall, 4, 19, 25, 27, 48-65 passim, 135, 178
Postclassic period, 5, 41—45 passim, 87, 112, 135, 174; pottery of, Rats, 296
12, 25, 27; defined, 19. See also Venta Salada phase wood Neotoma
( spp.): distribution, 135, 136, 139, 149, 156,
Pottery: in Tehuacan sequence, 24; helps define region, 25, 27; 169, and climate, 140—144 passim;
171; Allen’s ( N . alleni),
Codex Borgia relationships, 114, 117—123, 126 138, 149; Mexican (N. mexicana), 149
Prairie dogs, 158 cotton ( Sigmodon hispidus), 138, 139, 156, 169, 171; and
Preservation, food, 21, 297, 299-300 climate, 140-144 passim-, described, 149
Prickly pear cactus, see Opuntia Rattle, gourd, 129
Priests, 24 Ravens ( Corvus corax), 134, 156, 163, 168, 169, 173, 177
Proboscidian, 80, 83 Red Beds formation, 68
Prosopis juliflora, see Mesquite Red Conglomerates, 71-72, 82, 88
Pseudoscorpions, 277 Reeds, 134
Psidium guajaba, see Guava Refuse, 5, 1.5-16, 17, 21, 23
Puebla (state), 34, 35, 67-68, 74; climate, 49, 52; distribution of Relaciones geograficas, 23, 63, 114
squash, 213, 215 Religion, 23, 24
Puebla, Puebla, 25 Reptiles, 134, 135, 136, 145, 152-154, 176
Puebla Basin, 35 Rhodes, A. M., 213
Puebla-Oaxaca Trough, 34, 35, 54, 60 Ring-tailed cats ( Bassariscus astutus ), 176; local distribution,
Puebla, Valley of, 152 138, 139, 140, 158; described, 150; hair identified, 264, 278,
Pueblo Indians, 218 281, 285, 289
Pueblo Nuevo, Puebla, 70 Rio Atoyac Poblano, 35, 80
Pueblo Nuevo Dam, 154 Rio Blanco, valley, 34
Pueblos despoblados, 62—63 Rio Calapilla, 38; valley, 132, 152; arroyo, 134-135, 153
Puerto Marques, Guerrero, 308 Rio Comulco, 36
Pulque, 35, 63, 126, 250 Rio Grande de Cuicatlan, 35, 38, 46, 62, 74
Pumas (Felis concolor ), 139, 151, 159, 168, 296 Rio Hondo, 38, 74; valley, 67
Punin Calvarium, 112 Rio Ixcatlan, 62, 63
Purron Cave (Tc 272): stratigraphy, 16, 18; location, 37, 45, Rio Papaloapan, 35, 74, 83; watershed, 55, 58-59, 61, 64, 66, 82,
227-228; climate, 55, 64; burials, 93, 96-97, 100-101; faunal 83
remains, 163-165; plant remains, 194, 195, 213, 217, 228- Rio de Parian, 74
230; mentioned, 8, 9, 12, 28 Rio Salado: described, 35, 36, 38, 44, 45-46, 55, 66, 74; area
Purron phase, 18, 19, 24, 97, 165, 294, 300 fauna, 134-135, 145, 147, 152-155 passim, 168, 172, 173, 176;
Pyramids, 16, 24, 37, 45, 123 arroyo, 158, 168; earlier level, 159
Rio San Martin, 37
Quachilco site (Tr 218), 10, 12; stratigraphy, 16; burials, 98-100; Rio San Pedro, 38, 62
fauna and hunting patterns, 147, 168, 169 Rio Santo Domingo, 35, 66, 74, 87, 153
Quail: late Pleistocene, 140, 158 Rio Seco, 63
bobwhite ( Colinus virginianus ): local distribution, 134-139 Rio Tehuacan, 35—36, 74
passim, 158, 159, 165-169 passim; described, 154—155; snared Rio Teotitlan, 37, 45
or trapped, 171, 172, 174; diet of, 176 Rio Tilapa, 37
Quarries, 24 Rio Tomellin, 35; canyon, 74
Quaternary, 66, 72-77 passim, 80, 81 Rio Xiquila, 38, 62
Quebrachales, 134, 150, 155, 158 Rio Zape Cave, Durango, 208
Quecholac, Puebla, 74 Rio Zapotitlan: described, 38, 46, 61, 84; sites and towns, 45; ar-
Quercus sp., 239. See also Forests royo fauna, 134-135, 145, 150, 153
Quetzalcoatl, 130 Roasting and charring: of meat, 158—165 passim, 168, 302, 303;
Quids, vegetal, 21, 194, 235, 238, 250, 299 of plant materials, 204, 214, 216, 267-281 passim, 288, 289,
Quiotepec, Oaxaca, 41, 42, 43, 58, 60, 74 302, 330
Roberts, L. M., 2
Rabbits, 171, 277, 281, 296, 305; hunting (drives) of, 172 Robertson, Donald, 115
Audubon cottontail (S. audubonii ): local distribution, 132- Rocks, metamorphic, 67
139 passim, 147, 158, 162-167 passim; described, 146; trap- Rock shelters, see Caves
ping of, 171-174 passim; diet of, 175; hairs identified, 264, Rodents, 10, 135, 175-176, 216; local distribution, 137, 145,
269, 276-281 passim, 289 167, 168, 169-170; as climate indicators, 140—145, 169-170
Mexican cottontail (S. cunicularius): local distribution, 132- Rogers, David, 286
139 passim, 158, 162-167 passim described, 146—147; trap- Roman y Zamora, Fray Jeronimo, 114
ping of, 171-174 passim; diet of, 175; hair identified, 264, 278, Romano, Arturo, 4, 171
279, 280, 289. Roofs, 125, 126, 129, 130

