Sume

oi.uchicago.
edu
T H E ORIENTAL INSTITUTE OF THE UNIVERSITY OF CHICAGO
ASSYRIOLOGICAL STUDIES
JOHN ALBERT WILSON & THOMAS GEORGE ALLEN • EDITORS

oi.uchicago.edu
oi.uchicago.edu
THE S U M E R I A N KING L I S T
oi.uchicago.edu
oi.uchicago.edu
THE
SUMERIAN K I N G
LIST
BT THORKILD JACOBSEN
THE ORIENTAL INSTITUTE OF THE UNIVERSITY OF CHICAGO
ASSTRIOLOGICAL STUDIES . NO. 11
THE UNIVERSITY OF CHICAGO PRESS • CHICAGO • ILLINOIS
Internet publication of this work was made possible with the

generous support of Misty and Lewis Gruber
oi.uchicago.edu
International Standard Book Number: 0-226-62273-8

Library of Congress Catalog Card Number: 39-19328
THE UNIVERSITY OF CHICAGO PRESS, CHICAGO 60637
The University of Chicago Press, Ltd., London
© 1939 by The University of Chicago. All rights
reserved. Published 1939, Fourth Impression 1973.
Printed by Cushing-Malloy, Inc., Ann Arbor, Michigan,
United States of America.
oi.uchicago.edu
TO O. E. RAVN, TO THE MEMORY OF EDWARD CHIERA

AND TO H. FRANKFORT ARE THESE STUDIES DEDICATED
oi.uchicago.edu
oi.uchicago.edu
PREFACE
The incentive to the studies here presented was furnished by the excava-
tions of the Oriental Institute at Tell Asmar. When in the season of 1931/32
we opened up strata of Agade and Early Dynastic times, the chronology of
these periods naturally occupied our thoughts greatly, and the author felt
prompted to resume earlier, more perfunctory studies of the Sumerian King
List. The main ideas embodied in the present work took shape that season in
the evenings, after days spent in the houses and among the remains of the
periods with which the King List deals. The detailed working-out and re-
peated testing of these ideas have occupied much of the author's time in the
years since then. He releases them—although he feels that they will continue
to occupy his thoughts for a long time yet—in the sincere hope that they will
prove fruitful to other workers in thisfieldand contribute toward better under-
standing of the innumerable chronological problems which still await solution.
The author is indebted to many people for help and encouragement—first
of all to the three men to whom this book is dedicated. To the example of their
widely different but all truly scholarly personalities I owe much. O. E. Ravn,
my teacher, is the embodiment of his own sober concepts of what scholarship
should be and the best mentor a young Assyriologist could have. As for Ed-
ward Chiera—only those who had the good fortune to work with this warm-
hearted, vital, and inspiring scholar can fully realize how irreparable a loss our
science suffered at his untimely death in 1933. With H. Frankfort I have been
associated through ten years of work in the field and at home, fruitful years of
friendship and free exchange of ideas which I value highly. Toward the studies
here presented he has shown a never failing interest.
Sincere thanks are due, further, to the late James Henry Breasted for accept-
ing this book for the Oriental Institute series and especially to the present
director of the Institute, John A. Wilson, for making the fulfilment of that
promise possible although conditions have changed materially. John Wilson
also suggested comparing my results with Egyptian chronology and assisted
with the Egyptological literature on the subject.
To my colleagues in the Oriental Institute I am indebted for suggestions and
helpful criticism on many points. Several have undertaken to read the manu-
script completely or in part. I may mention Professors A. T. Olmstead and
F. W. Geers and Drs. George G. Cameron, Robert M. Engberg, C. W. Mc-
Ewan, Ignace J. Gelb, and Samuel I. Feigin.
ix
oi.uchicago.edu
X PREFACE
Of inestimable value has been the interest and painstaking care which Dr.
T. George Allen and his associates on the editorial staff of the Oriental Insti-
tute have shown in dealing with this volume. I wish especially to emphasize
my feeling of deep indebtedness to Mrs. Ruth S. Brookens, with whom every
point of both style and argument has been thoroughly discussed. Through her
unflagging interest, her fine scholarly approach, and excellent judgment this
essay has profited materially. The index at the end of the book is also due
to her.
THORKILD JACOBSEN
COPENHAGEN
April 14, 1939
oi.uchicago.edu
T A B L E OF C O N T E N T S
PAGE
LIST OF ABBREVIATIONS xiii
I. INTRODUCTION 1
I I . TEXTUAL PROBLEMS 5
The Individual Manuscripts 5
Derivation from a Single Original 13
Genealogy of the Manuscripts 14
The Postdiluvian Section 14
Evidence from the Principal Variants 14
The Atabba Variants 15
Variants in the Reigns of the Agade Rulers 23
Variants in the Formulas 28
Genealogical Table Based on the Foregoing Variants 42
Supplementary Evidence from Variants of More Limited Scope . . 50
The Fourth Dynasty of Uruk 50
The Interregnum in the Agade Period 52
The Ku(g)-Baba Variant 53
Influence from the Y Group on Sui, Su2, and Li 54
The Antediluvian Section 55
III. CRITICAL EDITION OF THE T E X T , WITH TRANSLATION AND NOTES . . . 69
IV. COMPOSITION 128

Date 128
The Collective i - b - LAL 129
The Disregard of Subject Element - n - and "Causative" - n - . . . 131
Traces of Early Redactions 135
The Term 8% u k u 1 sig 137
Basic Ideas in the King List 138
Conclusions 140
Place 141
Sources 141
General Character of the Sources 141
Sources of the "Notes" 142
Sources of the Material Which Constitutes the Main Body of the List . 147
Individual Sources 154
Methods 158
V. HISTORICAL VALUE 165
Relative Values of Features in the King List 165
Isolation and Reconstruction of the Sources 167
The Kish Source 167
xi
oi.uchicago.edu
xii TABLE OF CONTENTS

PAGE
The Uruk Source 169

The Ur Source 172
The Akshak Source 176
The Agade Source 177
Correlation of the Sources 178
Ur and Agade 178
Agade, Kish, and Uruk 178
Akshak and Kish 179
Uruk and Lagash 179
Table I. Tabulation of the Sources as Correlated facing 180
Comparisons with Data from Other Documents 180
"Missing Rulers" 180
Synchronisms 183
En-temena-Lugal-kinishe-dudu 183
En-shakush-Anna(k)-Enbi-Eshtar 183
En-shakush-Anna(k)-Elulu 184
A-Anne-pada-about Ur-Nanshe(k) 184
Gilgames-Aka 186
Types of Personal Names 187
VI. RESULTING CHRONOLOGY 191
Relevant Fixed Points of Absolute Chronology 191
Reckoning from Late Chronological Lists 191
Astronomical Dating of the Hammurabi Dynasty and of the Third Dy-
nasty of Ur 196
Connection of Datable Dynasties with Those in the King List Proper . . 202
The Reign of Utu-hegal 202
Gutium and Agade 204
Table II. Chronology of Mesopotamian Rulers down to the First Dynasty
of Babylon facing 208
INDEX 209
PLATE. CUNEIFORM TEXT OP J AND OF EPIC FRAGMENT K 11624 at end

oi.uchicago.edu
LIST OF ABBREVIATIONS
AH DELITZSCH, F. Assyrisches Handworterbuch (Leipzig, 1896).
AJSL American journal of Semitic languages and literatures (Chicago
etc., 1884 ).
An. Or. Analecta orientalia (Roma, 1931 ).
AO Paris. Mus6e national du Louvre. Antiquity orientales. (Fol-
lowed by catalogue number.)
AOF Archiv fur Orientforschung (Berlin, 1923 [Vols. I - I I called
"Archiv fur Keilsehriftforschung"]).
AS Chicago. University. Oriental Institute. Assyriological studies
(Chicago, 1931 ).
AS No. 6 JACOBSEN, THORKILD. Philological notes on Eshnunna and its
inscriptions (1934).
BE Pennsylvania. University. The Babylonian expedition of the
University of Pennsylvania. Series A: Cuneiform texts, ed. by
H. V. Hilprecht (Philadelphia, 1893-1914).
BE I 1-2 HILPRECHT, H. V. Old Babylonian inscriptions, chiefly from Nip-
pur (1893-96).
BE I I I 1 MYHRMAN, DAVID W. Sumerian administrative documents
(1910).
BE VI 2 POEBEL, ARNO. Babylonian legal and business documents . . . .
chiefly from Nippur (1909).
BE X X 1 HILPRECHT, H. V. Mathematical, metrological and chronologi-
cal tablets from the temple library of Nippur (1906).
BE X X X I LANGDON, S. H. Historical and religious texts from the temple
library of Nippur (1914).
BM British Museum. (Followed by catalogue number.)
Bu British Museum. Budge collection.
CBS Pennsylvania. University. University Museum. Catalogue of
the Babylonian section.
CR Acad6mie des inscriptions et belles-lettres, Paris. Comptes ren-
dus des stances (Paris, 1858 ).
CT British Museum. Cuneiform texts from Babylonian tablets, &c,
in the British Museum (London, 1896 ).
Dec. SARZEC, ERNEST DE. D^couvertes en Chald6e. II. Partie 6pi-
graphique et planches (Paris, 1884-1912).
EH A SMITH, SIDNEY. Early history of Assyria to 1000 B.C. (London,
1928).
GSG POEBEL, ARNO. Grundziige der sumerischen Grammatik (Ros-
tock, 1923).
HRETA NIES, J. B., and KEISER, C. E. Historical, religious, and econom-
ic texts and antiquities (Babylonian inscriptions in the collec-
tion of James B. Nies. II [New Haven, Conn., 1&20]).
oi.uchicago.edu
xiv L I S T O F ABBREVIATIONS
ITT Istanbul. Asan Atika Muzeleri. Inventaire des tablettes de Tello

conserves au Mus£e imperial ottoman (Paris, 1910-21).
ITT II 1-2 GENOUILLAC, HENRI DE. Textes de Pepoque d'Agade* et de Y6-
poque d'Ur (1910-11).
ITT IN DELAPORTE, LOUIS. Textes de P6poque d'Ur (1912).
JAOS American Oriental Society. Journal (Boston etc., 1849 ).
JRAS Royal Asiatic Society of Great Britain and Ireland. Journal
(London, 1834 ).
JSOR Society of Oriental Research. Journal (Chicago etc., 1917 ).
K British Museum. Kouyunjik collection.
KAHl MESSERSCHMIDT, LEOPOLD. Keilschrifttexte aus Assur his-
torischen Inhalts (WVDOG XVI [Leipzig, 1911]).
KAH I I SCHROEDER, OTTO. Keilschrifttexte aus Assur historischen In-
halts (WVDOG XXXVII [Leipzig, 1922]).
KAR EBELING, ERICH. Keilschrifttexte aus Assur religiosen Inhalts
(WVDOG X X V I I I 1 - 4 and XXXIV 1-5 [Leipzig, 1919 ]).
KAT* SCHRADER, EBERHARD. Die Keilinschriften und das Alte Testa-
ment (3. Aufl., . . . . neu bearb. von H. ZIMMERN und H.
WINCKLER; Berlin, 1903).
KAV SCHROEDER, OTTO. Keilschrifttexte aus Assur verschiedenen In-
halts (WVDOG XXXV [Leipzig, 1920]).
Ki British Museum. King collection.
MDP France. Delegation en Perse. Memoires (Paris, 1900 ).
MDP X X I I - X X I V SCHEIL, J. V. Actes juridiques susiens (1930-33).
MJ Pennsylvania. University. University museum. The Museum
journal (Philadelphia, 1910 ).
MVAG Vorderasiatisch-aegyptische Gesellschaft, Berlin. Mitteilungen
(Berlin, 1896-1908; Leipzig, 1909 ).
OECT Oxford editions of cuneiform texts (Vol. I : " . . . . inscriptions";
London, 1923 ).
OECT I I LANGDON, S. H . Historical inscriptions, containing principally
the chronological prism, W-B. 444 (1923).
OIP Chicago. University. Oriental Institute. Oriental Institute pub-
lications (Chicago, 1924 ).
OIP XLIII FRANKFORT, H., LLOYD, S., and JACOBSEN, T. The Gimilsin
temple and the palace of the rulers at Tell Asmar (in press).
OLZ Orientalistische Literaturzeitung (Berlin, 1898-1908; Leipzig,
1909 ).
PBS Pennsylvania. University. University museum. Publications of
the Babylonian section (Philadelphia, 1911 ).
PBS IV 1 POEBEL, ARNO. Historical texts (1914).
PBSY POEBEL, ARNO. Historical and grammatical texts (1914).
PBS VIII 1 CHIERA, EDWARD. Legal and administrative documents from
Nippur (1914).
PBS VIII 2 CHIERA, EDWARD. Old Babylonian contracts (1922).
PBSX2 LANGDON, S. H. Sumerian liturgical texts (1917).
oi.uchicago.edu
L I S T OF ABBEEVIATIONS xv
PBS X I 1 CHIERA, EDWARD. Lists of personal names from the temple

school of Nippur (1916).
PBS X I I I LEGRAIN, LEON. Historical fragments (1922).
PBSXY LEGRAIN, LEON. Royal inscriptions and fragments from Nippur
and Babylon (1926).
R RAWLINSON, SIR HENRY. The cuneiform inscriptions of Western
Asia (5 vols.; London, 1861-84; Vol. IV, 2d ed., 1891).
RA Revue d'assyriologie et d'areh&>logie orientale (Paris, 1884 ).
REC THUREAU-DANGIN, FRANCOIS. Recherches sur Forigine de P6cri-
ture cun&forme (Paris, 1898).
RISA BARTON, GEORGE A. The royal inscriptions of Sumer and Akkad
(New Haven, Conn., 1929).
RLA Reallexikon der Assyriologie, hrsg. von ERICH EBELING . . . . und
BRUNO MEISSNER (Berlin und Leipzig, 1932 ).
RTC THUREAU-DANGIN, FRANCOIS. Recueil de tablettes chaldeennes
(Paris, 1903).
SAK THUREAU-DANGIN, FRANSOIS. Die sumerischen und akkadischen
Konigsinschriften (Leipzig, 1907).
SEM CHIERA, EDWARD. Sumerian epics and myths (OIP XV. Cunei-
form series. I l l [Chicago, 1934]).
SL t)EiMEL, ANTON. Sumerisches Lexikon (2. Aufl.; Roma,
1930 ).
SRT CHIERA, EDWARD. Sumerian religious texts (Crozer Theological
Seminary. Babylonian publications. I [Upland, Pa., 1924]).
TMH Jena. Universitat. Texte und Materialien der Frau Professor
Hilprecht Collection of Babylonian Antiquities im Eigentum
der Universitat Jena (Leipzig, 1932 ).
TRS GENOUILLAC, HENRI DE. Textes religieux sum&riens du Louvre.
I - I I (Paris. Mus6e national du Louvre. Textes cuneiformes.
XV-XVI [Paris, 1930]).
U Joint Expedition of the British Museum and of the Museum of
the University of Pennsylvania to Mesopotamia. Ur collection.
UE Joint Expedition of the British Museum and of the Museum of
the University of Pennsylvania to Mesopotamia. Ur excava-
tions (London, 1927 ).
UEI HALL, H. R., and WOOLLEY, C. L. Al-<Ubaid (1927).
UEII WOOLLEY, C. L. The royal cemetery (1934).
UET Joint Expedition of the British Museum and of the Museum of
the University of Pennsylvania to Mesopotamia. Ur excava-
tions; texts (London, 1928 ).
UETl GADD, C. J., and LEGRAIN, L. Royal inscriptions (1928).
VAT Berlin. Staatliche Museen. Vorderasiatische Abteilung. Thon-
tafelsammlung.
W-B Oxford. University. Ashmolean Museum. H . Weld-Blundeli col-
lection.
oi.uchicago.edu
XVI L I S T OF ABBREVIATIONS
WZKM Wiener Zeitsehrift fiir die Kunde des Morgenlandes (Wien,

1887 ).
YOS Yale Oriental series. Babylonian texts (New Haven, Conn.,
1915 ).
YOSI CLAY, A. T. Miscellaneous inscriptions in the Yale Babylonian
collection (1915).
ZA Zeitsehrift fiir Assyriologie und verwandte Gebiete (Leipzig,
1886 ).
ZDMG Deutsche morgenlandische Gesellschaft. Zeitsehrift (Leipzig,
1847 ).
oi.uchicago.edu
I
INTRODUCTION
The first fragment of the Sumerian King List of any importance was pub-
lished by Hilprecht in 1906,1 the second by Scheil in 1911.2 The following ten
years saw a steady stream of new material appear: four important texts were
published by Poebel in 1914,3 two more by Legrain in 1920-21 ;4 and lastly, in
1923, came the magnificent Weld-Blundell prism,6 which in many respects was
to close the earlier phase of the study of our document.
The interest which this material aroused in the scientific world was consider-
able, and numerous scholars took up the problems which it presented. Besides
the names already mentioned we might cite Gadd, Langdon, Eduard Meyer,
Thureau-Dangin, Ungnad, and many others.6 As was natural, considering the
fragmentary state of the material and the gradual way in which it accumu-
1
BE XXI (1906) chap. iv.
*CR, 1911, pp. 606-20.
3
PBS V, Nos. 2-5. The reverse of No. 5 had been published earlier by Hilprecht (see
n. 1); the obverse was new.
4
MJ XI (1920) 175-80; MJ XII (1921) 75-77; PBS XIII, Nos. 1-2.
5
Langdon, OECT II, Pis. I-IV. Since then fragments of Elamite versions have been
published by Scheil in RA XXXI (1934) 149-66.
6
Without attempting completeness we may quote the following:
After Scheil's text (CR, 1911, pp. 606-20): F. Hrozn£ in WZKM XXVI (1912) 143-62;
F. X. Kugler in ZA XXVII (1912) 242-45; Eduard Meyer in Preussische Akademie der
Wissenschaften, Siizungsberichte XXXI (1912) 1062-88, and in his Geschichte des Altertums
I 2 (Stuttgart und Berlin, 1913) § 329a; F. E. Peiser in OLZ XV (1912) 154 f.; Poebel ibid.
col. 290; A. H. Sayce in Society of Biblical Archaeology, Proceedings XXXIV (1912) 165-72;
Scheil in RA IX (1912) 69; Thureau-Dangin ibid. pp. 33-37 and 81-83 and in his La chrono-
hgie des dynasties de Sumer et d'Accad (Paris, 1918) p. 65.
After Poebel's texts: G. A. Barton in MJ VI (1915) 55-58 and in Journal of Biblical Lit-
erature XXXIV (1915) 1-9; L. Delaporte in Revue de Vhistoire des religions LXXII (1915)
183-86; F. Hommel in J. B. Nies, Ur Dynasty Tablets ("Assyriologische Bibliothek" XXV
[Leipzig, 1920]) pp. 205-7; L. W. King, Legends of Babylon and Egypt in Relation to Hebrew
Tradition. (The Schweich Lectures, 1916 [London, 1918]) pp. 27-40; T. G. Pinches in Exposi-
tory Times XXVII (Edinburgh, 1915/16) 517-21; Poebel, PBS IV 1, pp. 71-140.
After Legrain's texts: A. T. Clay in JAOS XLI (1921) 241-63; C. J. Gadd, The Early
Dynasties of Sumer and Akkad ("The Eothen Series" I [London, 1921]); Hommel, "Zur alt-
babylonischen Konigsliste,," Beitrdge zur morgenldndischen Altertumskunde III (Munehen,
1922) 33-36; Langdon in Expository Times XXXII (1920/21) 410-13; Poebel in ZA XXXIV
(1922) 39-53; Thureau-Dangin in RA XVIII (1921) 153 f.; A. Ungnad in ZA XXXIV 1-14;
L. Waterman in AJSL XXXIX (1922/23) 233-47.
1
oi.uchicago.edu
2 T H E SUMERIAN K I N G L I S T
lated, most of these studies were concerned primarily with the reconstruction
of the text, the placing of the known fragments, and the filling up of gaps.
The reliability of the information contained in the fragments was rarely
seriously questioned. Most scholars inclined to accept it at face value and
saw the chief difficulty in the fragmentary state of our knowledge of the text:
"The royal canon when complete would be a most precious document and
would help us to fix the lines of Babylonian history from the legendary time
of the kings after the flood, down to the end of the Isin dynasty." 7 The opti-
mistic view prevalent and the enthusiasm at seeing the gaps gradually fill up
is vividly expressed by Gadd in the introduction to The Early Dynasties of
Sumer and Akkad (p. v):
The acquisition of a complete and trustworthy scheme for the foundations of the
oldest history of even one well-defined region in the ancient world is no mean addition
to science, and perhaps only those who work in less favoured fields can duly appreciate
the fortune of the Assyriologist in this important respect. Thanks to discoveries which
date almost entirely from the last ten years, this claim can now be made with confidence
for the early period of history in Babylonia. This short essay is an attempt to utilise
the latest available material, which seems to afford sufficient indications to furnish at
last an entirely connected scheme of chronology, which rests, not upon conjecture, but
upon the evidence of written records, that are, in comparison, almost as old as the
events which they commemorate.
Only two years later, however, Langdon published the Weld-Blundell prism,
which gives the text of the King List almost complete. The publication of this
text, it is true, confirmed in a very gratifying manner the results of the early
phase of the study of the King List. The placing of the known fragments and
the reconstruction of the gaps had for the major part been correct. The text
now available also showed, however, that a number of important rulers whom
one would certainly have expected to find in the King List were not there; and
when excavations at al-cUbaid in that same year revealed an inscription men-
tioning one early king who was listed, Mes-Anne-pada, it was at once obvious
that this king could not possibly have reigned so far back in time as his posi-
tion in the King List would suggest.
The result of these new facts was a wave of rapidly growing skepticism.
The discrepancy between the actual date of Mes-Anne-pada, as indicated by
the orthography of his inscription, and the early place to which he was assigned
in the King List gave point to a chronicle to which Weidner called attention in
1923 and again in 1926, which indicated that several dynasties listed as con-
secutive in the King List must in reality have been contemporaneous. Around
these observations and around the facts that so many kings who were to be
7
Legrain, PBS XIII 15.
oi.uchicago.edu
INTRODUCTION 3
expected in the King List are not mentioned there and that so many of the
older rulers mentioned appear with unbelievably long reigns, center most of
the comments on the King List after 1923.8 Only a few have taken as lenient
a view as Gadd, 9 that "in the main it presents a true summary of the order of
events in the early kingdoms of Sumer, it doubtless preserves substantially the
order in which cities rose to predominance, and the names of many of their
most celebrated rulers," or as Weidner,10 who only makes the reservation that
"seine restlose Auswertung fur die Geschichtswissenschaft wird freilich erst
moglich sein, wenn Chroniken und historische Inschriften uber das zeitliche
Verhaltnis der Dynastien zueinander die notige Aufklarung gegeben haben
werden." More scholars lean toward the view of Eduard Meyer, who, admit-
ting that some of the information of the King List may go back to reliable
historical sources, continues:
Wie weit es freilich den Gelehrten zu Ende des dritten Jahrtausends noch moglich
war, den wahren geschichtlichen Zusammenhang einigermassen festzuhalten oder
wiederherzustellen, bleibt fraglich genug; man wird vermuten dtirfen, dass Dynastien,
die in den einzelnen Stadten nebeneinander bestanden und um das Oberkonigtum ran-
gen, falschlich als aufeinander folgend betrachtet worden sind . . . . und dass nicht
selten einzelne Namen, die sich erhalten hatten, aneinander gereiht und zu Dynastien
verbunden sind. Daneben ist die Einwirkung von Volkssagen erkennbar.11
The extreme consequences of these premises, finally, were sharply drawn by
Landsberger in 1931:
Der Wert der Kdnigsliste, der selbst in historisch vollig klaren Perioden wegen ihrer
Gepflogenheit, gleichzeitig regierende Dynastien hintereinander aufzufuhren, be-
schrankt ist, ist fiir diese alten Zeiten noch geringer, wie sich aus den hohen Regierungs-
daten, dem Fehlen wichtiger Namen wie Me-silim und Lugal-kisal-si ergibt. Wir haben
8
Doubts concerning the reliability of the data for the dynasties of Kish and Akshak given
in ScheiFs list had been expressed already in 1913 by Meyer, Geschichte des AUertums I 2
§ 329a, and later, in 1918, by Thureau-Dangin, La chronologie ... , p. 65.
Of the literature on the King List after the publication of Langdon's text we may quote
H. Zimmern in ZDMG n.F. I l l (1924) 19-35 and A. Deimel, Die alibabylonische Konigsliste
und ihre Bedeutungfur die Chronologie (Rom, 1935), who are chiefly concerned with the text
of the list. The inconsistencies of its data and the extent of its historical value have been dis-
cussed by Gadd in UE1128-40; Langdon, Excavations at Kish I (Paris, 1924) 5 f.; B. Lands-
berger in OLZ XXXIV (1931) 118 f.; Meyer, Die dltere Chronologie Babykmien*, Assyriens
und Agypte?is (Stuttgart und Berlin, 1925) pp. 27-39; Sidney Smith, EHAf pp. 18-44;
E. Unger in OLZ XXXVII (1934) 363 f.; A. Ungnad, Subartu (Berlin und Leipzig, 1936)
p. 35; Weidner in Archivfiir Keihchriftforschung I (1923) 95 and in AOF III (1926) 198 f.
An effort to trace the genesis of textual errors in the Agade dynasty was made by me in
Acta Orientalia V (1927) 302-9. The question of what sources underlie the King List is dis-
cussed by H. G. Guterbock in ZA n.F. VIII (1934) 2-7 and by Sidney Smith, op. cit. p. 29.
9
History and Monuments of Ur (London, 1929) p. 67.
10 ll
AOF III 199. Die alttre Chronologie , p. 38.
oi.uchicago.edu
den Eindruck, dass die in der spateren Zeit durch die Sage beriihmt gewordenen Gestal-
ten an beliebiger Stelle als Dynastiengriinder untergebracht wurden. Jedenfalls sind
hier sehr verschiedenartige Quellen ohne richtige historische Tradition kompiliert wor-
den. Daraus ergibt sich, dass wir uns von der Konigsliste vollstandig emanzipieren mussen.
Hoehstens, dass wir ihr noch ganz rohe Kenntnis der relativen Aufeinanderfolgen zu-
trauen, so dass z.B. ein Name, der sich in der 9. Dynastie findet, nicht vor die 3. zu
setzen 1st.12
A strong element of uncertainty concerning the value of the King List has
unquestionably been introduced; for, although Landsberger may have stated
the consequences more sharply than others, these consequences have been
drawn in practice. In late years the study of the King List has come almost to
a standstill, and its evidence is hardly ever used for purposes of chronology.
But complete disregard of the King List and its evidence is not justifiable. It
must be the purpose of further study to penetrate this general uncertainty and
to define as far as possible just what is unreliable in the King List and what is
not. The present essay represents an effort in this direction and endeavors to
reach a clearer estimate of the historical value of this document by a study of
the development of its text, the time of its composition, the sources used by
its author, and the manner in which those sources were treated. We are here
in many respects continuing lines of investigation suggested by other scholars,
notably Weidner and Sidney Smith, but point in the main to new ways of ap-
proach. It is our hope that this essay will contribute to bringing the study of
the King List out of the dead water in which it now lies.
12
OLZ XXXIV 119; italics ours.
oi.uchicago.edu
II
TEXTUAL PROBLEMS
THE INDIVIDUAL MANUSCRIPTS
The texts which are of importance for the study of the Sumerian King List
are the following:
WB Ashmolean Museum 1923.444. Published by Langdon, OECT II, Pis.
I-IV. Transliteration and translation with valuable notes on pp. 8-21.1
WB begins with the rulers before the flood and continues with only
minor gaps to Sin-magir of the dynasty of Isin. The provenience of
the text is given as Larsa.la If we assume, as seems most probable, that
the scribe carried his copy up to date, WB must have been written in
the 11th year of Sin-magir; for the list ends with this ruler and assigns
to him a reign of 11 years.2 This dating is supported by the character
of the writing, which can hardly be placed later than the middle of the
1st dynasty of Babylon.3
P2 CBS 13981.4 Published by Poebel, PBS V, No. 2; photograph on
PL XC. Transliteration and translation in Poebel, PBS IV 1, pp. 73-78.
The obverse of P2 begins in the first half of the 1st dynasty of Kish
and carries on with only small gaps to the end of the 1st dynasty of Ur.
The reverse lists four kings of the Isin dynasty and has two valuable
columns of summaries of the dynasties originally listed by the tablet.
P2 was found in Nippur by the Babylonian Expedition of the University
1
Transliteration and translation of this source may be found also in Barton, RISA,
pp. 346-55, and Deimel, Die altbabylonische Konigsliste, pp. 14-29.
la
Langdon (OECT II 1) states that it was "written in Larsa." Deimel, op. cit. pp. 29 f.,
states that it comes from Kish. Is this a mistake?
2
Cf. Langdon, OECT II 1 and esp. 21, n. 6: "The prism must have been written at
the end of the reign of Sinmagir for Damifc-ili-shu the last king is omitted."
3
WB uses consistently the forms £*= and *%&& for DUMU and NAM. These forms became
relatively rare already before Hammurabi. In documents of this ruler and later DUMU gener-
ally takes the form £= or £=, and NAM is written >tti. Similarly WB uses >&i for MU and ts$
for BI, forms which are not frequent after the 30th year of Hammurabi, when ># and £* or
Pp become the current forms. Lastly GA, which in WB takes the form ^ , is seldom writ-
ten with more than two horizontal wedges after the reign of Hammurabi, when the forms pa
and £fr flourish.
* See Legrain, PBS XIII 17.
oi.uchicago.edu
6 THE SUMERIAN KING LIST
of Pennsylvania. The text has been dated by Poebel to the 4th year of
Enlil-bani.5
P3 CBS 13994.6 Published by Poebel, PBS V, No. 3; photograph on PL
XCI. Transliteration and translation in Poebel, PBS IV1, pp. 78-80.
The obverse of P3 covers—with two large gaps—the 1st dynasty of
Kish and ends with the first king of the 1st dynasty of Uruk. The re-
verse has portions of the dynasties of Akshak and Agade. P3 was found
in Nippur by the Babylonian Expedition of the University of Pennsyl-
vania. The date of the text cannot be determined with certainty. We
may guess at approximately the middle of the 1st dynasty of Babylon.7
P4 CBS 13293.8 Published by Poebel, PBS V, No. 4; photograph on PL
XCI. Transliteration and translation in Poebel, PBS IV 1, pp. 80 f.
P 4 has on the obverse the summary of the 4th dynasty of Uruk and
the beginning of the dynasty of Gutium. The reverse preserves rem-
nants of a final summary. P4 was found in Nippur by the Babylonian
Expedition of the University of Pennsylvania. Since the fragment is
small, definite epigraphical clues to its date are lacking. The general
character of the writing suggests, however, that it belongs in the period
of the 1st dynasty of Babylon. Poebel assumes that it was written in
the 23d year of Damiq-ilishu, but it is doubtful whether such close dat-
ing is possible.9
6
PBS IV 1, p. 98: ". . . . the summary in Column 10 assigns to the dynasty of Isin
only 11 kings and 159 years instead of 16 kings and 225 years, a fact which can be explained
only on the assumption that the eleventh king of Isin was the ruling monarch when the
list was drawn up. Figuring from the number of years given to the various kings of Isin in
list No. 5, the first year of Enlil-bani, the eleventh king of Isin, was the 156th year of the
kingdom of Isin, and it is thus clear that the list was finished in the fourth year of Enlil-bani.''
The writing on the tablet is in perfect accord with this dating.
• See Legrain, PBS XIII 17.
7
Even the writing does not give a clue, as the indications are too vague. We have based
our guess on the fact that such consistent use of the simple form & for MU as this ms. ex-
hibits fits in better around the middle or toward the end of the dynasty than at the begin-
ning, while the forms & and & for GA are relatively rare after the reign of Hammurabi.
8
See Legrain, PBS XIII 17.
9
Poebel arrives at this date (PBS IV 1, pp. 98 f. and 102 f.) by comparing the grand
total of P4, 139 kings in 32,243 years, * months, and 18(?) days, with that of P2, 134 kings
in 28,800(-far?)-f-76 years, y months, and 21(?) days, and by explaining the surplus of
kings and years as due to the fact that P4 was written later than P2. Having already dated
P2 to the 4th year of Enlil-bani, he can place P4, which has 5 more kings than P2, in the
reign of the fifth king after Enlil-bani, Damiq-ilishu. The 3,367 years which P4 has more
than P2 cannot, of course, be distributed among the five kings between Enlil-bani and Da-
miq-ilishu; but Poebel plausibly points out that the tens and units of the two totals 32,243
oi.uchicago.edu
TEXTUAL PROBLEMS 7
P5 CBS 19797.10 Published by Poebel, PBS V, No. 5.11 Transliteration