See also Jack rabbits Roots, collecting of, 305, 306
Raccoons ( Procijon lotor ), 171, 176; local distribution, 135-139 Rose, J.
N., 216
328
INDEX
Rose, J. S„ 216 Sideroxylon cf. tempisque (cosahuico): local distribution, 138,

Rye, 214 225-234 passim; described, 246—247; question of cultivation,
251-252, 254, 291-292, 295
Sacrifice: human, 23, 129; animal, 127 Sierra de Acultzingo, 35, 66, 67
Sahagiin, Fray Bernardino de, 114 Sierra de Atzingo, 67, 68, 70, 84
Salado Valley and watershed, 58, 59 Sierra de Huautla, 34
Salas, G. P., 67, 72 Sierra Madre de Oaxaca: boundary of Tehuacan Valley, 25, 34,
Salinas de Barranca, Puebla, 46, 58, 59, 60 66; described, 34, 35, 36, 46; climate, 50-55 passim, 58, 59
135
Salt production, 23, 24, 46, Sierra Madre 34
Oriental,
SaltsCave, Kentucky, 264, 265 Sierra Madre 85
del Sur,
San Andres, Puebla, 8 Sierra Mazateca, 35, 66, 67
San Antonio las Canadas, Puebla, 36 Sierra de Miahuatepec, 67, 70
San Bernardino lagoon, 73 Sierra Mixteca, see Mixteca Alta
San Juan sin Agua, San Luis Potosi, 65, 144 Sierra de San Bartolo, 70
San Juan Raya, Oaxaca, 68 Sierra de San Felipe de las Vaderas, 36
San Juan Raya formation, 68, 70 Sierra de Santa Rosa, 68
San Lorenzo springs and bottling plant, 137 Sierra de Tlacotepec, 34
San Marcos arroyo, 84 Sierra de Zapotitlan, 35, 66, 67
San Marcos Cave (Tc 254), 12, 28; stratigraphy, 16, 18; plant Sierra de Zongolica, 25, 34—35, 66, 77, 205
remains, 136, 172, 179, 213, 217, 227, 260; faunal remains, Simpson, L. B., 22, 41, 42
136-137, 153, 166—167, 170; ecology, 136, 226; subsistence, Sites, in Tehuacan project, 22, 27-28, 32. See also names of sites
227 Skeletons, human, 19, 21, 91—113
San Marcos bill, 73, 89 Skinning, of meat, 274, 28.5-286, 289
San Pablo, Valley of, 43 Skirts, 129
Sandals, 22, 24, 129-130 Skulls, human, 93—99 passim, 103—107, 110-113
Santa Isabel Iztapan, Mexico, 171 Skunks, 140, 159, 163, 167, 168, 171-174 passim, 296; hooded
Santa Maria Astahuacan, Mexico, skull, 111 or striped ( Mephitis macroura ), 132—140 passim, 151, 158,
Santa Maria del Monte, Puebla, 74 159, 165-170 passim, 176; spotted ( Spilogale augustifrons),
Santa Maria phase, 18, 22, 23; and Formative period, 19; de- 136-139 passim, 151, 158, 159, 162—167 passim; hog-nosed
scribed, 24, 98; burials, 98-100; fauna, 132, 151, 163, 165, ( Conepatus mesoleucus) 138, 139, 140, 151, 158, 162
,
169, 173; plants, 195, 205, 213, 217, 228, 230, 233-234; die- Slaves, 286
tary evidence from feces, 266, 272-274, 28.5-288 passim; agri- Slender pop (maize), 195, 197, 198, 294
cultural innovations, 294-295; subsistence, 300, 30.3-306 Smallpox, 42
passim Snails, 19, 171, 265
Santa Marta Cave, Chiapas, 4, 15, 307 Snakes, 158, 166, 167, 171, 174, 265
Sanvitalia fruticosa, 231 Snares and snaring, 172, 173, 304
Sapotes, 24, 63, 64, 126, 254. See also Casimiroa; Diospyros Social organization, 23, 24
Sauer, Carl O., 213 Soil types, Tehuacan Valley, 132, 136, 137, 147, 224
Sauer, Jonathan, 213 Sonora, 4, 14, 293
Sauk Valley, skull. 111, 112 Sonoran Desert, 201, 210
Sears, P. B., 4 Sosola, Oaxaca, 74
Sea urchin, 265 Sotomayor, Arturo, 12, 22, 87, 88, 90