and translation in Poebel, PBS IV 1, pp. 82 f.
The obverse of P6 contains the middle part of the 1st dynasty of
Kish; the reverse deals with the 3d dynasty of Ur and the dynasty of
Isin. P 5 was found in Nippur by the Babylonian Expedition of the
University of Pennsylvania. It may be dated from the character of
the writing to the second half of the 1st dynasty of Babylon.12
P6 CBS 15365.13 Only the reverse of this fragment has been published, and
in transliteration and translation only, by Poebel, PBS IV 1, p. 81. u
Col. i of the reverse has portions of a dynasty in which Sumu-abum—
probably Sumu-abum of Babylon—occurs. Col. ii has remnants of a
and 28,800(+£?) +76 have a better chance of being correct than the other parts of the fig-
ures, so that the original difference may have been 67 years, the difference between the
tens and units of the two totals. This difference, 67 years, is—according to Poebel—exactly
the distance in years from the 4th year of Enlil-bani to the last year of Damiq-ilishu, if we
accept the evidence of PeJ i.e., the difference in the totals of years points to the same ruler
as the difference in the totals of kings had already indicated. Poebel therefore concludes
that P* was written in the 23d year of Damiq-ilishu.
It will easily be seen that the strength of this argument lies in the convergence of two
independent lines of evidence upon the same reign. The difference in the number of kings
points to Damiq-ilishu, and the difference in years also points to the reign of that ruler.
However, as seen by Mrs. Ruth S. Brookens of the Editorial Department of the Oriental
Institute, the difference of 67 years does not, as assumed by Poebel, correspond to the
distance in years from the 4th year of Enlil-ban! to the last year of Damiq-ilishu as given
in P6, for that distance (204-3+4+4+11+23) amounts to no more than 65 years. In
other words, the 67 years which should represent the difference in date between the writing
of P2 and the writing of P< carry the latter beyond the reign of Damiq-ilishu, and there is
thus no longer agreement between the evidence from the number of kings and that from
the number of years. We must therefore accept Mrs. Brookens' conclusion that the evidence
does not permit exact dating of the fragment.
1° See Legrain, PBS XIII 17.
11
The reverse of the tablet had been published earlier by Hilprecht, BE XX 1 (1906)
PL 30, No. 47; photographs of both obverse and reverse are given on his PL XV.
12
To the latter half of the 1st dynasty point the forms £* and £*T of BI and DA.
That the text was written after Samsu-iluna is also shown, as Poebel has pointed out in
PBS IV 1, p. 83, n. 5, by the fact that the divine name Irra is written with the determina-
tive d i n g i r .
13
See Legrain, PBS XIII 17 f. According to the information given there the tablet has
now disappeared.
14
Hommel (in J. B. Nies, Ur Dynasty Tablets, p. 205), who claims to have been the
first to identify the fragment ("des erst von mir in seiner Wichtigkeit erkannten und richtig
eingereihten Fragments C.B.S. 15365"), states that the obverse is destroyed ("der abge-
wetzte Obv. enthielt Teile von Dyn. Ki§ I und Uruk I").
oi.uchicago.edu
dynasty summary. 15 P 6 was found in Nippur by the Babylonian Ex-

pedition of the University of Pennsylvania. Its date is unknown.
18
The identification of the dynasties listed on this fragment presents difficulties. Poebel,
who published it, left the question open (PBS IV 1, p. 95). HommeFs suggestion (in Nies,
op. tit. p. 205, n. 4) that col. i contains part of the dynasty of Isin and that the city name
missing in col. ii should be restored as Akshak cannot be correct and has rightly been re-
jected by Legrain (PBS XIII 18). A definite solution of the problem can hardly be given
until we get more material concerning the dynasties ruling in Babylonia in the Isin period,
but we may at least offer a suggestion.
Col. i contains a list of rulers not known elsewhere:
[ N. ]
4[-f-x mu i-a 6 ]
reigned 4 + z years.
d
l [r - r a - . . . . d u m u]
Irra(?)- . . . . (less likely:
divine Warad- . . . . ), son of
u[r-....] Ur- . . . . ,
dumu nu-mu-[un-tuk] did not get a son;
6 mu i - [as] he reigned 6 years.
s u - m u - a - b u - u [m] Sumu-abum
itu 8 i - as reigned 8 months.
[i - k] u - u n - p i (KA?) - i § t a r (e h - 1 a r ?) Ikun-pi-Eshtar
[. . m u] i - a5 reigned . . years.
Since the names so clearly belong in the Isin-Larsa period, we must assume that our list
was written in a city which at that period had independent rulers and that the scribe
added the local dynasty to his copy of the King List. This assumption is confirmed and
the choice of city is narrowed down when we compare col. ii:
[ 1 [ 1
[SU-NIGIN . . 1 u g] a 1 Total: . . kings
[m] u - b i 125 its 125 years
i b - as reigned
[a - r] a 6 - k a m six times
[§ a . .]. .ki - a in . . . . .
[SU-NIGIN . . 1 u g a] 1 Total: . . kings
etc.,
for we find here a city which has been ruling city as many as six times. The classical edition
of the King List knows only one city which comes up toward this figure, namely Uruk,
which was ruling city five times; and from contemporaneous documents we know that
Uruk actually did have independent kings during part of the Isin-Larsa period. If we as-
sume therefore that it is the local dynasty of Uruk which is treated in col. i (the name of
Sumu-abum might then stand for a short Babylonian domination), the total of six dynasties
in col. ii becomes correct, inasmuch as the scribe had the five classical dynasties plus the
local dynasty of the Isin-Larsa period to sum up.
One difficulty with this explanation must, however, be mentioned. The city name in
col. ii appears as [ ]. .ki - a . If the city name to be restored is actually Uruk, we should
have expected [ ]. . k i - g a ( ! ) , i.e., [ u n ] u k i - g a , Less important is the fact that the
figure giving the total number of years for the six dynasties does not agree with what we
might expect from the five known dynasties of Uruk plus a dynasty of the Isin-Larsa period,
for it is obvious that the total is corrupt; 125 years is much too small a span of time for
six dynasties.
oi.uchicago.edu
TEXTUAL PROBLEMS 9
U CBS 14220. Published by Legrain, PBS XIII, No. I.16 Transliteration

and translation on pp. 25-28. A photograph of the text is published at
the end of the volume on PI. II.
The text of Li covers, with several large gaps, the period from the
end of the 1st dynasty of Ur to the beginning of the Isin dynasty. Li was
found in Nippur by the Babylonian Expedition of the University of
Pennsylvania. It may be dated to approximately the first half of the
1st dynasty of Babylon.17
L2 CBS 14223. Published by Legrain, PBS XIII, No. 2. Transliteration

and translation on p. 24.
The obverse of L2 lists three rulers of the 1st dynasty of Kish. The
reverse of the fragment is uninscribed. L2 was found in Nippur by the
Babylonian Expedition of the University of Pennsylvania. Its date is
uncertain.
S BM 108857. Last published by Gadd, The Early Dynasties of Sumer

and Akkad (1921) Pis. 1-2.18 Transliteration and translation on
pp. 3-7.
S begins with the dynasty of Akshak and carries on, with a gap in
the middle of the Agade dynasty, to the end of the 4th dynasty of Uruk.
The provenience of the tablet is unknown. The character of the writing
suggests that it was written in the second half of the 1st dynasty of
Babylon.19
18
The text consists of two fragments which join. Both fragments had been published
earlier by Legrain, the first in MJ XI (1920) 175-80, the second in MJ XII (1921) 75-77.
A transliteration and translation of the first fragment, made from the photograph in Le-
grain's article, was given by Gadd in The Early Dynasties of Sumer and Akkad, pp. 9-12.
Cf. also the transliteration and translation given by Ungnad and Poebel in ZA XXXIV
(1922) 2-8 and 14 and 39-42 respectively.
17
An upper date limit is represented by the reign of Ishm£-Dagan, who is mentioned in
col. x, while the forms £#> for GA and ££ for BI and the consistent use of #& rather than >#
for MU are rare in the latter half of the 1st dynasty of Babylon and thus suggest the middle
of this dynasty as a likely lower limit.
18
The first publication of the tablet was by Scheii in CR, 1911, pp. 606-20. In RA IX
(1912) 69 Scheii gave further results obtained when the tablet was cleaned and an excellent
photograph. Thureau-Dangin also has published a copy of the tablet in La chronobgie des
dynasties de Sumer et d'Accad, pp. 59 f.
19
The forms £* and & for BI and GA respectively and the consistent use of ># rather
than >£& for MU point to the later part of the 1st dynasty of Babylon. Cf. Gadd, op. tit. p.
1, who dates the tablet merely to the 1st dynasty of Babylon.
oi.uchicago.edu
SUI Published by Scheil in RA XXXI (1934) 150 f. Transliteration and

translation on pp. 152-59.
Sui contains portions of the 1st dynasty of Kish (col. i), the 1st dy-
nasty of Uruk (col. ii), the 2d dynasty of Ur (col. iv), the dynasties of
Maeri and Akshak (col. v), the 4th dynasty of Kish and the 3d dynasty
of Uruk (col. vi), the Agade dynasty (cols, vi-vii), the 4th dynasty of
Uruk (col. vii), and the 3d dynasty of Ur and the dynasty of Isin (col.
viii). The provenience of the text is Susa.20 The writing suggests a date
around the middle of the 1st dynasty of Babylon.21
Su2 Published by Scheil in RA XXXI (1934) 160 as Fragment A. Translit-
eration and translation on pp. 159-61.
The obverse of the fragment contains portions of the 1st dynasty of
Kish (cols, i-ii) and of the 1st dynasty of Uruk (col. iii). The reverse has
not been published (prism?). The provenience of Su2 is Susa.22 The
writing suggests a date around the middle of the 1st dynasty of Baby-
lon.23
Su3+4 Published by Scheil in RA XXXI (1934) 162 and 164 as Fragments
B and C. Transliteration and translation on pp. 161-66. Only one side
of each fragment has been published. The distribution of Jbhe text in
cols, ii, iii, and iv of Fragment B when compared with that in cols, i,
ii, and iii of Fragment C provides clear indications that these two frag-
ments are parts of a single tablet (or prism?) severed by a break which
destroyed two lines at its narrowest point. We are therefore treating
these two fragments as parts of one text.24
20
Scheil in RA XXXI149 and in MSlanges Maspero I (Cairo. Institut francais d'arch^-
ologie orientale du Caire, "Memoires" LXVI [Le Caire, 1934]) 393-400.
21
Comparison of the sign forms used in the Su texts with those found in contracts from
Susa dated to the s u k k a l - m a f e ' s Shiruduh-Kutir-Nahhunte (MDP XXII-XXIV,
the texts listed in MDP XXIV i—ii under rulers 2-7) and belonging approximately to the
time of Warad-Stn, Rim-Sin, and Hammurabi shows close affinities, although those in Sui,
Su2, and SU3+4 seem to be slightly older. Note especially the forms GA a: fc^, GA b: £$>, is:
gr, LXJGAL a: &£»y and LUGAL b: wfc>- used in the Su texts. These forms are for the major
part still to be found in the group of contracts mentioned (GA b in No. 62, later forms—
mt etc.—in Nos. 200-201, 131, and 347; i§ in Nos. 328, 200, and 376, later form with
only two horizontal wedges in Nos. 131 and 347; LTJGAL 6 in No. 347, slightly more simplified
forms in Nos. 131, 202, and 376) but are in the process of disappearing there.
22
Scheil in RA XXXI 149 ff.
23
See n. 21 above.
24
If we place the two fragments so that the beginning of col. ii in Fragment C is exactly
two lines below the end of col. iii in Fragment B, as required by the context, then the gaps
oi.uchicago.edu
TEXTUAL PROBLEMS 11
S113+4 contains portions of the 2d dynasty of Uruk (col i), the 3d and
4th dynasties of Kish (col. ii), the 3d dynasty of Uruk (col ii), the
dynasty of Agade (col. iii), the 4th dynasty of Uruk (col. iii), the dy-
nasty of Gutium (cols, iii-iv), the 5th dynasty of Uruk (col. iv), and
the 3d dynasty of Ur (col. iv). The provenience of Su3+4 is Susa.25 The
writing suggests a date around the middle of the 1st dynasty of Baby-
lon.26
K K 8532+K 8533+K 8534. Published by L. W. King, Chronicles con-

cerning Early Babylonian Kings II ("Studies in Eastern History" III
[London, 1907]) 143-45. Transliteration and translation on pp. 46-56.
An improved transliteration and translation of the first part of the text
was made by Poebel, PBS IV 1, p. 84.
The fragment starts with the antediluvian kings and ends with
M&rbiti-apal-u§ur. Of the older sections hardly anything is preserved ex-
cept a few names belonging to the 1st dynasty of Kish. The provenience
of the tablet is, as shown by the museum number, Kuyunjik. It is
Late Assyrian.
between col. ii of B and col. i of C and between col. iv of B and col. iii of C will fit exactly
the number of lines which the King List has at these places:
Col. Ill Col. IV
rm»
To make such a restoration with fragments belonging to different tablets or prisms would
be possible only if the two tablets were exact duplicates, following each other line for line
and spacing the lines exactly alike on the columns. Such close correspondence is in itself
unlikely (one such rare case is that of WB and J; see p. 49). It may be ruled out altogether
in the present instance, for a glance at the fragments will suffice to show that the spacing
of the lines is so haphazard and varied that it can have been guided only by chance and the
immediate convenience of the scribe.
26
Scheil in RA XXXI 149 If.
26
See n. 21 above.
oi.uchicago.edu
G Published by H. de Genouillac, Fouilles frangaises d'el-Âkhymer. Pre-

miires recherches archiologiques A Kich II (Paris, 1925) PL 21, C. 112.
G, a fairly small fragment, preserves remnants of two columns. The
one to the left (vii) deals with the beginning of the Gutian dynasty, that
to the right (vi) with the beginning of the dynasty of Akshak.27 The
obverse (wrongly called "reverse"!) is destroyed. The fragment was
found in the French excavations at Kish and may be dated, from the
general character of the writing, to the second half of the 1st dynasty
of Babylon.
J Unpublished fragment in the collection of the writer.28
27
The fragment may be restored as follows:
Col. vii Col. vi
[ki-su-lu-ub4-gar g u - 1 i - u m*]
[1 u g a 1 m u] rn u - t u k1 U4-kusu k i -a u[n -zi l u g a l - a m
30 m u i - a5]
[im-ta-a lugal-am 3 m] u I - a5 un-da-lu-lu [12 m u 1 -as]
[in-ki-sus 6 m] u i - a5 ur-ur [6 m u i - a5]
[sa(?)-ar-la-gab 6] mu i - fas1 p ii z u r -dn i [r a fy 20 m u I - a5]
[sul-me-e ] 6 m [u 1 - a5] 'i-su-il1 [24 m u 1 - a5]
[e-lu-lu-me-e§] 7 m [u i - a5] [
[i-ni-ma-ba-ki-e]§ 5 [m u i - a5] etc.
[i-ge4-es-a-u]s '61 [m u 1 - a5]
28
See plate at end. The text reads:
OBV.
[ki-su-lu-ub4-gar] The host of
g u - 1 i - u mw Gutium
[*"]rt u k u l 1 b a - s i g was smitten with weapons;
[nam] -1 u g a 1 - b i its kingship
[u n u g] rkn -£& ba-ttim was carried to Uruk.
5 [u n u]rki1 - g a d
utu-k£-g£l In Uruk Utu-hegal
[1 u g a 1] - a m m u 7 ri t u1 6 U4 [15 1 - a5] became king and reigned 7 years, 6 months,
and 15 days.
1 1uga1 1 king
[m u - b i 7] i t u 6 u4 15 1 - ra,^ reigned its 7 years, 6 months, and 15 days,
[u n u g]rkil r*iS1[t u k u l ] [b] a - s [i g] Uruk was smitten with weapons.
REV.
[i-bf-] r d l [stn] I(b)bi-Sin,
d u m u § u ~rd1[s t n - k e4] son of Shti-Stn,
f }
mu 24 l - [a5] reigned 24 years.
5 1 u [g a 1] 5 kings
5 [m] u - b i [. . ! b - as] reigned its . . years.
[u] r ikI - m a fgi5t u k u l b a ] - s i g Ur was smitten with weapons;
[n] a m - rl u g a 1 - b i1 its kingship
[i] - s i - i nkl - § e b a -1 ii m was carried to Isin.
oi.uchicago.edu
TEXTUAL PROBLEMS 13
The obverse deals with the fall of Gutium and with Utu-hegal, the
reverse with the end of the 3d dynasty of Ur. The provenience of the
fragment is unknown.29 The character of the writing suggests a date
around the middle of the 1st dynasty of Babylon.80
DERIVATION FROM A SINGLE ORIGINAL

A comparison of the texts listed above will show extensive and detailed
agreement between them both in form and in content. The names of the rulers,
their mutual order, the distribution of the names over the dynasties, and the
order in which the dynasties appear are virtually the same in all the texts. A
similar fundamental agreement is found in the figures given for the reigns of
the single rulers and for the duration of the various dynasties. Finally, the
short historical or genealogical notes which occasionally are found added to
a name appear, whenever they can be traced, at the same places and with
practically the same wording.
Agreement so extensive and detailed as this is unthinkable except between
texts derived from a common source. A single example will suffice to illustrate
the point. The section dealing with Sargon of Agade is preserved in three of
our texts, WB, Li, and S. It takes the form: "Sargon—his . . . . was a date-
grower—cupbearer of Ur-Zababa(k), king of Agade, the one who built Agade
(var.: 'the one under whom Agade was built'), became king." In all the texts
this entry differs from those which precede and follow it by giving a certain
amount of historical information. If, now, these three texts were completely
independent of one another we would obviously have to assume (1) that three
different scribes had all independently decided to elaborate this particular
passage with historical information; (2) that it had independently occurred to
each of them to mention Sargon's origin, his position in life before he became
king, and his founding of Agade, but not any of his political and military
achievements; and finally (3) that all three happened to couch this information
in exactly the same words and to arrange the facts in exactly the same order.
Such an assumption is obviously absurd. The three texts cannot be independ-
ent; and on the basis of this and numerous similar instances we can with ab-
29
The fragment was bought in Baghdad in 1933 together with a small collection of odd
fragments. The dealer could give no information concerning where it came from, as it
had been in his possession a long time.
30
It corresponds in almost every particular to that of WB and cannot be far from that
text in date.
oi.uchicago.edu
solute certainty draw the conclusion that our texts are related, that they ulti-
mately descend from a common original.31
The next step must then be an investigation of the manner in which the
texts are related and how each of them derives from the original. In other
words, we must establish the genealogy of our manuscripts.
GENEALOGY OF THE MANUSCRIPTS
THE POSTDILUVIAN SECTION
The genealogy of versions derived from a common original has to be de-
termined through a study of textual variants and their distribution among the
versions. But our manuscripts of the King List give opportunity for such
study only to a very limited degree. The majority are small fragments. It
is therefore relatively seldom that many of them overlap, and passages com-
mon to several versions, where we might observe the spread of variants, are
few. This scarcity of material is to some extent remedied by the fact that most
of the fragments, even the smaller ones, preserve examples of the formulas
used for single kings, dynasty totals, and changes of dynasty, making it pos-
sible to compare parallel passages in versions which do not actually overlap.
Even so, however, the material must be considered meager. It is obvious that
we cannot hope to establish the pedigree of our versions in any great detail but
must content ourselves with determining some of its more prominent lines.
EVIDENCE FROM THE PRINCIPAL VARIANTS
The sections in which we have relatively many manuscripts overlapping so

that we can study the distribution of variants over a fairly large number of
versions are two, namely (1) the middle of the 1st dynasty of Kish and (2) the
dynasty of Agade. In addition we can make use of the evidence obtainable
from a study of the formulas for single rulers, dynasty totals, and changes of
dynasty and also of some overlaps of more limited scope, for example the varia-
M
The view which we have stated here (see also Acta Orientalia V [1927] 302), that our
texts are copies, or copies of copies, of a single original document, seems to be generally
held by scholars, if one may judge from such expressions as "the King List" and "dupli-
cates" used of the various texts concerned. But, since few have paid special attention to
the question, statements of a definite character are rare. We have found only two: Sidney
Smith, EHA, pp. 29 f.: "Where these fragments can be compared, they agree save for trivial
differences in spelling and in the figures assigned to the different reigns; there can be no
doubt that they all depend upon one original source Where there are discrepancies
in the lists they arise partly from scribal errors, possibly due to writing from dictation, partly
from omissions. It is justifiable to assume that the original wording of the source, the first
list, can be inferred from the extant fragments/' and L. W. King, Legends of Babylon and
Egypt* P« 28, n. 1: "The fragments belong to separate copies of the Sumerian dynastic
record." But already on p. 30 King refers to the copyist who wrote Pi as "the compiler of
one of our new lists"!
oi.uchicago.edu
TEXTUAL PROBLEMS 15
tion in the position of the 3d dynasty of Kish (Ku(g)-Baba). We may begin

the investigation with the variants in the 1st dynasty of Kish. These form a
group which from its salient feature may be named the "Atabba variants."
THE ATABBA VARIANTS
The middle part of the 1st dynasty of Kish is preserved in five of our manu-
scripts: WB, P2, P3, P5, and Sui. It takes the forms shown on page 16.32
Comparing these versions, we see that the last four, P2, P3, P5, and Sui, are
very much alike. The small variations which occur are easy to account for,
and the original form of the text can be established with confidence. It must
have been
1 ga-1f-b u-um 6 840 m u i - a5 10 900 m u 1 - a5
2 900 mu l-a 5 7 ka-lu-mu-um 11 a r - w i - t i - u m
3 d-ba 8 900mu i-a 5 12 dumu ma§-da-ke 4
4 [ ] m u 1 - a5 9 zu-ga-gi4-ip 13 720 m u 1 - a6
5 &~1ab-ba
We arrive at this text through the following considerations:
Lines 1-2.—As far as they are preserved all four sources agree on these lines,
which must accordingly have appeared thus in the original.
Lines 8-^—P3 reads [& -] b a [ ] m u 1 - a5. P2 has the same name,
r
a - b a1, but leaves the line giving his reign blank. P5 also has the name
k - rb a1, but the rest of the line is so damaged that we cannot decide
what form it took. Sui, finally, omits both lines. The origin of these varia-
tions is clear. P3 obviously preserves the original text, which gave both name
and reign. P2 was copied from a tablet in which the line giving the reign had
been damaged, just as is now the case with P5. The copyist therefore left
that line blank. Sui derives from a version in which both lines had been
damaged and which was copied without indication that a lacuna existed in
that place.
Lines 5-6.—These lines present a very similar picture. P3 reads W -
t a b - b a 84r01 m u i - a5. P2, P5, and Sui also have this name (in Sui only
its first sign remains); but the line which gives the reign is destroyed in P5,
32
For details of readings see the notes to this section in our edition of the King List (pp.
78-81). In this particular discussion we have, to make comparison easier, adopted a uni-
form distribution of the text in lines for all the manuscripts; and in order to make ortho-
graphical differences between the versions stand out clearly we have here kept closer to the
classical values of the signs than elsewhere. We thus transcribe g a - l u - m u - u m in-
stead o f q d - l u - m u - u m , k a - l u - m u - u m instead of q a - l u - m u - u m , to make
the differences in the writing more evident, and so on; but we do not go so far as to write
k a - g a - g i i - i p rather than z t i - g a - g i 4 - i p to distinguish it from z u - g a - g i i - i p .
The variation between m u x and x m u in the formula for introducing single rulers is
discussed in the section "Variants in the Formulas" (pp. 28 ff.); it does not concern us here.
oi.uchicago.edu
WB P2 Pa P6 Sui
g a - l i - b u - u m33 [g a -11 -] b u - u m [ga-li-b]u-um r
g a1 - 1 f - [b u -] u m â^H-tbu-um]
m u 960 1 - a 5 [90]0 m u i - ^ [900 m] u i - a 5 [m u 900 i -a6] 9[0]0 [m u i - a5]
f
ka-lu-mu a" - b a1 [i-] b a k- rb a1
m u 840 1 - a 5 [ ] m u i - 'as 1
[ ] om.
5zii-ga-gi4-ip 'd1 - 1 a b - b a r 1
& -1 a b - b a d-tab-ba d - [t a b - b a]
m u 900 i - a 5 8 4 ^ m u i - a5 [m u 840 i - a5] 720 [m u 1 - a5]
d-t ab [ka-]lu-mu-um-e ga-lu-lmul-fum1 ka-lu-mu-um qa-lu^lmu-um]
m u 600 i - a 5 [900] m u i - a 5 900 m [u i - a5] [m u 900 i - a5] ÔO1 m u i- W
d-tab-ba [z]u-ga-gi4-ip-e z u - g a - g i 4 [- i p] zu-ga-ki-ip [z] u - g a - g i4 - ri p 1
10 m u 840 i - a 5 '9001 m u ! - a 5 '9001 m u 1 - [a5] fm u 900 i - a5] [6]00 m u i - a 5
ar-wi-ii-um ar-wi fa^-bu-um ar-wi-ii [ar]-wi-um
dumu ma§-da-ke434dumu ma§-da-ke4 dumu ma§-EN-da-ke4dumu m a § - d a - k e 4 [ d u m u ma§]-da-ke4
mu720i-a5 720 m u i - a 5 [72]0 m u i - a 5 [m u 720 \- a5] [720 m u] 1 - a 5
33
Spaced roman has been used for Sumerian, italics for Akkadian. Where an Akkadian name occurs in Sumerian context, however, it has been
treated as a Sumerian "word" and vice versa, just as we would not ordinarily italicize foreign names in English context.
34
On the reading - k e4 rather than - g e* see S. N. Kramer, The Sumerian Prefix Forms b e - and b i - in the Time of the Earlier Princes of Lagas (AS
No. 8 [1936]) p. 8, § 2, and n. 95.
oi.uchicago.edu
TEXTUAL PROBLEMS 17
P2 leaves it blank, and Sui reads 720 [m u 1 - a5]. Of the figures given by P 3
and Sui it is clear that the 840 of P3 has the better chance of being original;
for while it is easy to see how fff (840) passing through a damaged form
Ml can become NF (720) in a later copy, the opposite development is im-
probable. The blank in P2 indicates, as we saw above, that the scribe copied
from a tablet in which this line had been destroyed.
Lines 7-10.—All four manuscripts give much the same text, which must ac-
cordingly represent that of the original. Doubt might arise as to whether the
name Qalumu(m) was written with g a - (P3), k a - (P6), or q a - (Sui) in
the original, whether this and the following name Zuqaqlp had originally
the subject - e which they take in P2 but not in the others, and whether Zu-
qaqlp was written with g i4 as in P2, P3, and Sui or with k i as in P5. These
questions are, however, of minor importance for the time being and may be
decided arbitrarily. Of more interest is the difference in line 10, where P2 and
P 3 have the figure 900, whereas Sui has 600. Of these 900 must be original,
for TW (900) can easily become J* (600) by passing through a damaged
form MW1, but there is no way which leads from 600 to 900.
Line 11.—There are four different forms: a r - w i (P2), r a r 1 - b u - u m
(P3), a r - w i - t i (P5), and [ a r ] - w i - u m (Sui). ArwiDum means "male
gazelle" and therefore goes with other animal names in this section,35 for ex-
ample Qalumu(m), "lamb," and Kalibum, "dog." The most correct writing
of ArwiDum would be a r - w i - i i - u m , and from such a form all our vari-
ants can be effortlessly derived: a r - w i - 6 , a r - w i , and a r - w i - u m
through simple omissions of lost signs by scribes copying damaged originals,
and a r - b u - u m through misreading of a damaged form 41— of #— (wi)
as %r~ (b u) and omission of a lost ti or through a mishearing in dictation after
ArwiDum had been contracted to Arwtim (ArwiDum>ArwuDum> Arwfim).
Lines 12-13.—Only one point could give rise to doubt as to how these lines
read in the original. This is the form d u m u m a § - EN; - d k - k e4 in P 3
as against d u m u m a § - d & - k e4 in all the others. M a S - d 4 can mean
both "gazelle" and muskinum, "plebeian," whereas m a § - E N - d & means
only muskinum™ The scribe of P 3 or of one of its ancestors must have con-
36
On these names cf. Poebel, PBS IV 1, p. I l l ; Landsberger quoted by Zimmern in
ZDMG n.F. I l l (1924) 30, n. 8; Albright in AOF III (1926) 181; and Guterboek in ZA
n.F. VIII (1934) 5. Langdon rejects for no reason the reading z t i - q a - q i « - i p and
proposes d u g - g a - g i - i b (OECT II 10, n. 7).
36
MuMnum, "plebeian," is a III/II participle of *ka?anum and means originally "one
who prostrates himself." This servile type of greeting would be characteristic for the lower
part of the population and could therefore be used as a designation of class: "plebeian."
From Old Akkadian, where the word must have had the form *mu$ka?enum (cf. the Sar-
oi.uchicago.edu
sidered it more likely that the king was son of a plebeian than son of a gazelle
and have introduced the unequivocal form m a § - E N - d & . In reality,
however, the fact that the king's name, ArwiDum, means "male gazelle" and
that other names of the dynasty are animal names makes it quite clear that
m a § - d k , "gazelle," is the original form.37
If we now compare the prototype of P2, P3, P5, and Sui at which we have
just arrived—we may call it "B"—with the text of WB, which we may call
"A," we shall notice a number of very striking differences:
A B
ga-li-bu-um ga-li-bu-um
m u 960 i - a5 900 m u i - a5
k a -1 u - in u a -b a
m u 840 i - a5 [ ] mu i-a5
zti-ga-gi4-ip a-tab-ba
m u 900 i - a5 840 m u i - a5
&-1 a b ka-lu-mu-um
m u 600 i - a5 900 mu l-a 5 .
£-1 a b - b a zu-ga-gi4-ip
m u 840 i - a5 900 m u i - a5
ar-wi-u-um ar-wi-ii-um
d u m u ma§-d&-ke4 dumu m a§-da •ke4
m u 720 1 - a5 720 m u i - a5
gon passage in Legrain, PBS XV, No. 41 vi 24-26: a-na dda-gan ul-ka-en), muskinum
seems to have passed into Sumerian; for in the pseudo-ideogram MAS-EN-KAK we may rec-
ognize an e m e - SAL form of *muska?en: m a s - d a - e n (i.e., m a § d a 3 e n). Transpo-
sition of signs, as here of EN and KAK(= da), is a common feature in pseudo-ideograms,
which are merely traditional orthographies inherited from the time before the signs had
to be written in correct order. Similar cases are dEN-zu for dz u - e n , ZTJ-AB for a b - z u ,
etc. The assimilation of the u to the following a in * m u § k a 3 e n > m a s d a 3 e n is
a common feature in Sumerian (see p. 171, n. 7a) and harmonizes with the well known
preference for a in eme-SAL (cf. Poebel in ZA n.F. I l l [1926/27] 259 and 270). For
the development from k to d cf. Poebel, GSG § 80.
37
That the author of the King List did not intend to state that Arwiûm was the son
of a plebeian is also clear from the form which the statement takes. When the author wants
to give information concerning the social status of the father of a king, he uses a set formula
which is quite different, namely N. a b - b a - n i x. Cf. d g i l g a m e s ab-ba-ni
1 i 1 -1A, "Gilgames—his father was a lilltirdemon" and s a r - r u - k i - i n . . . . - b a - n i
n u - g i r ii2, "Sargon—his . . . . was a date-grower." We should therefore have had
a r - w i - t i - u m a b - b a - n i m a § - d a , "Arwpum—his father was a plebeian," and
not a r - w i - t S - u m d u m u m a s - d a , "Arwiûm, son of m a § - d a ." The correct
interpretation of m a § - d a as a name meaning "gazelle" was suggested as a possibility
already by Poebel, PBS IV 1, p. I l l , but has hitherto only been accepted by Albright
(JAOS XL [1920] 329; AOF III [1926] 181); by Landsberger (see Zimmern in ZDMG
n.F. I l l [1924] 30, n. 8; Landsberger, Die Fauna des alten Mesopotamien nach der 14- Tafel
der Serie ffAR-ra^fyubullu [Leipzig, 1934] p. 100), who first pointed out that Arwpum^
arrnH, "male gazelle"; and by Gliterbock (ZA n.F. VIII [1934] 3 and 5).
oi.uchicago.edu
TEXTUAL PROBLEMS 19
It will be seen first of all that whereas A lists the two rulers k - 1 a b and
a - t a b - b a between Zuqaqip and ArwPum, B places them, as k-ba and
a - t a b - b a , between Kalibum and Qalumu(m). In A, further, Kalibum
is listed with a reign of 960 years, while B gives him only 900; and Qalu-
mu(m), who according to A reigned 840 years, reigned 900 years accord-
ing to B.
To decide which of these two texts represents the original text the more
faithfully we must consider the differences one by one. Little information is—
a priori—to be gained from the variant forms a - t a b (A) and k - b a (B).
It seems likely that one of these forms originated through influence from the
following name a - t a b - b a , but whether a copyist, having written a - ,
continued with - b a instead of - t a b because his eye fell upon the final
- b a in the next name, or whether his original had k - b a and he copied it
as a - t a b because by mistake he looked at k - 1 a b of the following a -
t a b - b a obviously cannot be safely decided at this point.
A similar case is presented by the varying figure for the reign of Qalumu(m),
which is 840 in A and 900 in B. These figures are written Mf and TV. Any
copyist knows how difficult it can be to read a figure like this if the surface
of the text is the least bit scratched or damaged, for FfF may equally well
represent either one. An uncertain form like this is probably responsible for
our variant. But whether an original PFF through such a form became PW
or vice versa remains uncertain.
More helpful are the variants in the reign of Kalibum, which lasted 960
years according to A and 900 according to B, and in the order of the rulers,
which is as follows in the two texts:
A B
Kalibum Kalibum
Qalumu(m) Aba
Zuqaqip Atabba
Atab Qalumu(m)
Atabba Zuqaqip
Arwiûm ArwPum
Here also, it is true, we must admit that the differences might have arisen
in various ways. But among the possible solutions one stands out as definitely
the most probable because it is so much more simple than the others. If we
assume the text presented by A to be original, a single scribal mistake of a
well known type will lead directly to the arrangement of rulers given by B,
and that mistake will at the same time account for the 900 years which B
ascribes to Kalibum in contrast to the 960 found in A. The scribal mistake
oi.uchicago.edu
with which we are concerned would have happened as follows: Having copied
the line g a - l f - b u - u m , the scribe jumped four lines of his original and
continued with the line after Zuqaqlp, m u 900 1 - a 5 , misled, no doubt, by
the close similarity between this line and the one he should have copied,
m u 960 i - as .38 He did not realize the mistake but went on copying the
lines dealing with Atab and Atabba:
Original Copy
ga-li-bu-um ga-li-bu-um
m u 960 1 - a5
ka-lu-mu
m u 840 1 - a5
z u - g a - g i4 - i p
m u 900 i - a5 m u 900 i - a5
d-tab a-tab
m u 600 i - a5 m u 600 1 - a5
d-tab-ba d-tab-ba
m u 840 i - a5 m u 840 i - a5
ar-wi-ti-um ....
dumu mas-d&-ke4
m u 720 l - a 5
When the scribe had copied the line m u 840 i - a 5 , however, and looked
back upon his original to find the continuation, his eye was arrested by the
identical figure, 840, in the line which gave the years of Qalumu(m), and he
discovered his omission. Rather than rewrite what he had already copied, he
took the easy course of inserting the two rulers he had missed at the point
to which he had come when he discovered the mistake. He thus got a list
ga-li-bu-um
m u 900 1 - a5
&-1 a b
m u 600 i - a5
d-tab-ba
m u 840 i - a5
ka-lu-mu
m u 840 1 - as
zu-ga-gi4-ip
m u 900 i - a5
ar-wi-u-um
dumu ma§-d&-ke4
m u 720 1 - a5
88
Cf. the exactly similar mistake in the Agade section discussed on p. 27.
oi.uchicago.edu
TEXTUAL PROBLEMS 21
which has the rulers in the exact order in which they appear in B and which
like B gives Kalibum a reign of 900 years instead of 960.
A single scribal mistake is thus sufficient to explain the two most conspicu-
ous characteristics of B. The explanation works only one way, however, from
A to B, and experiments will show that repeated and complicated errors must
be assumed if we want to account for the opposite development, from B to A.
Our evidence points accordingly to A as the better form of the text.
The conclusion that A is closer to the original than B is based upon the
inherent probability of the most simple solution. That principle is not, how-
ever, altogether without exceptions. It happens, although rarely, that events
do not follow the most probable course, and our argument is thus not abso-
lutely decisive. It is therefore fortunate that we possess confirmatory evi-
dence which settles the question beyond possibility of doubt.
In both of our versions occur the lines a r - w i - t i - u m d u m u m a § -
d k - k e 4 . There is here a slight irregularity because the Mashda mentioned
as father of ArwiDum is otherwise completely unknown. There are naturally
many cases in the list where the father of a king is mentioned, but the father
has then as a rule been king himself and appeared higher up in the list.39 That
Mashda has not similarly appeared before therefore suggests that the list as
we have it is incomplete and that a king Mashda has been left out just before
the ArwPum passage.
Examining our two versions we see that B, which deals with the reign of
Zuqaqlp just before ArwPum, shows no trace whatsoever of a missing ruler
at this place. In version A, however, the case is different. Here ArwPum is
preceded by the two rulers a - t a b and a - t a b - b a , and it cannot but
strike us that the latter of these names, Atabba, is nothing but the genitive
case A t a b b - a (k) of the former, Atab. How does the list come to record
as different rulers two forms of the same name? A glance through the list
itself will give us the explanation, for we find numerous passages in which—
as with Atab and Atabba—the same name appears twice in close succession,
the second time in the genitive. Such passages are those in which a king is
39
In the 33 other cases in which we have the formula N . d u m u P. there is only one
where the father has not appeared earlier as ruler. This one case is highly exceptional and
explains itself. It is [m e s - k i] - & g - g a - [s e - e r] d u m u d u t u , "Mes-kiag-gasher, son
of Utu (the sun-god)." Obviously the author could not well enter the sun-god himself as
king of Uruk in his list.
Of no importance—and therefore not included in our statistics—is the fact that Pa has
after Ur-Ninurta(k) an extra line not found in WB: d u m u rdliM [ ]; for we cannot
draw safe conclusions about the principles of the original from the latest parts of the list,
which are additions by later scribes.
oi.uchicago.edu
22 T H E SUMERIAN K I N G LIST
stated to be the son of his predecessor: "A reigned x years; B, son of A,

reigned y years."
If on the strength of this analogy we assume that the present text of A de-
rives from a slightly broken original through the succession
Original > Copy > A
d-tab d-tab d-tab
m u 600 1 - a5 m u 600 i - a6 m u 600 i- a5
[N. d u m u ] d - t a b - b a d-tab-ba d-tab-ba
m u 840 1 - a5 m u 840 1 - a5 m u 840 1 - a5
the repetition of Atab as Atabba will be satisfactorily explained.

This result is of considerable interest. We have just seen that the passage
a r - w i - u - u m d u m u m a § - d & - k e 4 suggested that a ruler's name
was missing immediately before it. Now we find that another feature of the list,
the absurd reappearance of the name Atab in the genitive form, is explicable
on that very supposition only; and in both cases the same point in the text,
the line before ArwiDum, is indicated. This cannot be coincidence, and we may
accordingly restore the original form of the passage as follows:
d-tab
m u 600 i - a5
ma § - d a dumu d-tab-ba
m u 840 i - a6
a r - w i - i i - u m d u m u ma§-d&-ke4
m u 720 1 - a5.
We may now consider the question whether version A or version B has pre-
served the more nearly original text in the light of these new facts. In A we
have just found a number of indications grouped so that they form an organic
and logical whole and point to a single conclusion. In B these same indications
are scattered, do not form any pattern, may even be unrecognizable; the tell-
tale relationship of the names Atab and Atabba is blurred; the names appear as
Aba, Atabba and are separated from the Arwiûm passage by two other rulers.
This difference between the two versions can mean only one thing: A has
preserved the original text. Here the traces left by the omission of the name
Mashda are still undisturbed and easy to read, while in B time has scattered
and partly obliterated them. To reverse the process and assume that B repre-
sents the original text is impossible. This would mean that the whole group of
indications in A would be due to coincidence. The existence of the ArwiDum
passage, which suggests that a ruler is missing, would be a coincidence. By
coincidence the name & - b a would have been changed so that it became
the nominative of the following Atabba and could convey the same suggestion,
oi.uchicago.edu
TEXTUAL PROBLEMS 23
that a ruler was missing. Finally, coincidence would have caused a copyist
to move two names en bloc from their place in the list and insert them at the
one place where the suggestion which they conveyed in their new form and
that conveyed by the ArwPum passage could be brought to bear on the same
line of the text. Such a series of extraordinary coincidences can safely be
ruled out. We can therefore consider it proved that in this section A has pre-
served the original form of the text, at least as far as the arrangement of the
rulers, the names Atab and Atabba, and related variants are concerned.
This result is of considerable importance for ascertaining the mutual rela-
tionships of our manuscripts. The version which we have called B is represent-
ed by no less than four sources: P2, P3, P5, and Sui. Some of the most promi-
nent characteristics of this version have proved to be due to mistakes in copy-
ing. Version A, on the other hand, represented by only a single source, WB,
is free of these errors. WB must accordingly have descended from the original
by another route than P2, P3, P5, and Sui, a route which avoided the text in
which the errors were made. The possibility that WB belongs to the same
line of descent as the B texts but should be placed above the text in which
the errors were made is excluded by the fact that our manuscripts are
roughly contemporaneous and that one of the B texts seems to be even older
thanWB. 40
We can therefore draw up the following elementary pedigree:
B represents here the copy or consecutive series of copies in which the errors
common to P2, P3, P5, and Sui were first made. The brace is meant to indicate
that the sources placed under it all derive from the original through a common
ancestor but that the exact way in which they descend from that ancestor is
unknown. P2, P3, P5, and Sui may thus have descended from B each in a direct
line, or one may have descended from another and that again from B, etc.
VARIANTS IN THE REIGNS OF THE AGADE RULERS
To a division of our sources very similar to that indicated by the Atabba

variants points a group of variants in the reigns of the Agade kings. The fig-
40
P 2 , which dates from the 4th year of Enlil-bani (p. 6, n. 5).
oi.uchicago.edu
ures for single reigns and dynasty totals preserved in the various manuscripts
are as follows:
WB Li P8 Pa S Sui Sua+4
Sharru(m)-k!n 56 55 [ J [ ] ] I 1
Rtmush 9 15 [ ] [ 1 ] I ]
Man-ishtushu 15 7 [ ] [ ] ] [ 1
Naram-Sin [ ] 56 [516 [ ] ] [ 1
Shar-kali-sharri [ ] 25 24 [ 1 ] 25
[1]57
Igigi, Nanum, Imi,
and E l u l u . . . t i I 1 3 [ ] [ ] [ 1
Years 161 177
Dudu 21 [ ] [ ] [ ] 21
Kings r9* r
9*
Shu-Durul 15 [ ] [ ] [ ] 15
Years 181 [ 1 [ ] [1]97 197
Kings 11 [ ] [ 1 [ 1 12
I t will be noted that the totals given in P3, P2, and S correspond to the fig-
ures for the single reigns preserved in Li and S in such a way that these lists
can be restored with full confidence41 as
Li Pa Pa s
Sharru(m)-kin .. 55 [55] [55] [55]
Rimush 15 [15] [15] [15]
Man-ishtushu.. 7 [7] [7] [7]
Nar&m-Sin 56 [5]6 [56] [56]
Shar-kali-sharri 25 24 [25] [25]
[1]57
Igigi, Nanum, Imi, and
Elulu [3] 3 [3] 3
Dudu [21] [21] [21] 21
Shu-Durul [15] [15] [15] 15
Years [197] [196] [U97 197
Kings [11] [HI [11] 12
41
The high degree of certainty which the interplay of totals and single figures gives may
be illustrated by an example. S has preserved the reigns of the second part of the dynasty
and the total 197 years. Li has the reigns of the first part. Added to each other these single
reigns make exactly the total 197 years, and the single reigns missing in S can therefore be
restored from thefiguresin Li. It is obvious that even if we restored S with figures different
from those of Li the sum of these different figures would have to be the same as that of the
figures of Li, for when we insert the figures of Lt the total is correct. Furthermore, any
differences between the individual figures which we restore in S and the figures of Li could
only have come about by scribal errors in S or Li or both, since a true rendering of the
oi.uchicago.edu
TEXTUAL PROBLEMS 25
The lists are very much alike. Only two points need comment: (1) the fig-
ure for the reign of Shar-kali-sharri, which is 25 in Lx and must have been 25
in P2 and S also, whereas it is 24 in P3; and (2) the curious summary in S giv-
ing 12 kings, although we know only 11 kings of the dynasty. The first of
these points, the different figures for Shar-kali-sharri, must be explained as a
copyist's mistake, but whether 24 became 25 or vice versa cannot be safely
determined.42 As for the second, it seems probable, as suggested by various
scholars,43 that a scribe counted the line after Shar-kali-sharri, a b a m
l u g a l a b a m n u l u g a l , a s a king when he added up.
A second group of sources is formed by Sui and Su3+4. Sui preserves a total
for the dynasty of 161 years and Wt kings. This total, 161 years, is the sum
of the first nine reigns of the dynasty as we know them from Li and S, and
the broken figure #? can be restored as fff (9). We can therefore conclude
that Sui had the samefiguresas Li and S but stopped its account of the Agade
dynasty with the ninth king, Elulu. The text Su3+4 also stops its account of
the Agade dynasty here,44 and the broken figure if for the number of kings,
in col. iii 14, can similarly be restored as ft? (9). Considering that Sui and
SU3+4 are roughly contemporaneous, that they come from the same place,
Susa in Elam, where copies of the Sumerian King List to use as originals can-
not have been abundant, and that both texts stop short in their account of
the Agade dynasty at exactly the same point, it is an obvious conclusion that
they derive from a single original, a text in which by some accident the final
section of the Agade dynasty had been destroyed. This conclusion becomes a
certainty when we compare other sections of the two lists, for elsewhere also
Sui and Su3+4 show the same omissions.
Only one more point needs comment. Sui and SU3+4 are not com-
pletely alike. Instead of the total 161 years given by Sui, SU3+4 has 177
original would obviously keep the texts identical. These scribal errors, however, would have
to be very peculiar; for, since the sum of the figures in Li and those restored in S must be
the same, we would have to assume that every time a scribal error changed one figure an-
other scribal error changed another figure and that by coincidence one error always hap-
pened to add exactly as many years as the other subtracted. Such an assumption is so un-
likely that we need not consider it. Thus the interplay of totals and single reigns assures a
high degree of safety for the restoration. In our table we have such assurance for all of S
and P2, for P3 down to Dudu, and for Li down to Igigi.
42
A similar problem is discussed on p. 19.
« Ungnad in ZA XXXIV (1922) 14; Poebel ibid. p. 46; and Langdon, OECT II 18, n. 2.
My own former objections in Acta Orientalia V (1927) 304, based on the fact that we know
of no analogous instance, are not serious.
44
See the text as restored above (p. 10, n. 24).
oi.uchicago.edu
years. Since both texts derive from the same original, it should natur-
ally be possible to give a reasonable explanation of this difference. This
is indeed the case. If the version from which the ancestor of Sui and
SU34.4 derived had for the reign of Dudu a damaged figure MT, but other-
wise looked like Li and S, the copyist would read 1 instead of 21 and get 177
years as total for the dynasty. When later on the text—with the section on
Dudu and Shu-Durul hopelessly damaged—served as original for Sui and SU3+4,
the scribe who wrote SU3+4 simply took over the total 177 years which he found
in the original, whereas the scribe of Sui checked the figures and, finding a
discrepancy, made a new total by adding up the single reigns. This gave him
161 years.
We have thus seen that the somewhat singular data of both Sui and Su3+4
become understandable if we assume that these versions in the section which
they preserve had the samefiguresas Li and S and derive from a common origi-
nal damaged at the end of the dynasty. We can therefore group them to-
gether with Li, P2, P3, and S. All six versions represent the same form of the
text.
The reconstructed original form of these six versions, based on all the factors
just discussed, is given below. Alongside it we have placed the totally differ-
ent text presented by the last of our versions, WB.
The Six Versions WB
Sharru(m)-kln 55 56
Rimush 15 9
Man-ishtushu 7 15
Nar&m-Sin 56 [ ]
Shar-kali-sharri 25 or 24 (P8) [ ]
Igigi, Nanum,
Imi, and Elulu 3 [ J
Dudu 21 21
SM-Durul 15 15
197 or 196 (P8) 181
As will be seen, only the figures for the last two rulers correspond in the two
lists; the reigns of the first three rulers and the totals disagree; it is not even
possible to restore the three missing reigns in WB from the other sources, as
disagreement with its total would result.
The fact that these two forms of the text dissent so strongly naturally raises
the question of which is the closer to the original. Fortunately the origin of
one of the variants concerned can be elucidated with sufficient certainty to
oi.uchicago.edu
TEXTUAL PROBLEMS 27
provide the answer. In the form of the text represented by the six sources—
for the sake of convenience we may call this form "B"—Rfmush is stated to
have reigned 15 years. In WB, however, the reign of Rimush is given as only
9 years, and it is his successor Man-ishtushu who reigned 15 years. To suggest
that a scribal mistake changed the 15 years of B to the 9 of WB seems out of
the question. These figures, <W (15) and ft (9), are too different to be mis-
read for each other. Looking at the full form of the text as given in WB we
note, however, that the phrase r f m u § d u m u § a r r u - k t n occurs twice
in close succession, so that a scribe copying such a text may easily have
jumped from the first line down to the exactly similar passage in lines 4-5:
WB Copy
rf-mu-u§ dumu 8ar-ru-ki-in ri-mu-u§ dumu (§ar-ru-ki-in
mu 9 i - a8
ma-ni-iS-ti-i§-§u
§e§-gal r 1-mu-uS Y rf-m u-u§
dumu Sar-ru-ki-in dumu) § a r - r u - k i - i n
mu 15 1-&5 mu 15 i - a6
Instead of r i m u § d u m u § a r r u - k t n m u 9 i - a5 the scribe would thus

g e t r l m u § d u m u § a r r u - k l n m u 15 i - a 5 ; that is, a scribal mistake
which may easily happen leads from the data of WB to the data of B. Since
a development from the form presented by WB to that presented by B is
thus easily explainable, whereas no way leads from the 15 years of B to the 9
of WB, we may conclude that WB has preserved the original text at this
point. 45
This result means that six of our versions—Li, P 2 , Ps, S, Sui, and Su344—
all contain the same copyist's error, and it is therefore probable that they
have all descended from the original through a single copy, that in which this
error was made. Within the group Sui and Su3+4 seem to be especially closely
related. They derive from a version broken at the end of the Agade dynasty,
a version through which none of the others has passed. WB, finally, stands
alone. It has preserved a better text than the others and derives from the
original without passing through the version in which the Rtmush mistake
was made. For the same reason as in the Atabba variant (p. 23) we cannot
place WB in the same direct line of descent as the other texts and higher than
44
This explanation purports to elucidate only the variant Rimush 15 in B. How B got
the further error Man-ishtushu 7 we leave an open question at present.
oi.uchicago.edu
28 THE SUMEKIAN KING LIST
the erroneous copy. Instead we must assume that it represents a different