Secretaria de Recursos Hidraulicos, 80 South America, 3, 14, 212, 218, 219, 254, 292
Sedimentation, 77, 82, 83, 85 Spain, 39
Seeds, 43, 210; characteristics, 203-204, 205, 208, 232; identified Spaniards, 42, 43
in feces, 261-265 passim, 268, 274, 277, 278, 279; role in diet, Spanish language, 39, 114
287, 299; preparation, 302, 303; collecting of, 305-306 Spear points, 171, 304
Selaginella sp., 231, 235 Spikes, flint, 127
Seler, Eduard, 114, 115, 119, 129 Spindle whorls, 129, 294
Semenov, S. A., 22 Spondius mombin ( ciruela ) : local distribution, 137, 138, 225,
Seminole, 202 230-234 passim, 290; question of cultivation, 230, 251, 254,
Servicio Meteorologico of Mexico, 49 291, 292, 295; described, 243
Setaria cf. macrostachya: local distribution, 225, 227, 231, 232; Spoons, 129
role in diet, 233, 249, 254, 287; described, 236; identified in Springs: and bottling plants, 38, 137; in Tehuacan Valley, 43,
feces,265—269 passim, 272—279 passim, preparation and use, 46, 81, 167; Quaternary deposits by, 75, 77; and salt industry,
288, 302; question of cultivation, 249, 287, 291, 292, 293 135
Settlement patterns, 22, 23, 24, 27, 44 Squash ( Cucurbit a spp.), 23, 24, 173, 250, 308; preparation,
Sewing, 129 214, 271, 302, 303; identified in feces, 270-281 passim, 285,
Sheep, 38, 43, 61, 62, 170 286, 287
Shellfish, collecting of, 307, 308 C. mixta, 212, 228, 232, 290; described, 213—214; cultiva-
Shells, 10, 17; artifacts of, 12, 17, 22, 128 tion and distribution, 216, 218-219, 253, 254, 293
Shields, 117, 129 C. moschata, 212, 217, 232; described, 214; cultivation and
Shoshone, 171, 172 distribution, 219, 254, 293
329
C. pepo, 212, 217, 225, 232, 290, 293; described, 214-216; 15, 130, 290; field and laboratory techniques, 25—33 passim;
cultivation and distribution, 219, 253, 294 mentioned passim
other species, 212, 213, 215 Tehuacan bottling plant, 137
See also Apodanthera; Crescentia; Lagenaria Tehuacan-Cuicatlan Basin, 66, 71, 74—75, 77, 80, 83
Squier, Robert, 12 Tehuacan formation, 72, 73
Squirrels: small ground, 140, 146, 147, 158, 171, 172, 174 Tehuacan Incised pottery, 117, 120, 123
gray rock ( Citellus spp. ), 175; local distribution, 135, 136, Tehuacan plain, 46, 137
137, 140, 158, 167, 168"; described, 147 Tehuacan Valley: mentioned passim; area defined, 25, 27, 66;
Starfish, 265 described, 34—46
Starr points, 304 Tehuacan Xantiles pottery, 119, 123
Stature, defined by human skeletons, 94—101 passim, 107-108, Temalacatl (gladiatorial stone), 117
112 Temperatures, 19, 49, 55
Steaming, of meat, 302 Temples, 123—126
Stew, bone or marrow, 302 Teopoxco, Oaxaca, 74
Steward, Julian, 23, 171, 172 Teosinte (Euchlaena mexicana ), 3, 14, 180, 183, 189
Sticks, whittled, 171 Teotihuacan, Mexico, 117
Stomach contents, analysis of animal, 175—177 Teotihuacan points, 304
Stone artifacts, 11, 12, 17, 159, 172 Teotipilco, Puebla, 82
Storms, 50—51, 52, 55, 58, 59 Teotitlan (San Pedro), Puebla, cf. San Pedro Tetitlan, 43
Stratigraphy, of Tehuacan sites, 15—17, 18 Teotitlan del Camino, Oaxaca: population, 41; described, 45;
Subsistence: and Tehuacan project, 6, 7, 19—24 passim, 261, and Tehuacan project, 46; climate, 51-59 passim; geology,
290; in Tehuacan sequence, 23, 24, 92-100 passim, 126—130; 70-71, 82; and documentary sources, 114, 115, 126; men-
and food preparation, 302—304; activities described by phase, tioned, 24, 39, 66, 119
304—307; developments in Tehuacan and Mesoamerica, com- Tepanco de Lopez, Puebla, 39, 41, 66, 71, 74, 82
pared, 307—309. See also Agriculture; Gathering; Hunting; Tepeaca, Puebla, 35, 42; Province and Alcaldia Mayor, 43
Sustenance Tepecintle (maize), 197, 295
Sugar, 214; cane, 36, 45, 63, 80, 214; mill, 37, 45 Tepelmeme, Oaxaca, 55
Survey, archaeological, 27—28 Tepetate, 61. See also Caliche
Sustenance, 19, 21, 126; means of estimating, 296—299; esti- Tepetzingo, Puebla, 44
mated volume by phase, 299-302; proportions from various Tepexpan Man, 110-111
subsistence activities, 302—307. See also Diet Tepostitlan, Puebla, 68
Swallow Cave, Chihuahua, 178 Tequixtepec, Oaxaca, 5
Syphilis, 99, 109 Termites, 265
Tetecho, see Cephalocereus hoppenstedtii
Tetzoyoean, Puebla, 74
Tabasco, 12, 14 Texmelincan, 117
Tamaulipas: search for corn origins in, 3, 4, 14; plant remains, Textiles, 22, 24, 32, 43, 129
203, 204, 205, 212, 216-219 passim, 253; dietary evidence Tezcatlipoca, 115
from feces, 265, 267, 287; plant distribution, 292—295 passim-, Thevetia peruviana, 225, 247
agricultural sequence, 307—308; subsistence, 307, 308, 309 Thornthwaite, C. W., 54
Tamayo, Jorge L., 34, 39, 43 Thread, cotton, 24
Taos group, Cucurbita seed types, 213 Ticks, 269, 277
Tapir, 171 Tilapa, Puebla, 24, 45, 58, 66, 70
Taxonomic method, 17, 18 Tilapa plain, 163
Tecamachalco (Tepemaxalco), Puebla, 35, 74, 80; climate, 25, Tile and brick works, 44
54, 58, 74, 80; population, 41, 42; described by Alcedo, 43 Tillandsia spp., 225, 237; T. recurvata and T. usneoides (Span-
Tecamachalco Valley, 35 ish moss), 237
Techniques, field and laboratory, 5, 12, 15, 27-33 passim Tizatlan, Tlaxcala, 115
Tecomavaca, Oaxaca, 25, 41, 45, 46 Tlacotepec de Diaz, Puebla, 25, 35, 39, 41, 43, 54
Tecorral Canyon: sites, 8; animals and ecology, 136—137, 147- Tlacoxcalco, Puebla, 45
155 passim, 166—167; subsistence in, 172, 173, 174; maize, Tlaloc, 119, 129
178-194 passim; plants and ecology, 226-227 Tlaxcala, 35, 115, 126; Senorio de, 35
Tecorral Cave (Tc 255), 28; stratigraphy, 16, 18; faunal re- Tonatiuh, 129
mains, 149, 150, 153, 166—167; plant remains, 213, 217, 227; Torquemada, Fray Toribio de, 114
ecology, 226 Tortillas, 273n, 276n, 303
Teeth: animal, 157, 162, 163, 165, 169 Tortoise, see Turtles
human: described, 93-103 passim; and changes in diet, 96, Towns, see Villages
98, 10.5-107; pathology of, 107 Trade, 23, 24, 39-43 passim, 46
Tehuacan, Puebla: described, 34-46 passim; climate, 48, 51-58; Trade winds, 48—55 passim, 64
geology, 66, 73, 74, 80, 82; and Codex Borgia, 115; mammoth Traps and trapping, 23, 92, 162—167 passim, 171—175 passim,
remains at, 171; plants sold in markets, 210, 213-216 passim; 304, 305
mentioned, 5, 12, 24, 25, 117 Travertine, 135, 136, 137, 224, 225
Tehuacan Archaeological-Botanical Project: described, 3-13, 14, Treponematosis, 99, 109
330
INDEX
Triassic, 67 Viru Valley, Peru, 253