branch of the tradition. We can therefore draw up the following pedigree:46
Original
/ B
/ Su
WB L, P2 P3 S Sut S u 5 U
Here B represents the copy in which the Rimush mistake was made, Su the
broken text from which Sui and Su3+4 were copied. The significance of the
braces has been explained above.
VARIANTS IN THE FORMULAS
The occurrence of the Atabba variants could be observed in five of our

manuscripts. For the variants in the reigns of the Agade rulers we could con-
sult seven. A still greater number of manuscripts can be compared through
observation of the formulas which they use. The King List is given to cliches
and uses set formulas whenever possible. Only three such formulas, however,
recur so frequently through the text that they are of real value for our purpose.
These are (1) the formula for change of dynasty, (2) the formula for dynasty
total, and (3) the formula for introducing single rulers.47 The forms which
they take in our various manuscripts are as follows:
46
In Acta Orientalia V (1927), 302-9, I some years ago proposed a different solution to
account for the variants in the Agade dynasty, in which I assumed that the total 12 kings
in 197 years of S as against 11 kings in 181 years of WB was due to the fact that S had listed
a usurper with a reign of 16 years between Nar&m-Stn and Shar-kali-sharri, whereas WB left
this usurper out. At the time when that article was written I did not have access to PBS
XIII and thus did not know that the two fragments which make up Li actually join, forming
part of a single tablet. It therefore seemed possible to consider these fragments parts of
different versions and to put one of them, the one I called L2, aside as totally corrupt. I thus
missed the significance of the fact that both of the two versions which preserved the section
where my usurper should be placed showed no trace of him, and I similarly overlooked the
highly suggestive correspondence between the single figures of these fragments and the total
197 years in S which indicates that S should be restored from them. Finally, I did not give
due attention to the place of the < -wedge in the damaged figure for the reign before Shar-
kali-sharri in P 8 . This wedge is placed so high that the traces can only be restored as WW (56),
the figure which L* gives as the length of Nar&m-Stn's reign, not as 4tiB (16) for the reign of
my usurper. Thus P 3 also has Shar-kali-sharri as the immediate successor of Nar&m-Sfn.
Since P 8 with its total of 157 years after Shar-kali-sharri should have listed the postulated
usurper, the theory must be abandoned.
47
This formula may, within the single texts, be elaborated in various ways. Its basic
constituents, with which we are here solely concerned, are (1) name of ruler -f- (2) length of
reign + (3) verb: "he reigned." As elaborations—to which we do not pay attention—may
be mentioned insertion after the ruler's name of (a) a patronymic: d u m u P.; (b) some
oi.uchicago.edu
TEXTUAL PROBLEMS 29
1. Formula for change of dynasty: A ki giSt u k u 1 b a - a n - s i g n a m - l u g a l -

b i B k i - § e b a - t u r n . Occurrences: ii 45-46: k i § i k i g i s t u k u l b a - a n - s i g
n a m - l u g a l - b i 6 - a n - n a - § 6 b a - t u m ; iii 37-38: u n u g ki g i l t u k u l
b a - a n - s l g n a m - l u g a l - b i u r i ki - § e b a - t t i m ; i v 17-19: a - w a - a n k i
gi5
[ t u k u l b a - a ] n - s i g n a m - l u g a l - b i k i § i k i - § e [ b a - t i i m ] ; iv 36-38:
k i § i k i g i 5 t u k u l b a - a n - s i g n a m - l u g a l - b i fr a - m a - z fki- £ & b a - t u m ;
iv 43-44: fr a - m a - z iki g i i t u k u l b a - a n - s i g n a m - l u g a l - b i u n u g ki - § &
b a - t i i m ; v 21-22: [ a d a b a k i ] g i i t u k u 1 b a - a n - s l g [ n a m - l u g a l - b ] i
m a - e r i k i 4 9 - § & b a - t t i m ; v 33-35: [ m a - e r i ki g i i t u k u l ] b a - a n - s i g
[ n a m - l u g a l ] - b i [ k i § i k i - § e ] b a - t u m ; v 42-43: k i § i k i g i s t u k u l b a - a n -
s i g n a m - l u g a l - b i u 4 - k u § u k i ( ! ) 5 0 b a - t u m ; vi 6-8: u 4 - k u § u k i g i i t u k u l
[ b a - a n - s l g n a m ] - l u g a l - b i k i § i k i - § & b a - t t i m ; vi 22-23: k i § i ki
gii
t u k u l b a - a n - s l g n a m - l u g a l - b i u n u g k i - § e b a - t t i m ; vi 28-30:
u n u g ki gi5 t u k u l b a - a n - s i g n a m - l u g a l - b i a - g a - d & k i - § & b a - t u m ;
vii 13-14: a - g a - d & k i g i i t u k u l b a - a n - s i g n a m - l u g a l - b i unugki-8&
ki g i l
b a - t t i m ; vii 24-26: u n u g t ukul ba-an-slg nam-lugal-bi ki-
s u - l u - u b 4 ( ! ) g u - t u - u m ( ! ) b a - t i i m ;51 viii 1-2: k i - s u - l u - u b 4 - g a r
epithet, such as s i p a ( d ) , "the shepherd"; (c) information concerning the ruler's ori-
gin, the name of his city, or the character or occupation of his father (this information is
given in the form x - n i y , "his x ['city/ 'father/ etc.] was y ['KuDa(ra)/ 'a littA-demon,'
etc.])"; (d) information concerning the ruler's exploits. This last type of information is
given in the form of a relative sentence introduced by 1 t i . The name of the ruler often
takes the subject element - e , but as the use of this element is generally very irregular
within the sources we are not considering its occurrence or nonoccurrence distinctive.
N.-e x m u i-a 5 and N. x m u 1 - as are thus not considered separate varieties of the
formula.
Related to the formula for introducing single rulers is the formula for introducing dy-
nasties: A k i -a N. l u g a l - a m m u x (or: x mu) i-a* (or: in-a*), "InAN.became
king and reigned x years."
48
WB contains two sets of formulas, one used in the section dealing with the antedilu-
vian rulers and one used in the postdiluvian section. As the various problems connected
with the antediluvian section are dealt with in detail further on (pp. 55-68), we shall here
limit ourselves to the formulas in the postdiluvian part. We quote from Langdon's auto-
graph copy (OECT II, Pis. I-IV) as corrected by collation with photographs of the origi-
nal (see pp. 76 ff.). In many cases Langdon's transliteration (op. cit. pp. 8-21) already has
the correct readings.
49
The reasons which make this reading preferable to m a - r iki are stated by Thureau-
Dangin in RA XXXI (1934) 83 f.
60
The scribe omitted - § e after Akshak here where it belongs and wrote it after Ak-
shak in the following line where it is incorrect. He must twice have looked at the wrong line
of his original when he copied this passage.
61
The scribe forgot GAR after k i - s u - l u - u b 4 - here and in the following line but
remembered it in viii 1. He also omits ki- 8 & after g u - t u - u m in this line, although
he correctly writes g u - 1 u - u mki in the next line. This accumulation of omissions
might suggest that the original of WB was slightly damaged at this place.
oi.uchicago.edu
g [ u - t u - u m k i ] (!)52 n a m - l u g a l - b i u n u g k i - § & [ b a - t u m ] ; viii 7-8:

u n u g k i e'H u k u l b a - a n - s i g n a m - l u g a l - b i u r [fki - § fc] b a - t u m ;
viii 21-22: u r f k i - m a (I)83 «iSt u k u 1 b a - a n - s i g n a m - l u g a l - b i 1-si-
i nki - § e b a - t u m .
2. Formula for dynasty total: x l u g a l m u - b i y i b - a 5 . If the dynasty
consists of one king only, the verb changes from collective to singular: i - a 5 . Occur-
rence^: ii 42-44: 23 l u g a l m u - b i 24,510 i t u 3 u 4 3 u4 \ i b - a 5 ; iii 35-36:
12 l u g a l m u - b i 2,310 i b - a 5 ; iv 15-16: 3 [ l u g a l ] m u - b i 3 6 1 [ + x f b - a 5 ] ;
iv 34-35: 8 l u g a l m u - b i 3,195 f[b-a 5 ] ; iv 41-42: 1 l u g a l m u - b i 6 § u - s i
i b - a 5 ; M v 13-14: [x] l u g a l m u - b i [y] i b - a 5 ; v 19-20: [x] l u g a l [mu-bi
y + ] 2 0 i b -as; 54 v 31-32: [ x + ] 5 l u g a l [ m u - b i ] 136 i (!)-a 5 ; 55 v 40-41: 1
l u g a l m u - b i 100 l - a 5 ; v i 20-21: 7 1 u g a l m u - b i 491 f b - a 5 ; v i 2 6 - 2 7 : l l u g a l
m u (!)» 25 l - a 5 ; vii 11-12: 11 l u g a l m u - b i 181 f b - a 5 ; vii 22-23: 5 l u g a l
m u - b i 30 i b - a 5 ; vii 50-51: [ x + ] 11 l u g a l [ m u - b i y] u 4 40 f b - a 5 ; viii 5-6:
1 [ l u g a l ] m u - b i 7 s u - § i 6 'u*1 [. . l-a 5 ] ; viii 19-20: 4 l u g a l m u - b i 108
i b - a5 ; viii 44-45: 13 1 u g a 1 m u - b i 203 i b - a 5 .
3. Formula for introducing single rulers: N . 6 7 m u x i - a s . Occurrences: i 46-
62
The passage is broken, but the space available does not allow us to restore the full
phrase g [u - 1 u - u mki *H u k u l b a - a n - s i g ] . The scribe seems therefore to have
jumped a line, *i5t u k u l b a - a n - s i g . If this is correct, it suggests that WB's origi-
nal was written with lines of only half the length of WB's, i.e., with the length of line usual
in the other manuscripts.
63
The form u r iki - m a is hardly intended as a locative, for the text does not as a rule
express locative. More likely it is a mere slip due to the fact that the scribe was more con-
versant with the genitive form of the name, as this form naturally occurred more fre-
quently than the nominative.
54
Remnant of an earlier form of the text in which this dynasty numbered more than
one ruler (cf. pp. 99, n. 174, and 102, n. 186).
58
1 b - a5 would have been correct.
56
The omission of - b i here is probably due to influence from 1. 25, m u 25 1 - a 5 .
67
The subject element is used sporadically when the name is followed by a patronymic, e.g.
b a - l i - i ^ d u m u e - t a - n a - k e i , as-ka d u m u e n - m e - e n - b d r a - g e - s i -
k e i , or by other epithets, e.g. d gi 1 g a m e s (dGis-BiL-GA-MEs) a b - b a - n i l i l - l d ,
e n k u l - a b - b a - k e 4 . The subject element may even be found before epithets with
- a m where it is incorrect (see Poebel, GSG § 152): k i § iki p u z u r - d s t n d u m u k u -
d
b a - b a 6 - k e 4 l u g a l - a m . Often, however, it is omitted, e.g. b a r - s a l - n u n - n a
d u m u e n - m e - n u n - n a , u r - d n u n - g a l d u m u d g i 1 g a m e s (doi§-BiL-GA-MEs);
and it is never used after a name which stands alone without epithet or patronymic, in other
words where there is not an obvious genitive relationship (cf. e.g. writings such as u r -
d
u t u [vii 21], although the name contains a genitive, with d§ u 1 - g i d u m u d u r -
d
n a m m u - k e4 [viii 11], where the genitive is "obvious")- This peculiar use of the sub-
ject - e belongs to a late stage of Sumerian at which the subject element had virtually
gone out of use. It occurs as a "fossil" only, namely in the sign - k e 4 , which originally
was used after subjects containing a genitive (N. - (a) k - e) but in this late period had
come to he considered a mere genitive indication (cf. Poebel, OSG §§373fT., where the
rules governing the genitive in late periods are set forth in detail).
oi.uchicago.edu
TEXTUAL PROBLEMS 31
47: g u l - l a d n i d a b a - a n - n a - r d a1 s i k i l 5 8 m u 960 l - a 5 ; ii 7-8: k a - l f -

b u - u m m u 960 i - a 5 ; i i 9 : q a - l u - m u m u 840 i - as ; passim.
P2
1. Formula for change of dynasty: A k i b a - [ g u l ] n a m - l u g a l - b i B ki -s&
b a - t i i m . Occurrences: ii 1-3: [k i § i ki b a - g u l ] n a m - l u g a [ l - b i ] 6 - a n -
n a - ^ b a - t u [m] ; iii 1-2: [ u n u g k i b a - g u l n a m - l u g a l - b i ] u r i [ki- s e
b a - t u r n ] ; iii 17-19: u r i k i b a 5 9 - [ g u l ] n a m - 1 u g a l - [b i] a - w a - a n ( ! ) 6 0
b a-1 u m .
2. Formula for dynasty total: x l u g a l m u - b i y m u f b - a 5 . Occurrence:
iii 14-16: 4 1 u g [a 1] m u - b i 171 [m u] i [b - a5] . Compare the form of the final
totals: x i l - 4 : SU-NIGIN r 5 r [ l u g a l ] m u - b i 18[000+x+]9 m u [y i t u z u4] i b -
[a 5 ]; xi 7-9: Û-NIGIN 1 22 l u g [a 1] m u - b i 2,61[0+x m u ] 6 i t u 13[+x] u 4 f b -
W ;xi 12-14: SU-NIGIN [1S} 1 u g a 1 m u - b i 3 9 6 m u f b - a 5 ; xi 17-19: SU[-NIGI]N
3 l u g a l m u - b i 356 m u i b - a 5 ; xi 22-23: [SU]-NIGIN 1 l u g a l m u - b i 7 m u
[i-a 5 ] ; xii 1-3: [SU-NIGIN X] l u g a l [ m u - b i y+]137 [mu] i b - a 5 ; xii 6-8: su-
^ I G I N 1 21 l u g a l m u - b i 125 m u 40 u 4 f b - a5 ; xii 12-14: [SU-NIGIN] 11 l u g a l
[m u - b] i 159 m u i b - a 5 .
3. Formula for introducing single rulers: N. 61 x m u i - a 5 . Occurrences: i 7-8:
[ q a ] - l u - m u - u m - e [900] m u i - a 5 ; i 9-10: [ z ] u - q a - q i 4 - i p - e '9001 m u
i - as ; passim.
p3
1. Formula for change of dynasty: not preserved.
2. Formula for dynasty total: x [ l u g a l ] m u - b i y [ ] i [b (?) - a5] . Oc-
currence: ii 7-9: 23 [1 u g a 1] m u - b i 18,000[+ x] i t u 3 u 4 3 i [b - a5] .
3. Formula for introducing single rulers: N. 62 x m u i - a 5 . Occurrences: i 1-2:
58
Probably not a proper name; see p. 77, n. 40.
59
Poebel reads this sign as m a in PBS IV 1, p. 76, and translates "the kingdom of
Ur passed to Awan" ( u r i m - a ( k ) n a m - l u g a l - b i etc.); but his copy shows a clear
b a ! On the restoration b a - [g u 1] see p. 46, n. 93.
60
The scribe omitted ki- s e .
61
The subject element - e is used regularly (1) when the name stands alone: [q a] -
l u - m u - u m - e , [ z ] u - q a - q i 4 - i p - e , e n - m e - n u n - n a - k e 4 (the only excep-
tions are ra] - 1 a b - b a and r& - b a1); (2) when the name is followed by a patronymic
only: a r - w i d u m u m a s - d a - k e 4 , w a - l i - i l j d u m u e - t a - n a - k e 4 , m e -
ldm-kisiki dumu e n - m e - n u n - n a - k e 4 , b a r - s a l - n u n - n a d u m u en-me-
n u n - n a - k e 4 , rsuMUG1 s a ^ r a u g 1 d u m u b a r - s a l - n u n - n a - k e 4 , [ t i - i z ] -
r
k & r1 d u m u b a r - s a l - n u n - n a - k e 4 . If the name is followed by other epithets
or by a note the scribe wavers. He omits - e with a simple epithet: l u g a l - b a n - d a
s i p a , where s i p a - d e would have been correct. When the name is followed by an
epithet ending with - a m he writes m e s - k i - i n - g a - § e - e r d u m u d u t u e n - a m
l u g a l - a m , which is correct (GSG §152), but he also writes incorrectly e n - m e -
e r - r u - k a r [ d u m u ] m e s - k i - i n - g a - s e - e r - k e 4 l u g a l u n uki - g a 1 u u n uki
-ga m u - u n - d a - d u - a lu g a l - a m .
62
Except for r a r v - b u - u m ' d u m u 1 m a s -EN- d a - rk e j the subject element does
not appear in any of the names preserved.
oi.uchicago.edu
r
[ k & - H - b u ] - u m [x m ] u l - a 5 ; i 3-4: $ ] - b a [x] m u \-W,i&-6: ^-tab-
b a 720[+ x] m u i - a 5 ; passim.
P4
1. Formula for change of dynasty: A ki g i 5 t u [ k u l b a - s i g ] n a m - l u g a l -
[bi] B[ k i -§e b a - t t i m ] . Occurrence: i 2-5: u n u g k i g U t u [ k u l b a - s i g ] n a m -
1 u g a 1 - [b i] k i - s u - l u - t i b - [ g a r ] g u - 1 i - u m[kri- [§ & b a - 1tim ] .
2. Formula for dynasty total: [ x l u g a l ] m u - b i y [ ] . Occurrence: i 1: [x
l u g a l ] m u - b i 24[+ y ] .
3. Formula for introducing single rulers: N. 63 x m u 1 - [a 6 ]. Occurrence: i 6-7:
i m - rt a1 - [a] 3 m u i - [a 6 ].
P*
1. Formula for change of dynasty: A ki b a l a - b i b a - a n - k u r n a m - l u g a l -
b i B ki -§& b a - a n - [ t t i m ] . Occurrence: iv 7: ru r i kil b a l a - b i ba-an-
ktir n a m - l u g a l - b i 1-si-inki-§ e b a - a n - [ t t i m ] .
2. Formula for dynasty total: x l u g a l - e - n e mu 6 4 y i n - a k - e § . Occur-
rences: iv 6: 5 l u g a l - e - n e ( I ) 6 4 117 i n - a k - e & ; iv 24: [x l u g a 1-e]- rn e1
[ m ] u 2 2 5 i t u 6 i n - ' a k 1 - [e g ] .
3. Formula for introducing single rulers: N. 65 m u x i n - a 5 . Occurrences: i 9:
q ^ - l u - m u - u m [ m u x i n - a 5 ] ; i 10: z u - q d - k i - i p [ m u x i n - a5] ; i 2 1 :
r
i P - t a - s a - d u - u m m u [x i n - a 5 ] ; i v l 3 : [ u r - d n i n ] - u r t a m u 28 i n - ' a s 1 ;
passim.
Pe
2. Formula for dynasty total: [x l u g ] a l [ m ] u - b i y i b - a 5 . Occurrence:
compare the final total in ii 1-3: [SU-NIGIN x 1 u g] a 1 [m] u - b i 125 i b - a 5 .
3. Formula for introducing single rulers: N , x m u l - a j . Occurrences: i 2 - 5 :
d
l [ r - ] u [ r - ] d u m u n u - m u - [ ] 6 m u l - [a5] ; i 6-7: s u - m u - a - b u -
u[m] i t u 8 i - a 5 ; i 8 - 9 : [i-k] u - u n - p i - i § t a r [x m u ] V-a*.
63
The fragment has not preserved a single complete name, so we do not know whether
it used the subject element or not.
64
In iv 6 the scribe omits m u . This omission, however, is obviously due to negli-
gence in copying, for in iv 24, where the formula occurs' again, traces of m u are pre-
served.
86
The subject element is in this text used differently in the different sections. In col. i
the subject element - e is used when the name is followed by a patronymic: a r - w i - ii
d u m u m a § - d & - k e 4 , b a - rl i - i h1 d u m u e - r t a1 - n a - k e4, m e - l d m - r k i s i1ki
d u m u e [n] - m e - n u n - rn a1 - k e*, SUMTJG s & - m u g d u m u bar-sal-nun-
n a - k e 4 , r t i 1 - i z - k d , r d u m u SUMUG s & - m u g - k e i . When the name stands alone
- e is not used. The only exception is [i] 1 - k u - u m - e . Similarly in iv 1-14 - e is used
when the name is followed by a patronymic: § u l - g i d u m u u r - d n a m m u - k e i
etc., but not when the name stands alone. It makes no difference whether the name is
followed by an epithet with 4 m o r not, for we find u r i k i - m a u r - d n a m m u - k e 4
l u g a l - ^ m (iv 1) but also [ I J - s i - i n ^ - n a i § - b i - d i r - r a l u g a l - k m (iv 8).
From 1. 15 on, however, all the names seem to be without - e , even when they are fol-
lowed by patronymics. We must here no doubt recognize a different hand; the scribe who
added the last section of the list did not bother with this grammatical feature (cf. p. 135).
oi.uchicago.edu
TEXTUAL PROBLEMS 33
U
1. Formula for change of dynasty: Aki gUt u k u 1 b a - s l g n a m - l u g a l - b i
B -§e b a - t u r n . Occurrences: iii 8: [urf1**] * m i t u k u l b a - s i g [ n a m -
ki
l u g a l - b i a - w a - a n k i - § ^ b a - t t i m ] ; iv 5-7: k i § i k i g i i t u k u l b a - s l g
n a m - l u g a l - b i b a - m a - z i k i - £ d b a - t t i m ; v l : [uri k i gii t u k u l b a - s i g
n a m - l u g a l - b i ] a d a b a ^ - ^ e 1 [b] a - t u r n ; v 8-10: a d a b a k l * i S t u k u l
b a - rs i g1 n a m - l u g a l - b [ i ] m a - e r i k i - § e b a - t t i m ; vi 2-4: u4 - k u § uki
* i S tukul b a - s l g n a m - l u g a l - b i ki§i k i -§& b a - t t i m ; ix4-8: k i - s u -
l u - t i b - g a r g u - t i - u mki [giil t u k u l b a - s l g [n a ra - 1] u g a 1 - b i
[unugki-§e] ' b a - t t i m 1 .
2. Formula for dynasty total: x l u g a l m u - b i y ib-a 5 . 6 6 Occurrences: iii
6-7: [x] l u g a l [ m u - b i y+]51 i b - a 5 ; iv 2-4: 4[+x] l u g a l m u - b i 3,792 i b -
a 5 ; v 6 - 7 : 1 l u g a l m u - b i 90 r ib 1 -[a s ]; vi 1: [x lugal] m u - b i 99 'fb-as 1 ;
ix2-3: 21 l u g a l m u - b i 124 u4 40 !
3. Formula for introducing single rulers: N . 6 7 x m u l - a 5 . Occurrences: iii 2-3:
[ e J - l u M u [x] mu i - a 5 ; iii 4-5: [ b a - l u ] - l u [x] mu i-a 5 ; passim. Twoneigh-
boring passages, viii 4-7 and 20-21, have a different formula: N. x m u i n - a5 .m
Occurrences: viii 4-5: i n - g i 4 - § u § 7 mu i n - a 5 (1. 3 still has l-a 5 ) ; viii 6-7:
z a r - a r - l a - g a - b a 6 m u i n - a5 (1. 9 has 1 - a6 again); viii 20-21: [ ] - a n -
g a b [x m u] i n - a 5 .
L2
2. Formula for dynasty total: not preserved.
3. Formula for introducing single rulers: N.69 x mu l - a 5 . Occurrences: i 1-2:
t e - [ x+]600m[u i - a 5 ] ; i 3 - 4 : p d - l a - k i - n a - r t i - i m 1 900 mu i - a 5 ; i
5-6: n a - a n - g i s - l i - i § - m a [ x ] f m u i - a 5 '.
S
1. Formula for change of dynasty: A k i -a(k) b a l a - b i b a - k i i r n a m -
l u g a l - b i B k i -§e b a - t t i m . Occurrences: obv. 8: u 4 - k u s t i k i - a b a l a - b i
b a - k u r n a m - l u g a l - b i k i § i k i - § e b a - t t i m ; obv. 19: ki§i k i b a l a -
b i b a - k [ i i r n a ] m - l u g a l - b i unug k i -§e* b a - t t i m ; obv. 22: u n u k i - g a
b a l a - b i b a - k t i r [ n a m - l u g a ] l - b i a - g a - d e k i - s e b a - t t i m ; rev. 9-
10: a - g a - d & k i b a l a - b i b a - k t i r n a m - l u g a l - b i u n u g k i - s e b a -
t t i m ; rev. 17-18: u n u k i - g a b a l a - b i b a - k t i r n a m - l u g a l - b i k i -
s u - l u - t i b - g a r r g u - 1 i1- u m k i - § £ b a - t t i m .
2. Formula for dynasty total: x l u g a l - e - n e m u - b i y i n - a k - e s .
If the dynasty consists of one king only the words l u g a l - e - n e and i n - a k - e §
66
This text seems to keep the collective i b - a6 even when the dynasty has only one
ruler.
87
The subject element is not used. Only two names are exceptions: p t i z u r - d s ! n
d u m u k u - db a - b ae - k e4 l u g a l - a m and u r - dz a - b a4 - b a4 d u m u p t i z u r -
d
s t n - k e4.
68
On the significance of this change see pp. 54 f.
69
The subject element is not used in the two names preserved.
oi.uchicago.edu
34 T H E STJMERIAN K I N G L I S T
change to singular 1 u g a 1 and i n - a 8 . Occurrences: obv. 7: 6 l u g a l - e - n e