Tribute, 46, 63, 203; rolls, 41 Volcanic activity, 34, 80
Tridax procumbens, 231 Volcanoes, 73, 80
Tripsacoid maize: early, 182-189 passim, 194, 195, 293; late,
189-190, 197 Wallrath,Matthew W., 12

Tripsacum, 14-15, 180, 183-184, 189, 293. See also Maize Warren, Bruce, 12
Tubes, bone, 167, 169 Wasps, 265, 280
Tularosa Cave, New Mexico, 205 Water: supply, 19, 63, 221-225 passim, 252; vapor, 49; table,
Tule points, 304 61-64 passim, 75, 81, 82; effect of on teeth, 107
Tump lines, 129 Weapons, 19
Turkey ( Meleagris gallopavo), 24, 290; remains, 163, 168, 169, Weaving, 24
174; distribution, 155, 295; feathers in feces, 273, 274; meat Wellhausen, Edward 3, 10
J.,
yield, 296 Wheat, 35, 38, 43
Turtles, 144, 166-173 passim-. Pleistocene, 147, 153, 166, 172; Whippoorwill, see Nightjars
mud Kinosternon integrum), 134, 135, 152—153, 159, 167-
( White, Theodore, 157, 171
173 passim, 176; gopher ( Gopherus sp. ), 140, 174; Texas Whitewater Draw, Arizona, 144
gopher (G. berlandieri) 153 Wind, 49, 64
Tuxpeno (maize), 197 Wolf, 171
Tuxtla, Chiapas, 259 Wooden artifacts, 22, 128—129
Typology, artifact, 17, 18 Wood-working, 128-129
Wool, 43
United States: southeastern, 3, 203, 208; southwestern, 64, 201-
218 passim, 265, 292-295 passim; eastern 293
Xaco, Puebla, 11, 44
Universidad Nacional Autonoma de Mexico, Instituto de Bio-
Xantiles, 23, 119
logia, faunal bone collection, 157
Xiquila, Oaxaca, 59, 60
Urban centers, see Cities
Xochichalco, Morelos, 308
Xochipilli, 119
Vallesia glabra, 134
Xochiquetzal, 120
Valley of Mexico, see Mexico, Valley of
Xochitlan, Puebla, 74
Valsequillo, Puebla, 152, 157
Valsequillo reservoir, 35
Vanilla, 43 Yanhuitlan beds, 72
Vasquez, Luis, 4, 5 Yarmouth interglacial, 15
Vavilov, N. I., 212 Yarn, 129
Vegetation, see Flora Yucatan, 208
Venta Salada phase: and Postclassic period, 19; population, 22;
Yucatan Indians, 202
salt industry, 23; described, 24; burials, 101-103; andCodex Yucca periculosa (izote), 135, 137, 225-229 passim, 238, 268,
Borgia, 114-130 passim; pottery types, 117-123; animals and 275, 276
hunting patterns, 145, 147, 163, 167, 168, 169, 304, 305; agri-
cultural plants, 203, 205, 213, 216, 217, 226, 227, 234, 295; Zacatenco, Mexico, 14
dietary evidence from feces, 266, 279-289 passim; subsistence, Zacatenco points, 304
300-307 passim Zapalote Chico (maize), 197, 294
Venta Salada plain, 58, 82, 134, 135, 146, 148 Zapalote Grande (maize), 295
Venta Salada site (Tr 75), 45, 115, 117 Zapotecs, 43, 114
Venus cycle, cult, 126 Zapotitlan formation, 68
Veracruz, 34, 35, 36, 55, 153; pottery, 24, 115, 117, 119; agri- Zapotitlan Salinas, Puebla, 39-43 passim, 46, 68, 74; rainfall
culture, 214, 215, 308, 309 data, 51, 54, 58, 59
Veranillo, 51 Zapotitlan Valley, 25, 46, 58, 66, 68
Verde Valley, U.S., 208 Zinacatepec, Puebla, 41, 45, 70
Viejon, Veracruz, 308 Zizyphus pedunculatus (cholulo), 231, 243, 286
Villa R., Bernardo, 145, 157, 17] Zones (excavated occupations), 15-17, 18
Villages: in Tehuacan sequence, 23, 24, 98, 100, 173; pit-house, Zoquitlan, Puebla, 43, 74
23; wattle-and-daub, 24; in Tehuacan Valley, 41-45 passim Zozutla, Puebla, 74
331
(Continued from front flap)
by the inhabitants of the valley. Dr. Man-