m u - b i 99 i n - a k - e § ; obv. 18: 8 1 u g a l - e - [ n e ] m u - b i 5 8 6 i n - a k - e § ;
obv. 21: 1 l u g a l m u - b i 25 i n - a 5 ; rev. 8: 12 l u g a b - e - n e m u - b i 197
i n - a k - e § ; rev. 16: 5 l u g a l - e - n e m u - b i 26 i n - a k - e & .
3. Formula for introducing single rulers: N.70 x mu i n - a 5 . Occurrences: obv.
2: u n - d a - l u - l u 12 mu i n - a 8 ; obv. 3: u r - u r 6 m u i n - a 5 ; obv. 4:
p t i z u r - d n i r a f c 2 0 m u i n - a 5 ; passim.
Sui
1. Formula for change of dynasty: Aki gi5t u k u l b a - s l g n a m - l u g a l - b i
B k i -§e b a - t t i m . Occurrences: v l : [ kigi8
tukul ba-slg nam-lugal-
b i u 4 - k u § t i k i - § e b a ] - t u r n ; v 18-20: u 4 - ku§ti k i tgi5]t u k u 1 b a - s i g
n a m - l u g [ a l ] - b i m a - e r i [ ki -£]e b a - t t i m ; vi 10-12: k i § i k i g i H u k u l b a -
s i g n a m - l u g a l - b i unug k i -§& b a - t t i m ; vi 19-21: u n ug k i «[i5t u k u 1
b a - s l g ] n a m - l u g a l - [ b i ] a - g a - d eki [- § e b a -1 ti m ] ; vii 5-7: a -
g a - d [e]ki gii t u k u l b a - s i g n a m - ' l u g a P - b i u n u gki - § e b a -1 ii m ;
viii 7-9: u r fki gii[t u k u l b a - s l g ] n a m - l u g a l - [ b i ] l - s i - i n k i - s e [ b a -
ttim],
2. Formula for dynasty total: x l u g a l m u - b i y mu i - a8 (?).71 Occurrences:
iv 14: [x] l u g a l [ m u - b i y mu i-a 5 (?)]; v 16-17: 6 l u g a l m u - b i 110[+x+]
3 m u 1 - a5 (?) ; vi £-9: 7 1 u g a 1 m u - b i 485 m u 1 - a5 (?) ; vi 17-18: 1 1 u g a 1
m u - b i 25 m u rl - a8 (?) 1 ; vii 3-4: [x+]3 1 u g a 1 m u - b i 161 m u i - a5 (?) ; viii
5-6: 2[+x l u g a l ] f m ^ - b i 120[+y m u i - a 6 ( ? ) ] .
3. Formula for introducing single rulers: N.72 x mu i - a 5 ( ? ) . Occurrences:
vi2-3: i § - m e - d § a m a § 11 mu l-a 5 (?); vi 4-5: S u - i - l i - S u 15 m u i- a8 (?) ;
vi 6-7: s f - m u - d a r 7 m u i - a 8 ( ? ) ; passim.
Su2
1. Formula for change of dynasty: Aki b a - g u l n a m - l u g a l - b i B k i -§e
b a - t t i m . Occurrences: ii 11-13: k i § i k i b a - g u l n a m - l u g a l - b i 6 - a n -
n aki - § e b a -1 ti m ; in 9-11: u n u gki b a - g [u 1] n a m -1 u g a 1 - [b i] u r i [ki -
§e ba-ttim].
2. Formula for dynasty total: x l u g a l m u - b i y m u i - a 5 . Occurrences:
ii 7-10: 23 1 u g a 1 m u - b i 20,940 m u 30 i t u 3 u4 2\ 1 - a 6 ; iii 6-8: 12 1 u g a 1
m u - b i 3,588 m u i - a 8 .
70
The subject element is used regularly when the name is followed by a patronymic,
otherwise not.
71
According to Scheil in RA XXXI (1934) 149 the text uses NA throughout instead of
a«. This spelling would be surprising even for a text written in Susa, and the present writer
is inclined to believe that the sign which Scheil reads as NA is in reality a careless form of
AK peculiar to the scribe who wrote SUL The two signs are not very different: *&• (AK;
MDP XXII, No. 42:9) and *^* (NA). It is unfortunate that Sui has not preserved the
line concerning Aka, so that we could see how the scribe wrote an indisputable AK.
72
The subject element is not used with names which stand alone and occurs only occa-
sionally where the name is followed by a patronymic: [ar]-wi-um [dumu m a §] ~
d a - k e 4 , [b a -1] i - i Jj d u m u e-[t]&-na-ke«, but £ u - ds t n d u m u i - § u - i 1,
u r - «% i g i r d u m u u r - n i e l n , l u K a l - m e - U m d u m u u r - ^g i g i r .
oi.uchicago.edu
TEXTUAL PEOBLEMS 35
3. Formula for introducing single rulers: N.73 x m u i - a 8 . Occurrences: ii 2-3:
e n - m e - b & r a - g a - e - s i 900 mu i - a 5 ; ii 4-6: a 5 - k a . . . . 6 2 5 mu i-a 8 ;
iii 2-3: t i - [ ] 75 m u 1 - [a 6 ]. The last reign preserved on the fragment has a dif-
ferent formula:74 N. mu x i - a 5 . Occurrence: iii 4-5: l u g a l - [ ] m u 7 i - a 8 .
SUM-475
1. Formula for change of dynasty: A k i * i S t u k u l b a - s l g n a m - l u g a l -
b i B k i -§e b a - t u m . Occurrences: ii 1-3: [ ] k i «"[tukul b a - s l g n a m ] -
l u g a l - [ b ] i [ki§i k i ]-§S b a - t u m ; ii 23-25: [ki§i k i « i s tukul] b a - s l g
[ n a m - l u g a l ] - b i [unug k i -§& b a ] - t u m ; ii 32-34: [ u n u g ^ H u k u l b a ] -
s i g [ n a m - l u g a l - b ] i [ a - g a - d e ^ - S e ^ b a - t u ] m ; iii 17-20: a - g a - d e k i
* i S tukul b a - s l g n a m - l u g a l - b i u n u g k i - § e b a - t t i m ; iii 34-37: unug k i
8iS
t u k u l b a - s i g [n] a m - l u g a l - b i [ m ] a - d a g u - t u - u m k i - § 6 b a -
t u m ; iv 3-6: m a - d a g u - [ t u - u m k i ] * " t u k u l b a - [ s l g ] n a m - l u g a l -
[bi] u n u g k i - § e [ b a - t u m ] ; iv 14-16: u n u g[ki * l i t u k u l b a - s i g ] n a m -
[ l u g a i - b i ] u r f [ k i - § e b a - t t i m ] ; iv 36-38: u r f - m a k i « u t u k u l b a - s l g
76
2. Formula for dynasty total: x l u g a l m u - b i y m u 1 - a 6 . Occurrences:

ii 20-22: [x lug]al [mu-bi y] mu [l]-a 5 ; ii 30-31: [x lug]al [mu-bi y mu]
i-a 5 ; iii 14-16: r 3l+x lugal] m u - b i 177 mu i - a 6 ; iii 31-33: 3 l u g a l m u -
b i 57m u l - a B ; iv 33-35: 5 1 u g a l m u - b i 123 m u l - a 5 .
3. Formula for introducing single rulers: N.77 x mu i - a 5 . Occurrences: ii 5-9:
k u - db a - b a 6 . . . . 100 m u 1 - a5 ; iii 1-2: [m a - a n] - i § -1 i - [s u x] m u 1 - a* ;
iii 3-5: 8 a r - f c a l P - f i a r - r i 25 m u l - a 5 ; i v l - 2 : i u g a l - [ ] 25 m u
[ l - a 5 ] ; iv 18-20: u r - d [ n a m m u ] 18 m u i-[a 5 ]; iv 21-23: d £ul-g[i]
. . . . 48 m u 1 - W ; passim. A single section, iii 22-30, has another formula:78 N.
m u x 1 - a 5 . Occurrences: iii 22-24: u r - n i g 1 n . . .. m u 15 i - a 5 ; iii 25-27:
u r - giSg i g i r . . . . m u 7 i - a* ; iii 28-30: u r - du t u . . . . m u 25 1 - a 5 .
K
1. Formula for change of dynasty: absent.
2. Formula for dynasty total: x l u g a l - e - n e b a l a A k i m u y i n - a 6 - m e -
e S . Occurrences: rev. i 13: 1[1 l u g a l - e - n e b a l a TiN-TiRki mu x i n - a 5 -
m e - e § ] ; i i 8 : 3 1 u g a l - e - n e b a l a k u r a - a b - b a mu 2 3 i n - a 5 - m e - e § ;
ii 12: [ 3 l u g a l - e - n e ] b a l a b t t - m b a - z i mu 20 i t u 3 i [ n ] - a 5 - m e - e s \
3. Formula for introducing single rulers: N.79 m u x i n - a 5 . Occurrences: obv.
78
The subject element is not used in the names preserved.
74
Cf. the similar cases presented by Li, which changes from 1 - a8 to i n - a6, and Su$+4,
which changes from x m u to m u x ; see pp. 54 f. for full discussion.
76
For the line numbering see p. 10, n. 24.
78
The text seems to have s u h u § fk e1 [ ] m u - u n (?) [ ], which may be restored
as s u k u § fk e1 - [e n - g i - r a] m u - ru n1 - [s i r], "The foundation of Sumer was torn
out."
77
The subject element is not used.
78
Cf. the similar cases presented by SU2, which also changes from x m u t o m u x,
and by Li, which changes from 1 - a» to i n - as; see pp. 54 f. for full discussion.
79
The subject element is not used.
oi.uchicago.edu
i 1: [N. m u x] i n - a 6 ; ii 2: m b a 1 i fr . . . . [mu x i n - a5] ; ii 3: m e n - m e n -

n u n - n a [ m u x i n - a 5 ]; rev. ii 5: m d 6 - a - m u - k i n - z h . . . . i t u 3 i n - a 5 ;
ii 7: m d k a § - § u - u - n a d i n - a k i . . . . m u 3 i n - a 5 ; passim.
G
2. Formula for dynasty total: not preserved.
3. Formula for introducing single rulers: N. 79 x m u i - a 5 . Occurrences: vii 3:
[i n - k i - § u § 6 m] u i - a 5 ; vii 4: [s a (?) - a r -1 a - g a b 6] m u 1 - a [5] ; vii 5:
[§ u 1 - m e - e] 6 m [u 1 - a5] ; vii 6: [e - 1 u -1 u - m e - e S] 7 m [u 1 - a5] ; vii 7: i -
n i - m a - b a - k i - e] § 5 [m u 1 - a5] ; vii 8: [ i - g e 4 - e § - a - u ] § f 6 ] [m u i - a 5 ] .
J
1. Formula for change of dynasty: A ki g i S t u k u l b a - s i g nam-lugal-
b i B k i -§& b a - t u m . Occurrences: obv. 1-4: [ k i - s u - l u - u b - g a r ] g u - t i -
u m k i ieikH u k u l b a - s l g [ n a m ] - l u g a l - b i [ u n u g] k i - § e* b a - t u m ; rev.
6-8: [u] r fki *iSt u k u 1 b a - s 1 g [n] a m - l u g a l - b i [ i ] - s i - i n k i - § & b a - t u m .
2. Formula for dynasty total: 1 l u g a l m u - b i x l-a 5 . 8 0 Occurrences: obv.
7-8: 1 l u g a l [ m u - b i 7] i t u 6 u 4 15 1 - W ; rev. 4-5: 5 l u [ g a l ] [m] u - b i [109
fb-a5].80
3. Formula for introducing single rulers: N. 81 m u x [ l - a 5 ] . Occurrence: obv.
5-6: d u t u - k 6 - g & 1 • . • • m u 7 i t u 6 u 4 [15 i - a 5 ] .
A glance through this list will show t h a t our three formulas v a r y not in-
considerably from one source to another.
I. The formula for change of dynasty occurs in three principal82 forms:
a) In WB, P 4 , Li, Sui, Su3+4, and J:
A k i g i S t u k u l b a - s l g (WB: b a - a n - s l g ) n a m - l u g a l - b i B k i - § &
b a - t tim
"The city A was smitten with weapons;83 its kingship was carried84 to the city B . "
80
In totals with more than one ruler the text presumably used i b - as.
81
There is only this one example of the formula for introducing single rulers. As the
name is followed b y l u g a l - a m , the subject element is (correctly) omitted.
82
For the sake of convenience and greater perspicuity we have here grouped WB with
P4, Li, Sui, Su3+4, and J, although strictly speaking it represents a separate form differing
from the others by having b a - a n - s i g instead of b a - s! g . In the same way we have
grouped S and P 5 together, although P6 writes b a - a n - k t i r and b a - a n - t u m while
S has b a - k a r and b a - t ti m . To separate WB from P4, Li, Sui, Su3+4, and J and to
separate S from Pg merely because of this small and insignificant (see p. 41) variation
would quite unnecessarily blur the basic grouping under * i S t u k u l b a - s i g , b a - g u l ,
and b a l a - b i b a - k t i r .
83
Giiterbock in ZA n.F. VIII (1934) 2 makes «iSt u k u l subject and translates "die
Stadt x schlug die Waffe"; but since the interest of the author of the King List so clearly
centers in the ruling cities (see below, n. 86), it is more natural to have the city name in
the most prominent position, i.e., as subject.
84
This formula is rendered "the kingdom of . . . . passed to . . . . " or the like by al-
most all scholars. But an occurrence of the same formula in an inscription of Utu-hegal
oi.uchicago.edu
TEXTUAL PROBLEMS 37
b) In P 2 (?) and Su 2 :
A k i b a - g u l n a m - l u g a l - b i Bki - § e b a - t u m
"The city A was destroyed; its kingship was carried to the city B . "
c) In S and P 6 :
A k i - a ( k ) (P 6 : A ki ) b a l a - b i b a - k u r (P 5 : b a - a n - k u r) n a m -
l u g a l - b i B k i - § e* b a - t u m (P 5 : b a - a n - t u m )
"The turn (to reign) of the city A was changed; its kingship was carried to the
city B . "
One text, K, omits the formula altogether.
2. The formula for dynasty total occurs in six different forms:
a) In Sui, Su2, and SU3+4:
xlugal mu-bi y mui-a5
"x kings reigned85 its (the city's) 86 years, y years."
b) I n P 2 :
xlugal mu-bi y muib-a5
"x kings reigned its (the city's) years, y years."
(RA IX [1912] 112-13 i 1-6), which contains our phrase: g u - t i - u m ki 1u

n a m - l u g a l k e - e n - g i - r a * k u r - § 6 b a - t u m - a , "Gutium . . . . , who had
carried the kingship of Sumer off to the mountains," leaves no doubt that DU represents
a transitive verb. We must therefore read b a - t u m ( ! ) and b a - a n - t d m (!) , "was
carried'' ( b a - t u m is the normal passive preterit 3d sg.; b a - a n - t u m is the corre-
sponding form used by the scribes who followed what Poebel calls "post-Sumerian sys-
tem A." See AJSL L [1933/34] 170). The kingship was thus not imagined as personal
and moving from city to city according to its own will but as an inanimate thing—prob-
ably closely tied up with its symbols, the scepter and crown—which had to be taken or
carried from one place to another.
88
We retain the traditional rendering of a (k) as "reigned," although, as Poebel has
pointed out in OLZ XV (1912) 291 f., a(k) means only "to do," "to perform." "To
reign" is n a m - e n — a (k) , "to perform lordship," or n a m - l u g a l — a ( k ) , "to
perform kingship." Poebel is inclined to believe that a (k) in our formulas takes m u ,
"years," as object: "so und so viele Jahre hat er gemacht." We consider his other explana-
tion, that "ein nam-lugal zu all den in-ag-Satzen dem Sinne nach zu erganzen . . . . ist,"
much more satisfactory.
86
The suffix - b i must here, as in the preceding formula, refer to the city in question
and cannot, as is generally supposed by translators, refer to its kings. This is plainly shown
by the fact that - b i is retained even when the dynasty consists of only one ruler and
the verb changes to singular (e.g. WB v 40-41: l l u g a l m u - b i 100 i - a5 ; WB viii
5-6: 1 [ l u g a l ] m u - b i 7 § u - § i 6 W [. . l - a 5 ] ; S obv. 21: 1 l u g a l m u - b i 25
in-a 5 ). If - b i referred to the kings, we should here have had 1 l u g a l m u - n i (!) x
1 - as (or i n - a 6 ). The list is primarily concerned with the kingship of cities, not of single
persons, as is clearly expressed in the framework as a whole: ". . . . The kingship was
in Kish. In Kish <3a. .ur(?) became king and ruled 1,200 years; N. ruled 960 years;
. . . . 23 kings ruled its 24,510 years, 3 months, and 3f days. Kish was smitten with weap-
ons; its kingship was carried to E-Anna(k). In E-Anna(k) "
oi.uchicago.edu
c) In WB, P«, Li, and J(?):

xlugal mu-biy ib-a6
u
x kings reigned its (the city's) y years."
d) In S:
xlugal-e-ne mu-bi y in-ak-e§
"x kings reigned its (the city's) y years."
e) InP 5 :
xlugal-e-ne mu y in-ak-e§
"x kings reigned y years."
/) In K:
x l u g a l - e - n e b a l a Aki mu y i n - a 5 - m e - e §
"x kings of the dynasty of A reigned y years."
3. The formula for introducing single rulers occurs in four forms:
a) InWBand J(?):
N. m u x i - a5
"N. reigned x years."
b) In P2, P3, P4» Pe, Li, L 2| Sui, Su2, Su3+4, and G:
N. x m u l-a 5
e) I n P 5 a n d K :
N. mu x i n -a 5
d) In S:
N. x m u i n - a5
U
N. reigned x years."
The origin and probable causes of most of these variants are fairly clear. We
notice first a certain number of variants which must be due to the application
of the grammatical and orthographical rules taught in the scribal schools of the
Isin-Larsa and post-Hammurabi periods. The variants b a - s i g : b a - a n -
s i g , b a - k t i r r b a - a n - k d r , and b a - t t i m : b a - a n - t u m of the
formula for change of dynasty and the variant 1 - as : i n - ag of the formula
for introducing single rulers obviously have their origin in the late custom
of expressing the subject - n - of the verbal form in writing. 87 The other
variant in the formula for introducing single rulers, m u x : x m u , corre-
sponds similarly to the change from the older Sumerian usage, according to
which the numeral follows the word for that which is counted, to the later
87
The date of this custom is discussed in detail on pp. 131-35.
oi.uchicago.edu
TEXTUAL PROBLEMS 39
usage in which the numeral precedes.88 Application of later grammatical

rules is likewise responsible for the variant l u g a l . . . . f b - a 5 : l u g a l -
e - n e . . . . i n - a k - e fi : 1 u g a 1 - e - n e . . . . i "n - as - m e - e S in the for-
mula for dynasty total. The old collective plural, which in the noun is ex-
pressed by the naked stem, in the verb by the form i - b - LAL, grew obso-
lete in the Isin-Larsa period,89 in which i-b-LAL was chiefly used for
3d sg. of the causative. It is therefore understandable that copyists of this
period should correct 1 u g a 1. . . . i b - a5 to the form which they consid-
ered correct, l u g a l - e - n e . . . . i n - a k - e § , and that still later copy-
ists should change to l u g a l - e - n e . . . . i n - a 5 - m e - e § .
Other variants, however, seem completely independent of any rules what-
ever, so much so indeed that we must ascribe their origin to idiosyncrasies
in individual copyists. A variant of this type is that in the formula for change
of dynasty, Aki gi§t u k u 1 b a - s i g n a m - l u g a l - b i Bki-§& b a - t i i m :
A k i b a - g u l n a m - l u g a l - b i B k i -se b a - t t i m : A k i -a(k) b a l a - b i
b a - k t i r n a m - l u g a l - b i B k i -§e b a - t 6 m . The three forms say vir-
tually the same thing, and it is quite impossible to imagine any rule which
would make scribes who found, for example, gi§t u k u l b a - s l g in their
originals change to b a l a - b i b a - k t i r o r t o b a - g u l .
Somewhat less clear are the causes which produced the variant m u - b i y
m u : m u - b i y : m u y in the formula for dynasty total. Such a syntac-
tical construction as x 1 u g a 1 m u - b i y 1 b - a5, "z kings reigned its (the
city's) y years," has close parallels in Old Sumerian.90 It seems possible that
such a form would sound incomplete to later scribes. The numeral has ref-
erence to the preceding m u - b i , "its years"; but, since the word for that
which is counted was in later times generally placed after the numeral, they
may have felt that something was missing and so have added a second m u .
Even if we admit the probability of such a feeling, however, we have no
means to judge its strength. Would it unavoidably suggest to every scribe
that he should add a m u after the numeral, thus acting with the force of a
grammatical rule; or was it merely a matter of personal feeling for the lan-
guage, leaving the individuality of the scribe a considerable amount of play?91
88 See Poebel, OSG §§ 303-5.
89
See the detailed discussion of this change on p. 129. That i - b - LAL is collective was
shown by Poebel, Grammatical Texts (PBS VI 1 [1914]) p. 101.
99
Cf. e.g. E-Anna(k)-tum Brick A (Z>&., p. xlv, No. 8; SAK, p. 26 h) iii 5—iv 9 : 6 -
a n - n a - t u m - e . . . . u m m a k i GIN-se b ! - s i SAQAR-DU-KID-bi 20 m u - d u b ,
"E-Anna(k)-tum . . . . defeated Umma and heaped up its twenty burial mounds."
91
The construction as a whole is so rare that our material does not even permit us to
form an idea of which form was prevalent in the various periods.
oi.uchicago.edu
The last form, x l u g a l - e - n e m u y i n - a k - e § , i s probably due to

the fact that the scribe in question was not aware that the - b i refers to
the city in which the dynasty reigned. He expected to find a simple sum-
mary: "x kings. They reigned y years" and got a smoother text by omit-
ting - b i .
Lastly there is the variant 1 - as in the formula for dynasty total as
against i b - a 5 , i n - a k - e s , or i n - a s - m e - e g . This very curious
use of the singular in passages where the context clearly demands col-
lective or plural becomes intelligible when we notice that it is limited to three
versions, Sui, Su2, and Su3+4, which all come from Elam. In Elamite scribes
we can expect but limited knowledge of Sumerian grammar. That they did
not understand the reasons for the occasional change in their originals from
i - a5 within the dynasties to i b - as in the totals and therefore disregarded
it is not to be wondered at, especially when we consider that their own tongue,
the Elamite, does not stress the difference between singular and plural in
the verb.92
We thus see that some of our variants are due to the application of later
linguistic norms, some have been caused by idiosyncrasies in individual copy-
ists, and one, the variant m u - b i y m u : m u - b i y : m u y, takes an in-
termediate position; it seems dependent upon a change in usage, but as we
do not know how complete this change was we cannot decide whether the
new usage acted as a fixed norm or left the decision with the individual scribe
and his feeling for the language. Lastly, a single variant found only in the
manuscripts from Elam seems to have originated through influence from the
native language of Elam combined with limited knowledge of Sumerian.
All of these variants may be characterized as intentional; they are (1) cor-
rections, (2) changes to forms which pleased the copyist better than those he
found in the original, or (3) systematic disregardings of differences which he
did not understand. This element of intention which characterizes the vari-
ants makes them in certain cases less valuable for our present purpose. There
is very little probability that two copyists should independently chance to
make the same unintentional deviation in copying at the same place in the
list. When we find the same error in two manuscripts we are therefore en-
titled to conclude that both inherited the error from a single original. In the
case of intentional variants, however, the value varies. We can consider it
92
F. Bork in his article "Elam," Reallexikon der Vorgeschichte, hrsg. von Max Ebert, III
(Berlin, 1925) 78 b: "Ein Vergleich der alten Sprache mit dem ]J6zI lehrt, dass auch jene
von Hause aus S i n g u l a r u n d P l u r a l n i c h t u n t e r s c h i e d e n hat."
oi.uchicago.edu
TEXTUAL PROBLEMS 41
improbable that two scribes should both independently feel dissatisfied with,
for example, the phrase gi§ t u k u l b a - s i g which they found in their orig-
inals and independently think u p b a l a - b i b a - k d r a s the suitable sub-
stitute; but we must admit that two scribes who accepted the same norms
for correct language (e.g. that the 3d sg. of the transitive verb was i - n - LAL,
that the plural of personal nouns was - e - n e , etc.) would both be very
likely to notice deviations from these norms in their originals and to correct
them in a similar way (1 - a5 to i n - a 5 , l u g a l to l u g a l - e - n e , etc.).
Variants of the latter type, corrections to fixed norms, accordingly give far
less certainty that sources in which they occur have inherited them from
a common ancestor. They may have been introduced independently by dif-
ferent scribes. The evidence from such variants should therefore be used with
great caution and preferably where it has the support of other evidence of
more reliable type. And evidence from such variants cannot outweigh that
of unintentional variants or variants due to idiosyncrasy if these point in a
different direction.
On a par with the variants due to the application of grammatical rules is
the use of the singular form of the verb in the manuscripts from Elam. The
background of this variant, limited knowledge of Sumerian subject to influence
from the undercurrents of a native language which does not stress the dis-
tinction between plural and singular in the verb, can be presupposed in most
Elamite copyists. We can consider it probable that two such scribes would
independently decide to disregard the difference between i - a5 and f b - a 5 ,
which they did not understand.
We are thus in a position to arrange our variants in order according to their
relative values as indicators of relationship:
1. Of high value
In formula for change of dynasty
giS
tukul b a - s i g WB, P4, Llf Sui, Su3+4, J
b a-gul Ps(?), Su2
bala-bi ba-ktir Pi, S
omitted K
2. Of unknown value
In formula for dynasty total:
rau-bi y mu P2, Sui, Su2, Su3+4
mu-bi y WB, P6, Li, S, J
mu y P8, K
oi.uchicago.edu
3. Of low value
In formula for dynasty total:
iugal . . . . i - a5 Sui, SU2, $U8-M
lugal fb-a5 WB, P2, P6, U J(?)
lugal-e-ne....in-ak-e§ P51 S
lugal-e-ne in-a5-me-e§ K
In formula for introducing single rulers:
l-a« WB, P2, P8, P4, Pa, Li, L2, Su,f Su2,
SU3+4, G, J
i n - a§ P5, S, K
In formula for introducing single rulers:
mux WB, P5, K, J
x mu P2, P3, P4, Pe, Li, L2, S, Sui, Su2| Su3+4, G
We may proceed to consider these groupings of the manuscripts in con-
junction with those at which we arrived on the basis of the Atabba and Agade
variants.
GENEALOGICAL TABLE BASED ON THE FOREGOING VARIANTS
The Atabba variants and the variants in the reigns of the Agade rulers
showed that the manuscripts involved derive from the original as follows:
ATABBA VARIANTS AGADE VARIANTS

Original Original
3L
WB P2 P3 P5 Su, WB L, P2 P3 S Su, S u j ^
In both cases we found on one side the source WB representing the original
fairly closely and on the other a text which we called " B , " in which one or
more errors had been committed. Through this text the remainder of the
sources descended. Although we have called the erroneous text " B " in both
cases it is clear that the Atabba mistake and the Eimush mistake need not
necessarily have happened in the very same copy. We should therefore make
a distinction and—admitting that the two texts may be identical—designate
the text in which the Atabba mistake was made as "Bi" and that in which
the Rtmush mistake was made as " B 2 . "
Comparing our two diagrams we see that, even if we suppose that Bi and
B 2 were different texts, they must at least have belonged to the same line of
descent from the original, B* having derived from B 2 or vice versa. This fol-
oi.uchicago.edu
TEXTUAL PROBLEMS 43
lows from the fact that three of our manuscripts, P2, P3, and Sui, exhibit the
errors of both Bx and B2, which presupposes that the line along which they
descended from the original passed through both these texts. We have ac-
cordingly only three possibilities to take into account:
P2 P 3 So,
We may investigate these one after another to see how each will influence the
relative positions of the manuscripts involved.
Assuming first that Possibility I is correct, we can place the manuscripts
Li, P2, P3, S, Sui, and SU3+4 below the brace, because we know from the Agade
diagram that they derive from B2. This leaves P5, which occurs in the Atabba
diagram. Since this version is only known to be derived from Bi, there is a
priori the possibility that it branched off directly after it had passed this text
and did not follow the others on their way through B2. In reality, however,
that cannot have been the case, as is shown by variants in the formulas. In
the formula for change of dynasty the form giSt u k u l b a - s l g i s found in
WB, in texts which have passed Bi, such as Sui, and in texts which have
passed B2, such as Sui and Li. Being thus represented in both main branches
of the tradition (WB and the B texts), g i 5 t u k u l b a - s l g obviously must
be the form used by the original. The variants b a - g u 1 and b a 1 a - b i
b a - k li r , which occur only in texts derived from Bi and B2, must thus
have arisen later. Now our text P5 (which derives from Bi) has the variant
b a l a - b i b a - k i i r ; this it shares with one other text, S, which descends
from B2. Since this is one of our "reliable" variants (see pp. 40 f.), we may con-
clude that Pfi and S derived it from a common ancestor, which we shall call
"Y." This conclusion is corroborated by another, in itself less reliable, vari-
ant, l u g a l - e - n e . . . . i n - a k - e S , which also is peculiar to S and P 5
and which similarly indicates that these two sources passed through one or
more ancestors peculiar to them alone. Y must, of course, be placed farther
down in the line of descent than Bi and B2, for, as we have seen, other texts
which have passed Bi and B2 (such as Sui and Li) have retained the orig-
inal gi5t u k u l b a - s l g . If Y is to be placed after B2, then naturally Pg,
which derives from Y, also must be placed after B2 and cannot have swerved
off after Bx.
oi.uchicago.edu
If Possibility I is the correct one, that is, if Bi and B2 are different texts
and B2 derived from Bi, the manuscripts with which we have here been deal-
ing must accordingly descend from the original as follows:
W B P P P S Su Su L
2 3 5 l 3*4 l
We may next look at Possibility II, which, like Possibility I, presupposes

that Bi and B2 are different but which has Bi derived from B2. If this possi-
bility is the correct one, we can obviously place the manuscripts P2, P3, P5,
and Sui under the brace, because the Atabba diagram shows that these derive
from Bi:
Original
WB P2 P5 P5 Su,
With the sources S, Su3+4, and Li, however, we must reckon with the possi-
bility that they did not follow the others, for they are only known to derive
from B2 and so may have swerved off directly after that text without passing
Bi. It is therefore necessary to examine these cases more closely.
(1) S. We have just seen that the source P5 can be placed under the brace
below Bi and must therefore have passed both B2 and Bi in its descent from
the original. When we discussed Possibility I we saw, too, that P5 and S
must have a common ancestor or series of ancestors (Y) peculiar to them, in
which their characteristic variants b a l a - b i b a - k u r and 1 u g a 1 -
e - n e . . . . i n - a k - e S arose. It was clear, furthermore, that Y must be
sought farther down in V$s line of descent than B2 and Bi, for as late as in Bi
the change to b a l a - b i b a - k i i r which is characteristic for Y has not
yet happened, as is shown by the fact that Sui, which also has passed through
both B2 and Bi, preserves the original gi§t u k u l b a - s l g intact. If Y is
farther down the line than Bi, however, S, which derives from Y, must also
be farther down than Bi and therefore cannot have swerved off after B2.
(2) Su3+4. When we discussed the variants in the reigns of the Agade rulers
oi.uchicago.edu
TEXTUAL PROBLEMS 45
we saw that Sui and Su3+4 must have had a common ancestor (Su) which had
a lacuna at the end of the Agade dynasty (p. 25). We know also that Sui has
passed through both B2 and Bi. But did Su come before or after Bx in the line
of descent of Sui? The question is easily answered because P2, in which the
total for the Agade dynasty indicates that nothing was missing, derives from
Bi, which must likewise have preserved this dynasty intact. Su, the text with
the lacuna, must therefore follow Bi in the line of descent of Sui, and Su's
other descendant, Su3+4, cannot have swerved off after B2.
(3) Li. With this text the material at our disposal does not permit us to
decide whether it followed the other sources through Bi or whether it swerved
off immediately after it had passed B2. We must therefore leave the ques-
tion open.
We have seen that six of our manuscripts—P2, P3, P5, S, Sui, and Su3+4—
must have descended through both B2 and Bi if Possibility II is the correct
one. As for Li, we are unable to decide whether it followed the others or
branched off after B2. The derivation of the manuscripts would therefore be
as follows:
Original
Y Su N%
WB P2 P3 P5 S Su7~Su^ L,? (or L, here?)
Lastly there is Possibility III. If this is correct, Bi and B2 are only different
names of the same text, which we may designate as "B." Since Bi and B2 are
here identical w6 can obviously place all the manuscripts derived from Bi in
the Atabba diagram and from B2 in the Agade diagram below the brace, that
is, P2, P3, P5, S, Sui, Su3+4, and Li. The reasons which made us assume a
common ancestor for P5 and S (Y) and a common ancestor for Sui and Su3+4
(Su) are naturally not affected by the identification of Bi and B2. The deriva-
tion of our manuscripts according to Possibility III is therefore:
s S u L
P2 P3 P5 Swi 3* .
oi.uchicago.edu
46 THE STJMERIAN KING LIST
Comparing the diagrams for our three possibilities we see that they lead
to practically the same results. In all three cases the sources P2, P3, Ps, S, Sui,
and Su3+4 must have descended through both Bi and B2; only Li may, if the
second possibility should be correct and Bi derive from B2, have taken a dif-
ferent course, swerving off directly after B2, but it is not necessary that it
should have done so. The chances that Li descended through both Bi and B2
are three out of four. Since the various possibilities lead to almost the same
result, we can draw up a single diagram which accounts for them all:
Original
jS B, B2
/ I or B or | \ ,
WB P2 P3 P5 5 Su, S u ^ L, (or, less likely, L, here)
The foregoing diagram is based upon the material presented by the Atabba
variants, the variants in the reigns of the Agade rulers, and a variation in the
formula for change of dynasty which can be considered absolutely reliable, the
variation between g i S t u k u l b a - s i g and b a l a - b i b a - k i i r .
This obviously does not exhaust the information which can be gained from
the variants in the formulas; for, although we must place the variants "of low
value" apart as unsuitable, we still have left the b a - g u 1 variant and
also the variant m u - b i y m u : m u - b i y : m u y i n the formula for
dynasty total, which for all we know may be of high value and which can
therefore be used if we exercise due caution.
The variant b a - g u 1 in the formula for change of dynasty occurs in Su2
and seems for reasons of space the only possible restoration of the broken for-
mula in P 2 : b a - [ ].93 Since the variant occurs in no other texts and is not
of the type likely to originate independently, we may conclude that Su2 and
P2 had a common ancestor in which the change to b a - g u 1 was made and
from which they inherited it.
Unfortunately this conclusion cannot be drawn with full certainty. Among
the variants in the formulas is one, the use of the singular verb form i - as
in the formula for dynasty total, which is peculiar to the sources from Elam
(Sui, Su2, and SU3+4) and suggests that Su2 should be grouped with Sui and
93
In the best preserved passage, iii 17, there is room for only one sign after b a - .
PoebePs reading - m a instead of - b a (PBS IV 1, p. 76) is not consistent with his copy;
see p. 31, n. 59.
oi.uchicago.edu
TEXTUAL PROBLEMS 47
Su3+4 rather than with P2, This variant is listed with those "of low value/'
and we have shown above (pp. 40-41) that it depends upon qualities which
may be presupposed in any Elamite scribe, so that it may well have originated
independently in two texts. Normally, therefore, its evidence would be of no
consequence compared to that of the b a - g u 1 variant, which belongs to
our most reliable type. In the present case, however, there is reason to men-
tion it; for the fact that b a - g u 1 is only partly preserved in P2 and has to
be restored introduces an element of uncertainty which is naturally stressed
when we find that the i - a5 variant, though admittedly of doubtful value,
points away from the combination P2 Su2. Since this grouping P2 Su2
depends wholly upon the restoration of b a - g u 1 in P2, we can accept it only
with reserve and as a tentative solution. Calling the supposed common origi-
nal "X," we may express the element of uncertainty by adding a query after X
and by using a dotted brace:
X?
P2 S^
In our larger diagram X must be placed under the brace below Bi and B2,
for we know that P2 passed through these texts and also that in them the origi-
nal gi§t u k u l b a - s i g had not yet been changed to b a - g u 1.
Another variant of interest i s m u - b i y m u : m u - b i y : m u y i n
the formula for dynasty total. The form m u - b i y occurs in WB, P6, Li,
and S. Comparing our main diagram we see that through these sources it is
represented in both of the principal branches of the tradition and must there-
fore be the form used in the original. The form m u - b i y m u occurs in
P2, Sui, Su2, and Su3+4. Assuming that the manuscripts mentioned inherited
this variant from a single text, we must obviously place that text, which we
may call "Z," higher up in the genealogical table than Su, from which as
we have seen above Sui and Su3+4 derive, and than X, which we have just
postulated as the immediate ancestor of P2 and Su2. The text Su is character-
ized among other things by a lacuna at the end of the Agade dynasty. But
this lacuna cannot yet have existed in Z, for P2, which also derives from Z, ap-
pears to have had the Agade dynasty intact. The lacuna must therefore have
originated between Z and the two texts which have it, Sui and Su3+4. Similarly
X, which is characterized by the change from gi§t u k u l b a - s i g to b a -
g u l , must be placed between Z and P2 Su2 which have this change;
that Z itself must still have had the original *iSt u k u 1 b a - s i g is shown
oi.uchicago.edu
by the fact that Sui and Su3+4, which also derive from it, preserve that form
unchanged. Acknowledging by queries and dotted lines that the value of our
variant is unknown and that we therefore cannot trust it implicitly, we may
draw up the diagram
Z?
X? Su
P2 Su2? S ^ S ^
In the main diagram Z must be placed under the brace below Bi and B2; for,
since a source such as S, which derives through both these texts, has inherited
the original m u - b i y, the change t o m u - b i y mu must have happened
later than Bi and B2.
Lastly there is the form m u y, which is peculiar to the two sources K
and P5. In other respects also K shows close affinity to P5. As will be re-
membered, K omits the formula for change of dynasty but gives the formula
for dynasty total a s x l u g a l - e - n e b a l a A k i m u y i n - a 5 - m e - e s .
This stands completely alone among our formulas but looks more than any-
thing else as if a later scribe had tried to compress into a single formula the
data given by the formulas for change of dynasty and for dynasty total as
these appear in P 5 : Aki b a l a - b i b a - a n - k i i r n a m - l u g a l - b i
B k i -§6 b a - t i i m and x l u g a l - e - n e mu y i n - a k - e § . The words
b a l a Aki were grafted from the formula for change of dynasty, and x
lugal-e-ne mu y i n - a 5 - m e - e § is the formula for dynasty total
of P5 except that i n - a k - e s has been changed to i n - a5 - m e - e s , which
was considered a better form in late times.94 When we consider the agree-
ment on the variant m u y in conjunction with the fact that on the whole
the singular formula of K stands nearer P5 than any other text, we may be
justified in placing K, at least provisionally, as a late offshoot of P 5 :
Y
S
ft
f
A?
We have now considered all the variants from the formulas except those
listed as "of low value," variants too unsafe to serve as basis for any conclu-
94
See Poebel, GSG § 456. That the formulas of K are related to those of older lists was
noted by Poebel in OLZ XV (1912) 293.
oi.uchicago.edu
TEXTUAL PROBLEMS 49
sions. We might therefore stop here. As it happens, however, there is among

the variants "of low value" a single one which deserves closer attention be-
cause its evidence can be supported in a rather curious manner.
If we would group our sources strictly according to the formulas which they
use we would find only three cases in which two sources have an almost identi-
cal set of formulas, namely WB J, Li P 4 , and Sui Su3+4. The group
which interests us here is WB J, which has the formulas A ki gi§ t u k u 1
b a - s i g (WB: b a - a n - s i g ) . . . . , x l u g a l m u - b i y i b - a 5 , and
N. m u x i - a 6 . Although WB and J are alike in all three formulas, the cru-
cial point upon which the grouping rests is in reality the variant m u x in the
formula for introducing single rulers, for only by using m u x instead of x
m u do WB and J stand apart from such sources as Li and P 4 as a separate
group. As this variant belongs to a type which gives little safety (cf. p. 42),
we cannot accept the grouping from its evidence alone. Comparing the two
texts WB and J we notice, however, that not only are the formulas the same
in both but also the arrangement of the text corresponds. The beginning and
end of the tablet upon which J is written are preserved. The text begins
[ k i - s u - l u - u b 4 - g a r ] g u - t i - u m ki lgi§1t u k u l b a - s i g , and at
this same point begins col. viii in W B : k i - s u - l u - u b 4 - g a r g [ u -
t i (?) - u m ki <gi5t u k u l b a - a n - s i g ) ] . The last two lines of J, n a m -
l u g a l - b i [ i ] - s i - i n k i - s e b a - t i i m , appear as line 22 in WB, al-
most exactly at the middle of the column, which has 45 lines.95 We must
therefore assume that J was written on tablets each of which took exactly
half a column in WB and every second one of which began at the same place
in the text as WB's columns. Such close correspondence in arrangement is
not found between any other two texts among our sources96 and can hardly
be accidental. Apparently the scribes who wrote J and WB copied from the
same text and took care to base their arrangement of the text in tablets or
columns upon that used in this original, so that they did not run the risk of
coming out with a final tablet or column which had only a few lines of writing.
Considering the fact that WB and J use the same set of formulas in con-
96
Mrs. Brookens calls my attention to the fact that the copyist who wrote WB omit-
ted the words 8i5t u k u l b a - a n - s i g after viii 1 (cf. p. 30, n. 52). Had the scribe not
jumped a line here, this column of WB would have had 46 lines and the phrase with which
J ends would have appeared as line 23, i.e., exactly at the middle of the column.
96
Su3 and Sm are no exception. As we have shown above (p. 10, n. 24), the correspond-
ence in arrangement indicates in that case that the two fragments are parts of the same
document.
oi.uchicago.edu
50 T H E SUMERIAN KING LIST
junction with the striking similarity in the way they have distributed the
text, we are justified in assuming that they derive from a common ancestor.
We may call this ancestor "A" and get the diagram
We may now proceed to add the various new items of information to our
main diagram. With the exception of the last, however, the derivation of J
and WB from a common ancestor A, these new groupings cannot be considered
final, because our evidence has not permitted us to exclude a small element of
uncertainty. Since it is of importance to be able to ascertain at a glance which
parts of the diagram are absolutely certain and which might still give reason
for doubt, we have indicated the latter by means of queries and dotted lines.
If the reader imagines the dotted lines and the queried texts blotted out, the
remaining diagram will represent what we know with certainty.
Original
B, B2
I or B or | \
WB J P3 f 5 S P2 5u2? Su, Su 3+4 L, (or, less likely, L, here)

J
K?
SUPPLEMENTARY EVIDENCE FROM VARIANTS OF MORE LIMITED SCOPE
The establishing of the main lines of the genealogical tabic should make it
possible to utilize for confirmation and new details variants of more limited
scope than those hitherto considered. We possess several such variants, for
which some four orfivemanuscripts can be compared, and we must accordingly
examine them to see whether they contain evidence of value.
THE FOURTH DYNASTY OF URUK
We have called attention earlier (pp. 25-26) to the fact that Sux and Su3+4
both give incomplete accounts of the Agade dynasty, suggesting that they
oi.uchicago.edu
TEXTUAL PROBLEMS 51
derive from a common ancestor in which the last part of the Agade dynasty-
had been damaged. We mentioned also that the conclusion that Sui and Su3+4
derive from a common damaged ancestor could be made with absolute cer-
tainty because the Agade dynasty is not the only passage in which they give
evidence of common lacunas. Such a case is the 4th dynasty of Uruk. The ac-
count of this dynasty is wholly preserved in WB and S. It contained five
rulers: Ur-nigin(ak), Ur-gigir(ak) the son of Ur-nigin(ak), Kudda, Puzur-ili,
and Ur-Utu(k). WB gives the total of the reigns as 30 years, S as 26, corre-
sponding to the fact that according to WB Ur-nigin(ak) reigned 7 years, ac-
cording to S only 3. P 4 , in which only the total is preserved, also gives 26
years; so we may assume that it had the same account of the dynasty as S.
Quite different is the account which Sui and Su3+4 give. Both these texts
begin like the others with Ur-nigin(ak) and Ur-gigir(ak), but then they stop.
Su3+4 passes directly to Ur-Utu(k) and the dynasty total, inserting a line to
say that Ur-Utu(k) was son of Ur-gigir(ak). Sui lists an otherwise unknown
king l u g a l - m e - l d m , who is similarly stated to be son of Ur-gigir(ak).
The reigns given by Sux and Su3+4 differ considerably from those of the
other texts. Su3+4 gives Ur-nigin(ak) 15 years, Ur-gigir(ak) 7, and Ur-XJtu(k)
25. The figures 15 and 7 reappear in Sui, but as the reigns of Ur-gigir(ak)
and the new king l u g a l - m e - l a m , while Ur-nigin(ak) according to
that text reigned 30 years.
To follow in detail the genesis of these complicated variants is hardly pos-
sible for the present, but so much seems clear: The original from which Su3-f4
derives had a lacuna in the middle of the dynasty wiping out Kudda and Pu-
zur-ili, so that the copyist had to pass from Ur-gigir(ak) directly to Ur-Utu(k);
and this same lacuna existed in the original of Sui, but there it had widened
still more, so that the name of Ur-Utu(k) also had become illegible. The curi-
ous restoration l u g a l - m e - l a m is probably taken from a still more
damaged parallel text of Y type. In the damaged total 5 l u g a l - e - n e
($\ 0&£^-W- P&4) the scribe has thought to recognize a proper name,
l u g a l - m e ( ! ) - l a m ( ^ ^ T- #£=T).97
If it was already highly improbable that two copies of the King List both
damaged at exactly the same spot, namely in the Agade dynasty, existed in
Elam, it becomes definitely impossible to imagine them damaged twice by
accident and both times damaged at exactly corresponding passages. Our vari-
ant is therefore decisive evidence that Sux and SU 3 -H derive from the same
ancestor.
97
On the use of more than one text in copying see pp. 53-55.
oi.uchicago.edu
52 T H E SUMERIAN KING L I S T
THE INTEKREGNUM IN THE AOADE PERIOD

The section dealing with the interregnum after Shar-kali-sharrI is preserved
more or less completely in five manuscripts: WB, P 3 , Li, S, and Su3-f4:
WB s
[a - b a - a m lu]gal a - b a - a m nu l u g a l a - b a - a [m 1 u g a 1 a - b a - a m
n u lugal]
m
[ -gi4]-gi4 lugal l - g i 4 - g i 4 [1] u g [a 1] ffii-mi
[l]ugal
m
[ ] lugal n a - n u - u m l u g a l mi - l u -
lu l u g a l
[ 1 lugal 4-bi 3mu in-a5
[ 1 lugal
[ ] lugal
[ ] ib-a5
Sll3 + « Pa Li
a-ba lugal [m]a-nu-um sdr ma-nu-um s[dr]
a-ba lugal-am ma-nu-um la sd[r] ma-nu-um l[a sdr]
i r - g i4 - g i4 l u g a l i-gii-gi* sdr ir-ki-k[i sdr]
na-nu-um lugal i-mi sdr na-nu-[um sdr]
i-mi lugal na-ni sdr i-[mi sdr]
i-lu-lu lugal e-lu-lu sdr [ ]
[ ] 1uga1 4-bi3mu'ib-a51 [ ]
[ 1
Disregarding those variants which appear in only a single manuscript and
which are therefore of no help in grouping the manuscripts, we notice (1) that
the Sumerian phrase a - b a - k m l u g a l a - b a - ^ m n u l u g a l of WB,
S, and—mutatis mutandis—SU3+4 varies wdth the Akkadian ma-nu-um sdr
ma-nu-um la sdr of P 3 and Li; (2) that the order of the rulers, which is Irgigi,
Nanum, Imi, Ilulu in Su3+4 and Li, is Igigi, Imi, Nani, Elulu in P 3 and S;
(3) that the name of the first ruler is written ir-ki-ki or ir ^ g i4 - g i4 in Lj
and Su3+4 but i - or i - g i4 - g i 4 in P 3 and S; and (4) that the total in WB
and Su3+4 contains the word l u g a l , whereas S and P 3 do not have that
word. If we arrange the variants successively it will be seen that their dis-
tribution among the manuscripts is peculiar:
a b a m l u g a l a b a m nu l u g a l WB Su3+4 S manum sar manum la sar P 3 Li
Irgigi, Nanum, Imi, Ilulu Su3+4 Li Igigi, Imi, Nani, Elulu P3 S
ir-gi4/ki-gi4/ki Su3+4 Li i / i - g i4 - g i4 P3 S
[ 4 - b i (?)] l u g a l [ m u 3 / 3 m u ]
ib/[l]-a5 WB Su3+4 4-bi3mu ib/in-a5P3S
We notice here that Li and S suddenly change places when we pass from
the first variant to the others. In the a b a m : manum variant S follows
oi.uchicago.edu
TEXTUAL PROBLEMS 53
S113-M and differs from Li, which follows P 3 ; in the following variants, how-
ever, S follows P3, and it is Li which follows Su3+4.
Such a distribution of the variants cannot possibly have developed through
normal transmission. To explain it we must assume either that one of our
variants arose independently in two or more manuscripts or that the scribe
of one of them used two different texts as originals for his copy. An example
may elucidate the latter possibility. If the line a b a m l u g a l a b a m
n u l u g a l was destroyed in the original of Li and the copyist restored it
from another text in his possession, a text of the same type as P3, we can
understand how Li might follow P 3 as far as this variant is concerned but dif-
fer from it already in the next line. It is, however, equally possible that it was
the section containing the three following variants which was destroyed in
the original of Li and that the scribe restored it from a text like Su3+4. There
is thus a variety of ways in which this distribution of variants may have
come about: independent origin of two or more variants, restoration of one
or another section from texts of different type, etc. Which of the many possi-
bilities is the correct one cannot be decided; and, as the evidence of this sec-
tion must remain ambiguous, we cannot use it with safety for the recon-
struction of the genealogical table.
THE KU(G)-BABA VARIANT
Queen Ku(g)-Baba of Kish constitutes a separate dynasty in manuscripts

WB, Li, P2, and Sui, whereas she is placed at the head of the 4th dynasty of
Kish in S and Su3+4. Several indications show that the arrangement with
Ku(g)-Baba as a separate dynasty is that of the original. It is—to mention
only one such indication—vouched for in both branches of our tradition.98 The
variant which places Kii(g)-Baba at the head of the 4th dynasty of Kish must
therefore be secondary. Since the 4th dynasty of Kish begins with Puzur-Sin,
a king stated to be the son of Ku(g)-Baba, our variant was no doubt introduced
by a scribe who considered it impossible to separate the reign of Ku(g)-Baba
from that of her son by the reign of a complete dynasty of another city. The
correction occurs in two of the manuscripts, S and Su3+4, and it is naturally
tempting to assume that they have inherited it from a common ancestor. Such
an assumption is, however, impossible. We have seen above that a number of
identical omissions in Sui and Su3+4 put it beyond doubt that these sources
derive from a common ancestor, which we have called "Su." Since Sui pre-
serves the original arrangement with Ku(g)-Baba as a separate dynasty, while
Su3+4 shows the correction, we would obviously have to assume that the cor-
98
See also the discussion of this variant on pp. 159 f.
oi.uchicago.edu
rection was made somewhere between Su and Su3+4. We know, however, that
S, which also has the correction, is not derived through Su; therefore it cannot
derive from a descendant of Su. Hence S and Su3+4 cannot have inherited
their correction from a common ancestor, and the distribution of the variant
must be due to "double origin"; that is, it was made twice independently, or
the scribe who wrote one of the two manuscripts must have used a text of
different type along with his original. Both of these explanations are possible.
The separation of Ku(g)-Baba from her son is in itself a very curious feature
which may well have been noticed and corrected independently by two dif-
ferent scribes. But there are special reasons which make the other possibility
even more likely. It will be remembered from the section on formulas that in a
single passage in Su3+4 (iii 22-30) the formula for introducing single rulers
differs from that used elsewhere in the text. The formula in this passage is
N.mu x i - a5; elsewhere it is N. x m u i - a s . The variant mu x appears
in the other manuscripts only in the A group—WB and J—and in the Y
group, in P5. Its sudden occurrence in a short section of Su3+4 must there-
fore mean that this section was destroyed in the scribe's original, so that he
had to restore from another text—a text of A or of Y type. Now if the scribe
of Su3+4 can have consulted a text of Y type along with the one he used as
basis for his copy, it seems almost certain that he got his Ku(g)-Baba correc-
tion from there; for S—the only Y text which preserves this section—has the
variant which places Ku(g)-Baba just before her son.
This Y influence upon Su3+4 we may express in our diagram by means of
an arrow. As it is not the only possibility, however, we must draw the arrow
in dotted line.
INFLUENCE FROM THE Y GROUP ON SUi, SU2, AND L*

The case just mentioned where we found evidence of influence from texts
of one group upon those of another is not the only one of its kind. We have
already earlier (p. 51) suggested that the inexplicable l u g a l - m e - l a m
which Sui gives as the name of a ruler in the 4th dynasty of Uruk may be a
restoration from a text of Y type in which a broken formula for dynasty
total, [x] l u g a l - r e ] - n e , was read as l u g a l - m e - l a m and
considered a proper name. Since both Sui and Su3+4 thus show influence
from a Y text, it seems reasonable to assume that our third Susa version also,
Su2, where we can observe foreign influence in the lines (iii 4-5) giving m u
x instead of the normal x m u , has been restored with the help of this Y text.
As a last case we may mention Li, which in viii 4-7 and 20-21 suddenly
changes from its usual formula for introducing single rulers, N. x m u i - a 5 ,
oi.uchicago.edu
TEXTUAL PROBLEMS 55
to N. x m u i n - as. We must assume here again that the passages in ques-

tion were damaged in the original which the copyist used as his basis and
that he restored them from a second text to which he had access, a text which—
as is shown by the variant i n - a&—must have been of Y type."
With the variants of limited scope investigated above we have considered

all the variants in the postdiluvian sections of the King List which can con-
tribute toward the reconstruction of the genealogical table, namely the vari-
ants which can be traced over at least four manuscripts. Since the discussion
of the problems connected with the antediluvian part of the list needs a sec-
tion by itself, we may add the results thus far achieved to our diagram:
T H E ANTEDILUVIAN SECTION
Four of our manuscripts—P2, P3, P4, and Sui—contain evidence showing

that they began their accounts with the 1st dynasty of Kish, that is, with
the first of the postdiluvian dynasties.100 Seven others—P6, Li, L2, S, Su2,
99
There is some reason to separate the ease of Y influence which we find in Sui, S112,
and S1134.4 from that of Li. For since Sui, Su2, and Su3+4 have the same provenience, Susa,
and we can hardly assume that imported Babylonian copies of the Sumerian King List
were abundant there (cf. p. 25), it seems likely that it is the same Y text which has in-
fluenced all our Susa versions. This Y text was no doubt so damaged that it could not be
used as a basis for later copies, but it was useful as a help in filling in gaps where the better
preserved texts were damaged. We have therefore used two arrows in the diagram, one for
the Y influence in Li and one for the Y influence in the Susa texts.
100
The evidence is as follows:
P2. The summary in col. xii gives only 11 cities of royalty, which agrees with a list be-
ginning with Kish I. Furthermore, the section missing at the top of the tablet can be estimat-
ed fairly accurately. Col. i ends with Tizkar, and col. ii begins—after the lacuna—with
the formula for change of dynasty referring to the change from Kish I to Uruk I. This
lacuna corresponds to 15 lines in WB. In col. i of P 2 the line following Kalibum's regnal
years stands opposite the first line preserved in col. ii. Counting 15 lines backward from
oi.uchicago.edu
SU3+4, and J—are too fragmentary to allow any conclusions as to where they
began. Three—WB, K, and P5—seem to have had an initial section dealing
with antediluvian rulers. This antediluvian section is preserved completely
in WB. In K traces of a whole column of rulers before the 1st dynasty of Kish
show that this text also must have listed antediluvian rulers.101 P 5 , a fragment
which constitutes the lower left-hand corner of a large tablet, begins with the
1st dynasty of Kish, but this dynasty could not appear so far down on the
tablet if it was not originally preceded by an antediluvian section.102 We thus
have
with antediluvian section WB, P§, K;
without antediluvian section P2, P3, P4, Sui;
indefinite P6, Li, L2, S, Su2, Su3+4, J-
There are obviously two possible ways in which this variant and its distri-
bution may have come about: (1) If we assume that the antediluvian section
is an original constituent of the King List, its absence in P 2 , P3, P4, and Sui
must be due to omissions in these versions or—more likely—to omission in a
common ancestor. (2) If we assume that the antediluvian section is second-
this place in WB we arrive a t i 4 1 : n a m - l u g a l a n - t a e n - d e - a - b a , which fol-