gelsdorf contributes an analysis of the ar-
chaeological corn discovered by the Project.
Hugh C. Cutler, Thomas W. Whitaker,
Lawrence Kaplan, Stanley G. Stephens, and
C. Earle Smith, Jr., discuss squashes and
pumpkins, beans, cotton, and other plants
brought under cultivation or used by man.
An inventory of resident species of wild ani-
mals by Kent V. Flannery and a discussion
of seasons when they were hunted, cast light
on habits of the people.
The ancient inhabitants, represented by
distressingly few skeletons, are discussed by
Dr. James E. Anderson. New light on their
food habits is shed by Eric O. Callen. Geo-
logical studies by Jean Brunet suggest rea-
sons for the progressive dessication of the
valley. A comparison of Codex Borgia with
archaeological discoveries points to the pos-
most colorful of aboriginal
sibility that this
Mexican books came from the vicinity of the

Tehuacan Valley.
Future volumes will cover other aspects
of the Project: Volume II, non-ceramic arti-
facts; Volume III, a study of pottery from

simplest beginnings to sophisticated vessels
of the sixteenth century; Volume IV, a chron-
ology of the Tehuacan Valley; Volume V, a
description of the excavations and problems
encountered, as well as ancient irrigation
works; Volume VI, a summary of the Project.
The entire series is edited at the R. S. Pea-
body Foundation for Archaeology, Phillips
Academy, Andover, Massachusetts, under
the direction of Douglas S. Byers, director of
the Foundation.
UNIVERSITY OF TEXAS PRESS

AUSTIN & LONDON
Printed in U.S.A.
THE AUTHORS
James E. Anderson, M.D., is Professor of Physical An- Walton O. Galinat is Research Fellow, Bussey Insti-
thropology and Paleopathology at the University of tution, Harvard University.
Toronto.
Lawrence Kaplan is Professor of Botany, University
Jean Brunet is a geologist on the staff of the Centre of Massachusetts, Boston.
Scientifique et Technique Franfais au Mexique.
Richard S. MacNeish, Director of the Tehuacan Proj-
Douglas S. Byers is Director, Robert Peabody Foun-
S. ect, is now Head of the Department of Archaeology,
dation for Archaeology, Phillips Academy, Andover, University of Calgary, Calgary, Alberta.
Massachusetts.
Paul C. Mangelsdorf is Fisher Professor of Natural
Eric O. Callen is Associate Professor, Department of History, Harvard University.
Plant Pathology, MacDonald College, McGill Uni-
versity. C. Earle Smith, Jr., is Research Botanist, United States
Department of Agriculture.
Robert Chadwick is now an archaeologist with the
Archaeological Salvage Program for the Rio Balsas. Stanley G. Stephens is Professor of Genetics, Institute
of Biological Sciences, School of Agriculture and Life
Hugh C. CutlerExecutive Director of the Missouri
is Sciences, North Carolina State University at Raleigh.
Botanical Gardens, St. Louis, Missouri.
Thomas W. Whitaker is Geneticist, Agricultural Re-
Kent V. Flannery is Associate Curator of Archaeology, search Service, United States Department of Agri-
Smithsonian Institution, United States National Mu- culture, United States Horticultural Field Station,
seum. * La Jolla, California.

Prehistory of Te Hua Can

Uploaded by

Copyright:

Available Formats

Prehistory of Te Hua Can

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prehistory of Te Hua Can

Uploaded by

Copyright:

Available Formats

•

R. S. MacNeish, Douglas S. Byers, and Others

hunting parties for traces of early com and

its course plant breeders set about improv-

is concerned principally with geography,

(Continued on back flap)

Environment and Subsistence

Edited by Douglas S. Byers

Published for the

ROBERT S. PEABODY FOUNDATION

UNIVERSITY OF TEXAS PRESS • AUSTIN & LONDON

Library of Congress Catalog Card Number 67-17873

The preparation and publication of

Typesetting by Service Typographers, Indianapolis, Indiana

13. A Cotton Boll Segment from Coxcatlan Cave 256

182 Cultivated Plants .... following 304

ENVIRONMENT AND SUBSISTENCE

time assistant manager of the Hotel Penafiel in Te-

to finish the first season in Coxcatlan Cave. All this March.

news some minor triumph might be followed by

he immediate objective Tehuacan Archae-

T was the establishment

plying an answer to this problem. we shall call them.

refuse designated Zone D 1 which covered sterile soil.

which there was charcoal. Next came Zone F 1 a hard- ,

tent. A definite floor (Zone Super-E or E 1

ENVIRONMENT AND SUBSISTENCE

Fig. 2. The interdisciplinary solution of archaeological problems.

ENVIRONMENT AND SUBSISTENCE

Barghoorn et al., 1954. Kidder et al, 1946.

Field and Laboratory Techniques

he more general problems dis-

^EL RIEGO CAVE

K- TECORRAL CAVE '

QUACHILCO ,LAS CANO AS

i ^ ( r ' -*Jk.O >XCATLAN TERRACE

- 'V. fa"" del Camino-

« Santa Maria lx cat lan'

Fig. 3. The Tehuacan Valley. Triangles indicate archaeological sites.

ideal pattern. First of from the preliminary survey

removed the isolated squares, standing out like but- —

If came from a surface collection, it was

The Region and Its People

THE REGION AND ITS PEOPLE

1548 1565 1568 1580 1595

population was drastic. The population of Mexico at

been in the neighborhood of 9,120,000 (C. A. Nieve,

500 families of Indians who concern themselves with

Quiotepec . . . has 42 families of Zapotec Indians who

Indians who speak Chocha and Mexican and support

102 of mestizos, and 481 of Indians. ... In the provincial

visible instrument other than the plow and oxen, so

To the west of Tehuacan, on the road to Zapotitlan,

Indians to whom the populace of Tehuacan offers a

Large haciendas formerly occupied the valley, and

Climate and Hydrology

T Rains occur there chiefly two periods of six

a sweater again becomes comfortable. from the oceans.

their early stages the storms usually move westward

tial downpours which can do severe damage, destroying