lows the record of the Deluge. Thus P 2 began with Kish I and had no antediluvian rulers.
P4. This text likewise counts 11 cities of royalty and accordingly cannot have had any
antediluvian rulers.
P3. Col. i ends with a lacuna after Etana, and col. ii begins after a lacuna with the note
on En-me(n)-barage-si. To the missing passages correspond 18 lines in WB. The lacuna
at the top of col. i, which goes to [Kalibu]m, one line lower than the first line preserved in
col. ii, can therefore at most have amounted to 19 WB lines; probably it was somewhat
less. Counting 19 lines back from Kalibum in WB takes us to the end of the antediluvian
kings (i 35), the dynasty total for Shuruppak. In P 3 also there is thus no room for a list of
antediluvian rulers.
Sui. Col. i ends with Samug before a lacuna. Col. ii begins with the years of En-me(r)-
kar. The missing portion corresponds to 29 lines in WB. Opposite the first line in col. ii is
Zuqaqlp in col. i; but, since the position of Atab and Atabba is different in WB as com-
pared with Sui, it is better to count from Arwi'um, who stands opposite the years of Lugal-
banda. The lacuna is then 31 WB lines. Counting back 31 lines from Arwi'um in WB takes
us only to the end of the dynasty of Sippar (WB i 30). Thus even if col. i in Sui had ended
at Samug, so that all that is missing had to be restored at the top of the tablet, we would
not have room enough for the antediluvian section. This text also must accordingly have
started with Kish I.
101
This conclusion was drawn already by King, Legends of Babylon and Egypt, p. 31, n. 4
102
See the drawing of the fragment in PBS V, PL IV, and photograph in BE XX 1,
PL XV. Considering that the tablet originally had two columns on each side, making it
twice as broad as the fragment which is left, we must, to get reasonable dimensions, suppose
that it was at least twice as high as the fragment.
oi.uchicago.edu
TEXTUAL PROBLEMS 57
ary,103 a later addition to the King List, its occurrence in WB, P5, and K must
be due to independent insertion in WB and in P5, from which K would then
inherit it; for, since any common ancestor of WB and P5 must also have been
an ancestor of P2, P3, and Sux, WTB and P 5 cannot have inherited the section
from such a text.
Neither of the two main possibilities here mentioned seems a priori more
probable than the other. As for the first we may, it is true, doubt that any
scribe would voluntarily omit a section which ' 'handed down the tradition
relating to (things) prediluvian."104 Such precious information is not willingly
left out.105 But the omission need not have been voluntary. Clay is brittle
writing material, and a copyist may very well have found this section so seri-
ously damaged in his original that he had no choice but to begin his copy with
the postdiluvian rulers. As for the second possibility, that the antediluvian
section is secondary, it is obvious that the exceptional interest of the section
would strongly favor its insertion. Any copyist who did not find the antedilu-
vian rulers mentioned in his original of the King List but knew of them from
elsewhere must have felt a natural impulse to round off his account of bygone
dynasties by adding such important information to his copy.
Since both possibilities thus seem equally probable, we must examine the
antediluvian section itself more closely to see if it contains any clues which
will decide the question. The first thing to be noted is the curious and signif-
icant independence which characterizes this section in contrast to all the rest
of the list. The other sections are true parts of the King List, do not appear
outside it, and have existence as parts of that composition only. Not so with
the tradition concerning the antediluvian rulers. This part has an individuality
of its own; it is not tied to the King List but can be found in the literature of
103
This seems to be the view of King, who was aware of the existence of this section in K
and of its absence in some of the Nippur fragments. He does not discuss the question in
detail, but his opinion may be inferred from his suggestion that "the exclusion of the Antedi-
luvian period from the list may perhaps be explained on the assumption that its compiler
confined his record to 'kingdoms/ and that the mythical rulers who preceded them did not
form a 'kingdom' within his definition of the term" (be. tit.). Note, however, that the form
which the antediluvian section takes in WB speaks of the "kingdoms" of the various ante-
diluvian cities.
104
R. Campbell Thompson, The Epic of Gilgamish. A New Translation (London,
1928) p. 9 i 6.
106
The keen interest which the ancients took in antediluvian matters may be seen from
the line of the Gilgames epic just quoted. Cf. also Ashurbanipal's occupation with reading
"stones from the time before the Deluge" (M. Streck, Assurbanipal II ["Vorderasiatische
Bibliothek" VII 2 (Leipzig, 1916)] 256:18; D. D. Luckenbill, Ancient Records of Assyria
and Babybnia II [Chicago, 1927] § 986).
oi.uchicago.edu
the period as a separate entity or even as part of a composition other than the
King List. As a separate entity the antediluvian tradition appears on Tablet
W-B 62, published by Langdon.106 This tablet gives the list of antediluvian
kings very much as does WB, but it is there unconnected with lists of other
rulers and has individual features which show that it is not merely a section
of the King List gone astray. 107 Chief among these is a complete absence of
the formulas so characteristic of the King List proper.
Of still greater interest is it to find our tradition as part of another literary
composition. As No. 1 of his Historical and Grammatical Textsm Poebel has
published a Sumerian epic which deals with the beginning of the world, and
there can be no doubt that a very close connection exists between this epic
and the antediluvian tradition as we have it in WB. The epic109 describes in
the first column the creation of men and animals; then, after a lacuna, col. ii
proceeds to relate the descent of kingship from heaven:
[me]n n a m - l u g a l - l a a n - t a en - d [& (?)] - a - b a
lgil
*§ i b i r gi5rg u - z a1 n a m - l u g a l - l a a n - t a e n - a - b a
"when the crown of kingship was lowered from heaven,110
when the scepter and the throne of kingship were lowered from heaven,"
and the founding of five cities, which are mentioned in the order Eridu(g),
Bad-tibira(k), Larak, Sippar, Shuruppak. After a new lacuna col. iii begins
the story of the Deluge, which, interrupted by lacunas, continues through
cols, iii-iv. The end of the Deluge is related in col. v in the following words:
a-ma-ru u-gu kab-dug4-ga b a - a n - d a - a b - u r - e
u4 7 4 m gi 6 7-km
a-ma-ru kalam-ma ba-ur-ra-ta
™OECT II, PL VI, and JRAS, 1923, p. 256. See also Zimmern in ZDMG n.F. I l l
(1924) 20, who dates the tablet correctly to the end of the 3d millennium B.C. and shows
convincingly that it is written in Sumerian.
107 The reigns of the kings listed differ considerably in length from those given by the
antediluvian section of WB; WB gives only Ubar-Tutu(k) as ruler of Shuruppak, whereas
W-B 62 has Shuruppak(-gi) and Zi-u-sud-ra; W-B 62 has an antediluvian dynasty of Larsa
which WB does not give; Eridu(g) of WB appears as Kuâ(ra) in W-B 62; etc. See also
S. Smith, EHA, pp. 20f., who similarly stresses the independence of W-B 62: "The scribe of
the second list did not therefore take the first list and wilfully alter it, simply to glorify his
own city"; but cf. our reservations to details in Smith's view stated below, p. 71, n. 17.
108 pBS V.
109
Cf. PoebePs translation and commentary in PBS IV 1, pp. 9-70, and King's discus-
sion of the text in Legends of Babylon and Egypt, pp. 41-101.
110
There is no necessity to render a n - t a as "from Ami," as Deimel proposes in Orien-
talia No. 17 (1925) p. 35, instead of "from heaven."
oi.uchicago.edu
TEXTUAL PROBLEMS 59
"The Deluge sweeps at the same time over the . . . .s. m

After the Deluge had swept over the land
for seven days and seven nights/'
and the tablet ends with the tale of Zi-u-sud-ra's sacrifices and how he was
made immortal.
The close relation of this account to that of WB is obvious. WB begins
with the phrase
[n a m] - 1 u g a 1 a n - t a eu-de-a-ba
"when the kingship was lowered from heaven,"
goes on to describe the antediluvian dynasties in the order Eridu(g), Bad-
tibira(k), Larak, Sippar, Shuruppak, and ends:
a - m a - r u b a - u r RA TA112
egir a-ma-ru ba-ur-ra-ta
"The Deluge swept.
After the Deluge had swept."
As will be seen, the ideas in these two texts—(1) the kingship being lowered
from heaven (WB: [ n a m ] - l u g a l a n - t a e n - d e - a - b a ; PBS V, No.
1: [ ] n a m - l u g a l - l a a n - t a e n - d [ 6 ( ? ) ] - a - b a ) , (2) the identi-
cal order of the primeval cities, and (3) the raging of the Deluge (WB:
a - m a - r u b a - i r ; PBS V, No. 1: a - m a - r u . . . . b a - a n - d a -
a b - u r - e ) — a r e very much the same, follow in the same succession, and
are expressed in almost identical language. 113
That the tradition concerning the antediluvian rulers can thus be shown to
exist as a separate entity outside the King List, appearing either by itself or
111
The word k a b - d u g« - g a is still not certainly explained. Poebel (PBS IV 1,
p. 41) thinks it is an epithet of the Deluge-demon and in ii 21 translates it as denoting the
gods to whom the antediluvian cities were assigned. Poebel is followed by King, who (Leg-
ends of Babylon and Egypt, pp. 58 f.) tentatively translates "divine rulers." Deirnel (he.
cit.) translates "Kultort"; and this rendering, which suits the context perfectly, is probably
correct, even though Deimel's reasons for it, "k a b - d ii (g) = k a -f- s u (arch. (k a b')
— ikribu, kar&bu" carry little conviction.
112
The signs RA TA here are obviously dittography from b a - u r - r a - t a in the next line.
113
The close correspondence in wording indicates that we are dealing with literary de-
pendency and not with different records of a common oral tradition. It is true that the epic
in its present dilapidated condition does not contain a passage giving the exact information
which we should expect on that assumption. It has names and order of the cities but lacks
names, order, and reigns of the antediluvian rulers. That a passage containing this informa-
tion once stood in the large lacuna covering the upper half or more of col. iii is, however,
highly probable. Col. ii ends with the allotment of the antediluvian cities. Where, after the
lacuna, the preserved part of col. iii takes up the narrative, Enlil has decided to send a
Deluge, and the gods are troubled. Of the passage which once bridged this gap we must
therefore expect that it continued the subject begun in col. ii, telling who ruled these ante-
diluvian cities, and then gave the reasons for EnhTs decision to wipe out mankind with the
oi.uchicago.edu
as part of an epic, naturally raises the question of where its original setting
should be sought. The answer is not difficult to find. In the epic the god Enki
plays a most conspicuous role. He is mentioned with Enlil and Ninhursag
as creator of mankind, and he is indisputably the hero of the following Deluge
story, for it is through his activity alone that mankind is saved. We can
therefore safely conclude that the epic hails from Eridu(g), the chief city of
Enki, and draws on its mythical lore. Toward Eridu(g) points also the tradi-
tion of the antediluvian rulers as we have it both in WB and in the epic, for
in both versions Eridu(g) is given the place of honor at the top of the list as
the foremost antediluvian royal city.114 The tradition therefore fits into the
epic hand-in-glove, and we cannot doubt that its original setting should be
sought in the lore of Eridu(g) and not in the King List.
We have thus found that the original setting of the antediluvian tradition
Deluge. And such a passage actually exists. Dr. Geers has discovered in the British Museum
a fragment of Assyrian date (K 11624) which, like our conjectural passage, begins with
an account of the antediluvian cities, rulers, and reigns and then goes on with lines of an
epic nature apparently dealing with the reasons that made Enlil send the Deluge. (The
lines mention Enlil and huburu, "noise," recalling the similar passage in the Atar-hasis
story, ingeniously restored by Sidney Smith [RA XXII (1925) 67 f.], which states that it
was the noise [huburu] of the humans that kept Enlil awake and made him send the Deluge
to restore peace and quiet.) These lines, with which the fragment breaks off, lead up to ex-
actly the situation found where col. iii of PBS V, No. 1, is preserved: Enlil has decided to
send the Deluge, but his decision is as yet known only to the gods. In view of the exceptional
ease with which the fragment thus fits into the gap in the narrative of PBS V, No. 1, it
seems indeed more than likely that it represents a late version with interlinear translation
of this very epic. Since the gap in epic PBS V, No. 1, may thus plausibly be restored from
the fragment and since the fragment contains the very information wanted, there is no
reason to doubt the existence of literary dependency between the epic and the King List
(WB). What form this dependency may have taken is discussed below (p. 64, n. 119).
Dr. Geers has kindly permitted me to publish a drawing from his copy of K 11624 (see
plate at end). We give here a transliteration and translation:
r
s i p p a rki1 [ b a l a - b i b a - a n - k t i r ] Sippar, its period of reign was altered.
Ia7-ra-ak-aki e[n-sipa-zi-an-na] In Larak En-sipa(d)-zi(d)-Anna(k),
1 l u g a l - e b a l a 3[6,000 m u] 1 king, period of reign 36,000 years.
1 a? - r a - a k - aki b a l a - b [ i b a - a n - k t i r ] Larak, its period of reign was altered.
LAM+KUR-RUki u b u r -1 [u - 1 u] In Shuruppak Ubar-Tutu(k),
zi-114-sud-ra d u m u u [ b u r - t u - t u ] Zi-u-sud-ra, son of Ubar-Tutu(k),
2 - a m l u g a l - e - n e b a l a [. . mu] 2 kings, period of reign . . years.
5uru-dil-dili91ugal-re1-[ne....] Five separate cities, 9 kings
d
en-lil-le nam-[ ] Enlil
d
en-lil ni-[. . . .]
KA-fLi-KA+Li i b - [ ] the (Akkadian: "their"?) noise
&w4wr4ti(?)[-»w(?)l
We have included the evidence of the text quoted above—designated as "epic fragment
K 11624' '—in the notes to the text of WB, where the discussion of details will be found.
114
That this position was indeed a desirable one may be seen, e.g., from the fact that
later Babylonian tradition, which has been preserved by Berossus, places Babylon there.
oi.uchicago.edu
TEXTUAL PROBLEMS 61
does not seem to be the King List and that this tradition exists as part of an-
other literary composition and as a separate entity in the Sumerian literature
of the Isin-Larsa period. The scribes who wrote our copies of the King List
must therefore have had ample opportunity to know this tradition even if it
did not appear in their originals. All this agrees perfectly with our second
possibility, that the antediluvian section is secondary in the King List. It
proves that all conditions for the making of such an addition were extant at
the Isin-Larsa period. It is not enough, however, to exclude the first possi-
bility, that the antediluvian section is original, for it is clear that traces of an
earlier setting such as we have found would remain if the author himself had
taken this section of his work from a tale from Eridu(g); and the tale in which
the tradition originally belonged could live on along with the King List down
to the Isin-Larsa period.
The peculiarities of the antediluvian section mentioned thus far, its inde-
pendence and its partiality to Eridu(g), are not, however, the only features
which set it apart from other sections of the list. It has a peculiar set of formu-
las. Since we have it fully preserved only in WB, we must study these pecu-
liarities there.
As will be remembered, the formulas characteristic of the main body of
WB, the postdiluvian section, are:
Formula for change of dynasty:
A ki giSt u k u l b a - a n - s i g n a m - l u g a l - b i B k i - § e b a - t u m
"The city A was smitten with weapons; its kingship was carried to the city B . "
Formula for dynasty total:
x l u g a l m u - b i y f b - a 5 (if the dynasty has only one king: i - a 5 )
"x kings reigned its (the city's) y years."
Formula for introducing single rulers:
N. m u x i - a s
I n t h e antediluvian section, however, the formulas a r e :
Formula for change of dynasty:
A k i b a - § u b - b 6 - e n n a m - l u g a l - b i B k i - § & b a - 1 u m115
"I (the author!) drop the city A;116 its kingship was carried to the city B . "
115
The passages where the formula occurs are: i 8-10: e r i d a k i b a - § u b ( - b 6 -
e n ) n a m - l u g a l - b i b a d -1 i b i r aki - § e b a -1 u m ; i 18-19: b a d -1 i b i r a ki
b a - § u b - b £ - e n n a m - l u g a l - b i l a - r a - a k k i (-§e) b a - t l i m ; i 24-25:
l a - r a - a k k i b a - s u b - b e - e n n a m - l u g a l - b i z i m b i r<ki> - § e b a - t t i m ;
i 30-31: z i m b i r k i b a - £ u b - b 6 - e n n a m - l u g a l - b i § u r u p p a k k i ( - § e)
b a- t u m .
116
I.e., "I drop the subject of A (and proceed to deal with another subject)." The author
himself is speaking. The Sumerian § u b , "to fall," transitive "to drop," can be used in
oi.uchicago.edu
Formula for dynasty total:

x l u g a l m u - b i y i b - a 5 (even if the dynasty has only one king!)117
"a; kings reigned its y years."
Formula for introducing single rulers:
N. m u x i-a 5 118
It will be noted that these sets of formulas differ on two points. In the
antediluvian section the formula for change of dynasty begins: "I (the au-
thor) drop the city A"; in the postdiluvian section the formula begins: "The
city A was smitten with weapons." In the antediluvian section the collective
form of the verb, f b - a 5 , is used in every total; in the postdiluvian sec-
tion i b - a5 is used only in totals of dynasties with several rulers, whereas
if the dynasty has but one ruler the singular i - a5 is employed. These dif-
ferences are important. The whole King List bears witness that its author
the meaning "to leave," "not to occupy oneself with a thing any longer," exactly like "to
drop (a subject)" in English. On S u b , "to leave," cf. the equation s u b : ezebu in SL,
No. 68:3; also such passages as SRT, No. 3 iii 12-13: u - m u - n i - k r i u - m u - n i -
n a g - rg a1 - 1 a i n f g - d i r i - g a g a - m u - n a - r a - a b - s u b , "Excess of oil be-
yond what he can eat and drink I shall let be left over ( - r a - ) for him," i.e., "I shall
serve him more oil than he can drink, so that some will be left" (- r a - is the disjunctive
verbal infix; our passage shows that—contrary to GSG § 513—this infix follows the da-
tive infix, just as do other directional infixes; cf. also H. de Genouillac, Textes economiques
d'Oumma de V&poque d'Our [Paris. Musee national du Louvre, "Textes eunciformes" V
(Paris, 1922)] No. 6167 rev. 3 : m u - n e - r a - e , "came forth for them," and No. 6164
rev. 4: m u - n a - r a - n e - e - e s , "came forth for him"); SEM, No. 49 obv. 5:
d
g a § a n a n n a r an m u - u n ' - s u b r ki m u - u n - s u b 1 k u r - r a b [ a - e - a - e u ] ,
"Gashananna(k) left heaven, left earth, and descended to Hades" (cf. now also S. N. Kramer
in RA XXXIV [1937] 98: "she forsook"). Langdon's translation of the formula in OECT
II 8 f. (accepted by Smith, EH A, p. 23, and Zimmern in ZDMG n.F. I l l 22), "(the city A)
was overthrown," is excluded by the form of the verb, b a - s u b b - e n , which must be
first or second person and cannot be third.
117
The passages in which this formula occurs are: i 6-7: 2 l u g a l m u ( - b i ) 64,800
i b - a5 ; i 16-17: 3 1 u g a 1 m u - b i 108,000 i b - a5 ; i 22-23: 1 1 u g a 1 m u - b i 28,800
i b - a5 ; i 28-29: 1 1 u g a 1 m u - b i 21,000 i b - a5 ; i 34-35: 1 1 u g a 1 m u - b i 18,600
i b - a6.
118
The formula occurs in i 5: a - l a l - g a r m u 36,000 1 - a6; i 13-14: e n - m e -
e n - g a l - a n - n a m u 28,800 i-agj i 15: d d u m u - z i s i p a m u 36,000 1 - a 6 .
The related formula for introducing new dynasties, Akl - a N. l u g a l - a m m u x i - as,
is used carelessly: i 3-4: e r i d a k i a - l u - l i m l u g a l (!) m u 28,800 1 - a5 ; i 11-12:
b a d - 1 i b i r aki e n - m e - e n - l u - a n - n a ( I ) m u 43,200 i - a5 ; i 20-21: 1 a - r a -
a k k i e n (!) - s i p a - z i - a n - n a (!) m u 28,800 1 - a5 ; i 26-27: z i m b i r k i e n - m e -
e n - d t i r - a n - n a l u g a l - a m m u 21,000 1 - a5 ; i 32-33: & u r u p p a k ki u b u r -
t u - t u l u g a l - a m mu 18,600 1 - a 5 . See pp. 67 f., where the significance of the variants
in this formula is discussed.
oi.uchicago.edu
TEXTUAL PROBLEMS 63
was a man who was fond of formulas and used them with singular precision
and consistency. It is inconceivable that a man of this type should have made
the purposeless and totally unnecessary change from "I drop the city A" to
"The city A was smitten with weapons" which we find in the formula for
change of dynasty. People with precise minds do not start before they have
the exact formula to be used clearly in mind, and then they stick to it. They
do not arbitrarily reject it when they get halfway, to evolve a different one.
The existence of this change therefore points to a different hand; a later copy-
ist is trying in the antediluvian section to adapt a source with different phrase-
ology ("I drop the city A and proceed to deal with the city B") to that used
in the King List ("The city A was smitten with weapons; its kingship was
carried to the city B").
But the final and irrefutable proof that the antediluvian section is not of a
piece with the postdiluvian is furnished by the other difference in the formulas,
the varying use of the collective. In the postdiluvian parts of the King List
the collective f b - a5 is, as mentioned above, used only in totals of dynasties
numbering several rulers, whereas we find the singular i - a5 in dynasties
with only one king. This section must therefore have been written by a man
who understood the difference between i b - as and i - as and used the two
forms correctly. In the antediluvian section, however, the collective i b - as
is used in every total, even when the dynasty has only one king. This sec-
tion cannot, therefore, have been written by the same man who wrote the
postdiluvian section but must have been composed later when the collective
was no longer in living use and by a man who did not know what it stood for.
It is also clear how this later writer arrived at the peculiar rules which govern
his use of the form. Cases of dynasties with only one ruler are rare in the post-
diluvian section; so a cursory inspection of his original would seem to him to
show that, while i - a5 should be used in the formula for introducing single
rulers, the form i b - as, which he did not understand, was characteristic
for the totals. Faced with the task of adding the material on antediluvian
rulers which he had found in another source and of adapting its phraseology
to that of the King List, he therefore used his self-made rule and wrote
i b - as in all his totals. It was, however, only in the new section which he
added, the antediluvian section, that he had to bother thus with grammatical
details. When he came to the postdiluvian section he could settle down to
merely reproducing his original sign by sign. Thus his self-made rule did not
disturb the old correct use of the two forms which appeared there.
The role of the antediluvian section in the tradition of the King List can
thus no longer be doubtful; it is a later addition. We have seen that it most
oi.uchicago.edu
likely originated in Eridu(g), that it was current in various settings in Su-

merian literature at the time when most of our copies of the King List were
written, and that it must have seemed valuable and desirable information to
the ancients. We have found also that the form in which it appears in WB
shows unmistakable signs that it was written later by a person different from
the one who composed the postdiluvian section of the list. This means that
the original form of WB, without the antediluvian section, agrees with the
manuscripts P2, P3, and Sui which belong to the other main branch of the
tradition, the B branch. Since the form without the antediluvian section is
thus represented in both main branches, it must have been that of the origi-
nal.119
119
It is ot interest to determine as accurately as possible the extent of the antediluvian
addition and to distinguish between those parts of it which come from the source used by
the scribe and those which he himself added to fit it into its new setting.
To material which the scribe must have got from his source obviously belong the actual
names and reigns given. As for the formulas, it is clear that the first part of the formula
leading from one dynasty to another must have been taken over from the source, for there
must have been a reason why the scribe did not simply use the formula as given in the
King List proper, A k i * i 5 tukul b a - a n - s i g n a m - l u g a l - b i B k i -§e b a - t i i m ,
but substituted Aki b a - § u b - b 6 - e n for Aki *H u k u l b a - a n - s i g . This reason
can only be that he found A k i b a - § u b - b 6 - e n in his source for the antediluvian sec-
tion. Since the formula A k i b a - s u b - b 6 - e n , "I drop (the subject) A," calls for the
mentioning of a new subject, B, immediately afterward, we can conclude that the source
for the antediluvian section, like the King List proper, introduced new dynasties by men-
tioning the name of their city. But it is unlikely that it also had l u g a l - a m after the
name of the first ruler, for in the antediluvian section our scribe has the greatest difficulty
in remembering to use this l u g a l - a m and forgets it completely both in i 11-12 and in
i 20-21. Whether the source for the antediluvian section used a verb "he reigned" after
the various reigris and whether it totaled up the reigns and dynasties is impossible to decide.
As we have mentioned above (p. 59, n. 113), there exists between the antediluvian sec-
tion of WB and the epic PBS V, No. 1, a similarity of phraseology so striking that we must
assume literary interdependence between the two texts. The fact that the antediluvian tra-
dition seems at home in the epic (pp. 58-60) but appears in only a few copies of the King List
as a late secondary addition makes it highly unlikely that the epic obtained its material
from the King List. On the other hand we cannot assume that the scribe who inserted this
section in WB took it directly from the epic. We have seen that his source used a formula
"I drop the city A"; and such a formula is unthinkable in an epic, where the author cannot
suddenly commence speaking in the first person. The source was therefore more likely a
separate, self-contained version of the tradition, a list of the type represented by W-B 62,
where such a formula would fit in well. This leaves us three possible ways to explain the
relationship: (1) If we may assume that the names and reigns of the antediluvian kings were
given in the lacuna covering the upper half of col. iii in PBS V, No. 1 (see p. 59, n. 113),
the separate list used as source by WB may have been excerpted from this epic. (2) The
epic may have used the same separate list which WB used. (3) The epic and the separate
list used by WB may both derive from a third, as yet unknown, literary composition and
may have inherited their common material and phraseology from there. Now it is clear that,
whichever of these possibilities is correct, the existence of a close literary relationship be-
oi.uchicago.edu
TEXTUAL PROBLEMS 65
It remains to examine the question when in the tradition of our manuscripts

the insertion of the antediluvian section may have been made. In the texts of
tween the epic and the source used by WB forms a valuable clue for separating those phrases
in the antediluvian section of WB which the scribe has taken over from his source from those
he himself added to fit the section in. If the source of WB and the epic are closely related,
it is obvious that phrases which appear in both WB and the epic are likely not to be inven-
tions by the scribe but to come from the source. Comparing, then, the phrases which intro-
duce and close the antediluvian section in WB with those of the epic PBS V, No. 1,
[me]n n a m - l u g a l - l a a n - t a en -
[nam]-lugal an-ta e u - d e - a - b a d[e(?)]-a-ba
l«!l*8 i b i r * i 5 r gu-za 1 n a m - l u g a l - l a
an-ta en-a-ba
eridaki n a m - l u g a l - l a
(antediluvian cities and kings) (antediluvian cities) x
a-m a-r u b a - u r a-ma-ru u-gu kab-dug4-ga ba-

an-d a-ab-ur-e
u* 7 - a m gi 6 7 - a m
egir a-ma-ru ba-ur-ra-ta a-ma-ru kalam-ma ba-ur-ra-ta
nam-lugal an-ta en-de-a-ba
kisiki n a m - l u g a l - l a ,
we notice that the phrases n a m - l u g a l a n - t a e u - d & - a - b a and a - m a - r u
b a - u r both have close parallels in the epic and accordingly can be considered part of
the source for the antediluvian section, whereas the phrase e r i d a k i n a m - l u g a l - l a
does not have any counterpart.
Leaving aside for the moment the third correspondence, between e g i r a - m a - r u
b a - u r - r a - t a and a - m a - r u . . . . b a - u r - r a - t a and utilizing what must have
formed part of the source—names of kings, reigns, the formula A k i b a - s u b - b 6 - e n ,
etc.—we can reconstruct the source as follows:
nam-lugal an-ta en-d^-a-ba

eridaki £ -1 u -1 i m m u 28,800
£-1 a l - g a r m u 36,000
eridaki b a - s u b - b e - e n
b a d - 1 i b i r a ki e n - m e - e n - l t i - a n - n a m u 43,200
etc.
a-ma-ru ba-ur
"When the kingship was lowered from heaven:

(In) Eridu(g): A-lulim(ak), 28,800 years
Alalgar, 36,000 years
IdropEridu(g).
(In)Bad-tibira(k): En-men-lu-Anna(k), 43,200 years
etc.
The Deluge swept in."
There are in this reconstruction several points worth noticing. First its abrupt, almost
account-like, form agrees with our assumption that the source for the antediluvian section
oi.uchicago.edu
66 THE SUMEBIAN KING LIST
the B branch we can do little more than make a guess. The two which show
traces of the insertion, P5 and K, both belong to the small group of manuscripts
deriving from Y. Only one other text, S, belongs to that group, and S also—
it is too fragmentary to give safe indications—may have had the antediluvian
section. Since two of the texts in the group have the antediluvian section and
the third may have had it, it seems a likely assumption that the insertion of
the section was originally made in their common ancestor, Y, and inherited
by the others. This assumption is the more probable because the scribe who
wrote Y must be considered a person especially likely to have searched for
and utilized additional material. We know that this scribe subjected his text
of WB was of a type similar to W-B 62. Such a form, introduced by short catchwords culled
from the surrounding text, is also what we should expect to find in a concentrated excerpt
from a longer literary composition.
Of most interest, however, is the fact that the author of this document apparently has
no intent whatever to present the dynasties as successive. He defines his period by an upper
limit: "When the kingship was lowered from heaven," and a lower: "The Deluge swept in";
but his phrase "I drop the city A. In the city B . . . . " does not imply that the dynasty
of B followed that of A in time; rather it seems chosen to avoid giving this impression. This
view, that the antediluvian dynasties were more or less contemporaneous, is clearly incom-
patible with the King List proper, which directly aims at following the route of "the king-
ship" from one city to another. If the two sources were to be harmonized it was therefore
necessary to supplement the vague framework of the antediluvian section, to establish where
the kingship was when it was sent down from heaven, and then to follow it in its vagaries on
earth. That the scribe who incorporated the section in WB realized this we may see from
the phrase e r i d a k i n a m - l u g a l - l a , "the kingship (was) in Eridu(g)," which has
no counterpart in the epic and therefore is likely to have been added by the scribe himself.
That this line does not come from his source is also indicated by the late form used here,
n a m - l u g a l l - a , in contrast to n a m - l u g a l in the line which he copied from the
original: n a m - l u g a l a n - t a ' e n - d & - a - b a (cf. pp. 67 f.).
It remains to discuss the lines e g i r a - m a - r u b a - u r - r a - t a n a m - l u g a l
a n - t a e u - d & - a - b a k i § i ki n a m - l u g a l - l a which form the actual link between
the antediluvian and postdiluvian sections. The only new thing which these lines bring is
the preposition e g i r . . . . t a , "after"; otherwise they are a mere repetition of the for-
mulas used higher up in the section. Since these lines would be completely out of place
in a separate list of antediluvian rulers but are eminently suitable to link an antediluvian
section to a postdiluvian and since, moreover, they contain the form n a m - l u g a l l - a
which is characteristic for the scribe's own efforts (see above), we can hardly doubt that
they were written by the scribe himself on the basis of the earlier phrases to form the neces-
sary connecting link between the two sections which he was joining. No importance can
thus be attached to the similarity of e g i r a - m a - r u b a - u r - r a - t a in the lines
here discussed and the line a - m a - r u . . . . b a - u r - r a - t a in the epic. This simi-
larity is purely accidental and due to the reuse in WB of a phrase, a - m a - r u b a - t r r ,
common to both versions. The preposition, which is the only new thing in the phrase, is
expressed differently in the two texts: PBS V, No. 1, has . . . . a - 1 a ; WB, e g i r . . . .
a- t a .
oi.uchicago.edu
TEXTUAL PKOBLEMS 67
of the King List to a thorough grammatical revision,120 and a man who took
so much trouble to improve the form of his copy would hardly be less interested
in improving and completing the content.
Clearer than in the B branch is the situation in the A branch; for the point
in WB's descent from the original at which the antediluvian section must have
been inserted can be determined with a fair degree of certainty from a grammat-
ical peculiarity common to the beginning and end of this version. The last
section in our manuscripts of the King List generally shows peculiarities of
some kind or other, for since the scribes as a rule copied from somewhat older
originals they had to add new material—names and reigns of one or more new
rulers—to bring their copies up to date. In these additions which the copyists
themselves composed their scribal habits are apt to show, so that we often
notice differences from the earlier parts of their copies, where they had merely
taken down the text of their originals sign by sign. A difference of this kind
in WB is the change in the formula for introducing new dynasties from Aki(- a)
N. 1 u g a 1 - & m , "In A N. was king," which is used in the main body of
the text, to A ki -a N. l u g a l (!), "In A N. (was!) king," which appears
at the end: u r f k l - m a u r - [ d n a m m u ] l u g a l (!) (viii 9) and i - s i -
i nki - n a i § - b i - i r - r a 1 u g a 1 (!) (viii 23). As will be noted, the verb
"was" (- km) is omitted in the latter form, and the name (Ur-Nammu(k),
Ishbi-Irra) is merely placed beside the substantive with which it is to be identi-
fied ( l u g a l ) . This change in the formula at the very end of the text can
only be explained by assuming that the section in which the new construction
appears was added by a scribe who was bringing his copy of an older original
up to date and who was accustomed to using this construction without -& m.
The construction without - k m is also found, however, in the antedilu-
vian section, where we have [e r i] d aki n a m -1 u g a 1 -1 a (!) , "(In) Eri-
du(g) (was!) the kingship" (i 2), and k i § iki n a m - l u g a l - l a ( I ) ,
"In Kish (was!) the kingship" (i 42). Here the name (Eridu(g), Kish) is
merely placed beside the substantive ( n a m - l u g a l l - a ) , exactly as in
the closing section of the list. It is further noteworthy that the copyist here
uses the form n a m - l u g a l l - a which is characteristic for late texts,121
whereas he uses n a m - l u g a l in the formula for change of dynasty, Aki
b a - § u b - b 6 - e n n a m - l u g a l - b i B ki - § 6 b a - t i i m , which he has
modeled on the formula in his source for the King List proper (Aki gi5t u k u 1
120
The changes from Aki *i5t u k u l b a - s i g . . . . t o A k i b a l a - b i b a - k u r ,
from 1 - a& to i n - as, from i b - a B to i n - a k - e s , and from l u g a l to l u g a l - e - n e
were made in this text.
121
GSG § 121.
oi.uchicago.edu
b a - a n - s i g n a m - l u g a l - b i B ki - § 6 b a -111 m) , and of course

in the postdiluvian sections, where he merely copies that source.122 The con-
struction without - k m occurs elsewhere in the antediluvian section also.
The first time the scribe uses the formula for introducing new dynasties (i 3)
he writes e r i d a k i d - l u - l i m l u g a l ( ! ) , "(In) Eridu(g) A-lulim(ak)
(was!) king," which is exactly like u r lki - m a u r - [dn a m m u] 1 u g a 1 (!)
and i - s i - i n k i - n a i § - b i - i r - r a l u g a l ( ! ) at the end of the list.
Then he completely forgets that there is a special formula with 1 u g a 1
for introducing new dynasties (i 11-12 and 20-21), clearly because his source
for the antediluvian section did not use it. When at last he discovers his for-
getfulness, he looks up the formula in his source for the postdiluvian section
and writes correctly z i m b i r k i e n - m e - e n - d u r - a n - n a l u g a l -
km (i 26-27) and § u r u p p a kki u b u r - t t i - t i i l u g a l - l t m (i 32-33).
The tendency to omit - k m which we notice in the antediluvian section
whenever the scribe is not copying from his original of the King List proper
and the similar omission of - h m in the additions at the end to bring the
list up to date show clearly that both sections were inserted by the same hand.
Since this hand has added the lines introducing the dynasty of Isin it must
obviously be dated later than the beginning of that dynasty. As a likely place
we may suggest the reign of Shu-ilishu, where traces in WB point to the end
of an earlier version. It will be noted that Shu-ilishu is written with the deter-
minative for divinity in line 25, which deals with his own reign, but not in
line 27, where he is mentioned as father of I(d)din-Dagan. This inconsistency
is most naturally explained if we assume that our text has descended through
a version written under Shu-ilishu and ending with him. In such a version
his name would be written with the determinative for divinity, for the scribe
would, of course, acknowledge the divinity of the ruling monarch. Since the
Shu-ilishu version stopped with that ruler, the following line, where he is
mentioned as father of I(d)din-Dagan, must be the work of a later scribe bring-
ing the Shu-ilishu copy up to date. Such a later scribe, writing after Shu-
ilishu^ reign, might naturally consider it unnecessary to give him the sign
of divinity.
132
Cf. also his use of n a m - l u g a l in the phrase n a m - l u g a l a n - t a e n - d e -
a - b a (i 1 and 41), which he probably took over as it stood from his source for the ante-
diluvian section, while he had to make up the line A ki n a m - l u g a l - l a himself.
oi.uchicago.edu
Ill
CRITICAL EDITION OF T H E TEXT, WITH
TRANSLATION AND NOTES
As basic text that of WB, collated with photographs of the inscription, has been
chosen. Letters in the margin show in which mss. each line or traces of it are extant,
even if these traces amount to no more than a blank space indicating that the scribe
found the line destroyed in his original. For definite omission of a line in a ms. we have
inclosed the ms. symbol in angle brackets: <Su2>. When—as is the case with P2—
lines of the text can be supplied from a final summary, the symbol is added in parenthe-
ses: (P2).
The numbering of the lines is based on the form of the text presented by WB (we
follow Langdon's numbering of the lines in OECT II). In emendations supplying pas-
sages presumably lost by WB in the course of transmission the lines have been num-
bered with the number of the preceding WB line followed by letters in alphabetic order,
e.g. 41a, 416, 41c.
Portions of the text which presumably represent later additions are indicated by the
use of smaller type. Passages presumably contained in the original but lost in WB have
been supplied in angle brackets: < >. Emendations of the text of WB to a form which
must be considered closer to that of the original have been added in parentheses:
(emend to . . . . ). In cases where it is doubtful whether the text of WB or that of
other mss. is preferable, the variant has been added in parentheses: (var. . . . .) or
(or . . . .). All restorations and emendations based on readings found in other mss.
are printed in letter-spaced roman; for conjectural restorations italics have been used.
Only the more important variants are noted in the text; those of less importance will
be found in the notes, where the reasons for all emendations etc. are given in detail.
Variants in the formulas have been discussed on pp. 28-42 and are not reconsid-
ered here.
In quoting Sumerian and Akkadian Words or names in English context, we have as a
general rule tried to give or to approximate toward the oldest form. Thus we prefer
Ubar-Tutu(k) to Ubur-Tutu(k), as the latter seems a later development. Similarly
we use Man-ishtushu in preference to Man-ishteshu etc. The grammatical form chosen
as basis for our English renderings of Sumerian words and names is the undeclined
stem. The amissible final consonants (sometimes syllables!), which were probably
indistinctly pronounced and which when Sumerian was a living language were not ex-
pressed in writing except before grammatical affixes beginning with a vowel, we have
usually inclosed in parentheses. In certain names, however, where forms based on
later explicit writings are already more or less current in English, e.g. Uruk, Shuruppak,
etc., we have retained those forms and have not added parentheses. Similarly we have
not considered it feasible to change well established forms such as Ur and Kish to
Uri(m) and Kishi. In Akkadian names we have allowed ourselves even more freedom,
giving the names sometimes with, sometimes without, case endings and mimatidn. In
our choice of form we have here largely been guided by the form found in the King List.
69
oi.uchicago.edu
70 T H E SDMERIAN KING L I S T
Col. i1
WB 1 [n a m] -1 u g a 1 a n - t a en 2 -de—a-ba 3
ki
WB [e r i] d a nam-lugal-la4
ki 5
WB erida d-lu-lim6 lugal7
WB mu 28,800s i - a5
10
WB 5 a -1 a 1 - g a r mu 36,000n i - a5
WB 2 lugal
WB m u <- b i>12 64,800 f b - a5
WB e r i d aki b a - § u b (~be-en)u
WB n a m - l u g a l - bi bad-tibiraki-§&
WB 10 b a - t li m16
WB b a d -1 i b i r aki 17 e n - m e - e n - l i i - a n - n a 1 8
1
The antediluvian section (i 1-42) is not an original part of the King List but was added
later (see pp. 61-68). We have indicated in the notes the chief points on which the form it
takes in WB differs from that found in PBS V, No. 1, in W-B 62, in epic fragment K 11624
(see p. 59, n. 113), and in Berossus. The last is quoted from Paul SchnabeFs edition, Berossos
und die babylonisch-hellenislische Literatur (Berlin, 1923) 2. Teil. The forms of the Greek
letters used in our efforts at reconstruction are patterned after the letters found in the Greek
transliterations of Sumero-Akkadian texts published by Pinches in Society of Biblical
Archaeology, Proceedings XXIV (1902) 108-19.
2
Thus according to photograph (cf. 1. 41, which has this same spelling), not UD-DU = e as
in Langdon's copy.
8
On the relation of this phrase to phrases in PBS V, No. 1, see pp. 58-59 and 64, n. 119.
4
On the grammatical construction used here see pp. 67 f.
5
PBS V, No. 1, agrees with WB on Eridu(g) as the first city. W-B 62 has SA-Aki, i.e.,
k u 6 - a k i (on this reading see n. 126 below), and Berossus has Babylon. The difference
must be due to a natural tendency in the narrator or scribe who was handing on the tra-
dition to locate the first kings in his own city (cf. pp. 59 f., esp. n. 114). Most likely the
tradition was original in Ku3a(ra) and was in early times adopted—and adapted—in
Eridu(g). These two cities were situated fairly near each other (see Poebel, PBS IV 1,
pp. 121 f.; Zimmern in ZA n.F. V [1930] 255, n. 1; Ungnad, Subartu, pp. 28-30). That they
were separate cities is rightly stressed by Landsberger (Der kultische Kalender der Babylonier
und Amyrer, 1. Halfte [Leipzig, 1915] p. 73, n. 1), but the close connection in which they
occur in Sumerian literature suggests intimate cultural relations (cf. ibid, and Zimmern,
loc. cit.). Finally the tradition was appropriated by Babylon, as happened with so much
ancient lore in Neo-Babylonian times (see e.g. Langdon, OECT II 2; Zimmern in ZDMG
n.F. Ill [1924] 27). Sidney Smith (EHA, p. 19) suggests that the appropriation "depends
on a New Babylonian reading of the actual cuneiform signs used by the scribe of list No. 1
[i.e., WB: NUNki], which, when originally written, denoted Eridu"; but this reading is of
course itself an outcome of the extensive appropriation of Eridu(g) lore, so it is not necessary
to assume that our substitution was dependent upon it.
8
W-B 62: [a] -1 u 1 i m (the second sign was identified by Langdon, OECT II 8, n. 1,
as a form of REC} No. 229, which varies with 1 u -1 i m) . Berossus: 'AXopos, in which,
oi.uchicago.edu
C R I T I C A L E D I T I O N O F T H E T E X T , C O L . i 1-11 71
Col. i 1
1 When the kingship was lowered from heaven
the kingship was in Eridu(g).
(In) Eridu(g) A-lulim(ak) (became) king
and reigned9 28,800 years;
5 Alalgar reigned 36,000 years.
2 kings
reigned its 13 64,800 years.
I drop (the topic) Eridu(g); 15
its kingship to Bad4ibira(k)
10 was carried.
(In) Bad-tibira(k) En-men-lu-Anna(k)
as sometimes happens, Babylonian I is represented by Greek r (Zimmern in ZDMG n.F.

I l l 23, n. 1).
The name seems to mean "staghorn" ( a * - l u l i m - a k , "horn of the stag"). The
genitive is not expressed in writing, probably because we are dealing with a "historical"
orthography from the time before the Sumerians expressed grammatical elements in writing,
i.e., before the time of E-Anna(k)-tum. On 1 u 1 i m , "stag," see Landsberger, Die Fauna
des alien Mesopotamien, pp. 98 f.
7
On the omission of - a m here see p. 68.
8 9
W-B 62: 67,200; Berossus: 36,000. On this translation of 1 - a5 see p. 37, n. 85.
10
W-B 62: [&]-lal-gar. Berossus: 'AXairapos, which may represent original *b*rb.rr^foc
in which rr was corrupted to n. Langdon, OECT II 2, and Zimmern in ZDMG n.F. I l l
23, n. 2, assumed <\r^n*j»oc < *<\/*A rAfoc. Landsberger, quoted by Zimmern (he. cit.),
suggests that 'AXavapos is correct and represents an e r n e - SAL form Alabar; but the
usual e r n e - SAL form of g a r is m a r , not b a r .
11
W-B 62: 72,000; Berossus: 10,800.
12
Accidentally omitted by scribe, who was not yet used to the formula; cf. 11. 17, 23,
29, etc.
13
I.e., Eridu(g)'s; see p. 37, n. 86.
14
Accidentally omitted by scribe, who was not yet used to the formula; ef. 11. 18, 24,
and 30.
15
See p. 61, n. 116, where the translation of this formula is discussed.
16
On the reading and translation of Dtr see p. 36, n. 84.
17
Bad-tibira(k) is given as the second city in PBS V, No. 1, and in Berossus, where
the name appears as Pautibiblon (Alexander Polyhistor ap. Eusebium [var. Parmibiblon]
and Abydenus ap. Eusebium), as he UavTifiipXwv (Polyhist. ap. Syncellum), and as he IIau-
Tt,pt0\apj var. he ir6\€us HavnpLpXios (Abyd. ap. Sync). W-B 62 has a dynasty of Larsa
[Footnote 17 continued on page 72]
18
W-B 62: [ e n - m e ] - e n - l t i - a n - n a . Berossus presumably had *^r~\r^ctaiN.
The extant forms are: Polyhist. ap. Euseb.: Almelon ( * & r w n c r t u N <*&nr\zraLN)f
ap. Sync.: 'AM\W (**MHW*cu N < * ^ n t - \ t raiN); Abyd. ap. Euseb.: Amelon (derived in
[Footnote 18 continued on page 72]
I
oi.uchicago.edu
WB 12 mu 43,20019 l-a5
WB en-me-en-gal-an-na20
WB mu 28,80021 i - a5
d
WB 15 dumu-zi s i pa 2 2 m u 36,00023 l - a 5
WB 3lugal
WB mu-bi 108,000 fb-a5
ki
WB b a d -1 i b i r a ba-sub-b^-en
[Footnote 17 continued from page 71]
as second dynasty and that of Bad-tibira(k) as third. On the reading of b a d -URUDU-
NAGARki as b a d -1 i b i r a ki see Ungnad in ZA XXXI (1917/18) 276 and Langdon in JRAS,
1923, p. 258, n. 2. References to the city may be found in RLA I 389 f. and in the literature
quoted there. The identity of Bad-tibira(k) and *IIai>Tt0i/3Xa was noted by Langdon in
JRAS, 1923, pp. 253 and 258, n. 2. Zimmern in ZDMG n.F. I l l 25, n. 1, suggests that an
original Uavnîpa became IlavTipipa and then, under influence of |8(/3Xos, UavripLpXa, while
Lehmann-Haupt in RLA II 5 prefers a slightly different derivation: dittography of /3 and
"der bekannte lautliche Wandel von p zu A."
The dynasty of Larsa which W-B 62 inserts before the Bad-tibira(k) dynasty consists of
4
[ ] - k i - d u - ru n1 - n u § a k i n k i n m u 72,000
5
[ a] 1 i m - r m a1 m u 21,600 «[ ] r b i1 (?) - g a r
7
[2] l u g a l 1 a r s aki - m a
". . . . kidunnu 72,000 years;

. . . . -alimma(k) 21,600 years . . . . ;
2 kings of Larsa."
The insertion of this dynasty must be considered due to local patriotism in a copyist who
wanted to see his own city, Larsa, represented among the antediluvian cities (thus also
Langdon, OECT II 3, and Zimmern in ZDMO n.F. I l l 27). Sidney Smith (EHA, pp. 20 f.)
stresses the many other differences between WB and W-B 62 and concludes: "The scribe
of the second list [i.e., W-B 62] did not therefore take the first list [i.e., WB] and wilfully
alter it, simply to glorify his own city; to all appearance he has recorded numbers, and prob-
ably names, based on authority which we have at present no reason to consider less than
that of the other list." This objection is correct in so far as WrB 62 cannot be just the ante-
diluvian section of WB gone astray (see p. 58, n. 107) but must represent a separate branch
of the (literary!) tradition concerning the antediluvian kings. It cannot, however, weaken
the case for considering the Larsa dynasty of W-B 62 a secondary insertion made by a
scribe from that city.
The name of the second of these Larsa kings is usually read [ ] - u k (?) - k u (?) (thus
Langdon, OECT II 2, and Zimmern, op. ait. p. 20); but the last sign is clearly - m a (see
Langdon's copy, OECT II, PL VI), and the first sign can be part of a 1 i m . A restora-
tion [ l a l - u r - a ] l i m - m a is therefore perhaps worth considering. From LudM Ml
nSmeqi we know Lal-ur-alimma(k) as an ancient ruler of Nippur (PBS IV 1, p. 42, n. 1;
Langdon, Babylonian Wisdom [London, 1923] p. 51, 1. 25 and n. 3). Our scribe can have
known that name and used it when he made up his spurious dynasty of Larsa.
[Footnote 18 continued from page 71]
same way), ap. Sync: 'AjutAXapoy (probably restoration of *b,nr^*////*ozwM <*&r\r\eraiN
influenced by the similar name MeyaXapos). In an Assyrian omen text (KAR> No. 434
rev.[?] 14), as first noted by Weidner (see Langdon, OECT II 3, n. 3, and Zimmern, op.
cit. p. 23, n. 3), the name occurs as am-me-lu-an-na. This latter form, which seems to rep-
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COL. i 12-18 73
12 reigned 43,200 years;

En-men-gal-Anna (k)
reigned 28,800 years;
15 divine Dumu-zi(d), a shepherd, reigned 36,000 years.
3 kings
reigned its 108,000 years.
I drop (the topic) Bad-tibira(k);
resent the pronunciation of the name in later periods, is obviously the basis for Berossus'
Berossus agrees with WB in placing ""A/z/zeXop as the first ruler of the dynasty. W-B 62
has Dumu-zi(d) as the first, En-men-lu-Anna(k) as the second, ruler. After *'AWJL6\WP Be-
rossus has a ruler 'Awepoov (Polyhist. and Abyd. ap. Euseb.: Ammenon, ap. Sync: 'Awe-
vo)v), whom Langdon (OECT II 3) identifies with En-men-gal-Anna(k). Langdon is fol-
lowed by Smith (EHA, p. 19); but Zimmern (op. cit. p. 23, n. 4) has rightly seen that
En-men-gal-Anna(k) must correspond to Berossus* MeyaXapos, Amegalaros, etc. (see be-
low, n. 20). He therefore assumes (following Weissbach in Schnabel, Berossos, p. 179, and
King, Legends of Babylon and Egypt, p. 32) that the tradition used by Berossus had a ruler
En-me-nunna ( = 'knpevw) here. This En-me-nunna would then be identical with En-me(n)-
nunna of the 1st dynasty of Kish and would, like Dumu-zi(d), appear both as an antedilu-
vian and as a postdiluvian ruler. To us this construction seems most unlikely. We prefer
to consider 'Awtvuv with his reign of 12 sar (i.e., 43,200 years) a mere doublet of the pre-
ceding *'Afj.fj.€\a>p with 13 sar (i.e., 46,800 years) and to assume that one of these got into the
text from a variant reading written in the margin or above the line in some early manuscript
of Babyloniaca. Note also that the 12 sar of 'Awevw correspond exactly to the figure for
En-men-lu-Anna(k) given in WB.
19
W-B 62: 21,600; Berossus: 46,800.
20
W-B 62 omits this ruler. Berossus had presumably *'AfXfieya\avos (Polyhist. and Abyd.
ap. Euseb.: Amegalaros [var. Amelagaros], ap. Sync: MeyaXapos). As Zimmern has seen
(ZDMG n.F. I l l 23), this reflects earlier 'AneyaXapos, which we assume to derive from a still
earlier * ^ r\W/C r A / s ^ N o c < * ^ h n e r <\ A ^ N O C . This *'AweyaXapos corresponds to
En-men-gal-Anna(k) (late form: *Ammegalan(na)) as does *^ju/)ie\a»'toEn-men-lu-Anna(k)
(late form: Ammeluan(na)); cf. Zimmern, he, cit.
21
Berossus: 64,800.
22
W-B 62: [ d d u m u ] - z i s i p a . Berossus: Aawws woi^v or perhaps better *Aa«fos
iromr\v (Polyhist. ap. Sync: Aawos TCO^P) Abyd. ap. Sync: Aaws iroifx^p. Polyhist. and
Abyd. ap. Euseb. both give the name as Daonos). W-B 62 places this ruler as the first
king of the dynasty, before En-men-lu-Anna(k), while Berossus, whose intrusive 'A/j/iowr
should be disregarded (see n. 18 above), agrees with WB in placing him third. The variant
forms of the name in Berossus, Aawws andAaws (<Aaa>[*>o]s), may represent an original
Aawfos (original T- corrupted to N} as suggested by Sayce [quoted in OECT II 3, n. 2]),
for which we may compare the late Babylonian form Duûzi and the form u l j (Ta : uz;
used by the §abeans of Harran in the Middle Ages (D. A. Chwolson, tfber TammUz und die
Menschenverehrung bei den alien Babybniern [St. Petersburg, 1860] p. 38; OECT II 3, n. 2;
cf. Burrows in Orientalia No. 7 [1923] p. 55 and n. 1). Zimmern in ZDMG n.F. I l l 23 rightly
assumes the existence of Babylonian forms *Da°uzi and *Damuzi as variants of Dumuzi.
23
Berossus agrees with WB on 36,000; W-B 62 gives 28,800.
oi.uchicago.edu
74 T H E SUMEKIAN K I N G L I S T
WB 19 nam-lugal-bi l a - r a - a kki<-$e> b a -
t lim
WB 20 l a - r a - a k k i 2 4 e n (!) - s i p a - z i - a n - n a 2 8
WB mu 28,80026 i-a5
WB 1 lugal
WB l a - r a - a k ki ba-sub-be-en
WB 25 nam-lugal-bi z i m b i r<ki> - § 6
WB b a - 1 ti m
ki27
WB zimbi r en-me-en-diir-an-na28
29
WB lugal-am m u 21,00030 i-a5
WB 1 lugal
ki
WB 30 zimbi r ba-&ub-b6-en
WB n a m - l u g a l - b i §uruppakki<-^)31 ba-
tiim
WB § u r u p p a k ki ubur-tu-tu32
24
Thus also W-B 62. PBS V, No. 1, omits the determinative k i . Epic fragment
K 11624 writes l a 7 - r a ~ a k - a k l , which corresponds closely to Berossus' *Aapa7xa (Poly-
hist. ap. Sync: h Aapayxw, ap. Euseb.: Lanchara [vars. Ilanchara, Chanchara]).
PBS V, No. 1, W-B 62 (if we ignore the spurious Larsa dynasty), and Berossus agree
on Larak as the third antediluvian city. Note, however, that in Berossus *Ebe8a)payxos,
who corresponds to En-men-dur-Anna(k) of Sippar, has been counted as the last king of
Bad-tibira(k), so that Berossus may originally have had Sippar as the third and Larak as
the fourth city. This is strongly supported by the fact that K 11624 actually shows Sippar
preceding Larak.
26
Thus according to collation. W-B 62: [e] n - s i p a - z i - a n - n a ; K 11624: e [n -
s i p a - z i - a n - n a ] ; Berossus: "kfxeîvos (Polyhist. ap. Euseb.: Amempsinos, ap.
Sync: 'k/xefixf/ivos). The form *kfieiA\pivos probably represents a late form of E n -
s i p a (d) - z i (d) - A n n a (k), *am-sip-zi-an-na. Zimmem in ZDMG n.F. I l l 24 suggests
that the initial 'A/* represents an e m e -SAL form of e n , a m , also found in *'k(j.ii€\uv —
am-me-lu~an-na = e n - m e - l u - a n - n a etc. The following € can be a misreading of
C , and n may stand for carelessly written W OY) above which the scribe had added
T to make it clearer (rtr) but which was copied as fir (the explanatory v being con-
sidered correction of an omission). This would give a development Ensipzi(a)n(na) =
*^t^CtrtN0C > *&M€rtlNOC > ^ H C M Y I N O C .
26
W-B 62, Berossus, and probably K 11624: 36,000.
27
Epic fragment K 11624 has Sippar before Larak; and Berossus, who now does not list
Sippar, may also originally have had that order (cf. n. 24).
28
W-B 62: e n - m e - d u r - a n - n a ; Berossus: *Efc8upayxos (Polyhist. ap. Euseb.:
Eved6ranchos [var. E(v)edo{r)anchos], ap. Sync: Efedupaxos; Abyd. ap. Euseb.: E(ve>-
doreschos, ap. Sync: Efa&copevxos. The form given by Abyd. may be explained through
^ r > me > 4C The name of this king is found also in the bard ritual K 2486 (Zimmern,
Beitrdge zur Kenntnis der babyhnischen Religion. Die Beschwdrungstafeln Surpu, Ritual-
oi.uchicago.edu
19 its kingship to Larak was carried.
20 (In) Larak En-sipa(d)-zi(d)-Anna(k)

1 king
I drop (the topic) Larak;
25 its kingship to Sippar
was carried.
(In) Sippar En-men-dur-Anna(k)
became king and reigned 21,000 years.
1 king
30 I drop (the topic) Sippar;
its kingship to Shuruppak was carried.
(In) Shuruppak Ubar-Tutu(k)

tafelnfur den Wahrsager, Beschworer und Sanger ["Assyriologische Bibliothek" XII (Leipzig,
1901)] No. 24 obv. 1), where it appears as e n - m e - d u r - a n - k i . This form is ob-
viously behind the variants in Berossus, which, as Zimmern has pointed out (ZDMG n.F.
I l l 24), reflect an original *Eife8c*)payxos. Of the two forms e n - m e - d u r - a n - n a
and e n - m e - d u r - a n - k i , the first seems original; for the sign >4r (NA) could
easily be misread 4r (KI) if slightly damaged, and we must reckon with influence from
the name of the ziggurat in Nippur, Dur-an-ki(k). Note the similar wavering in the name
of the ziggurat in Larsa: e - d u r - a n - n a / 6 - d u r - a n - k i (Ebeling in RLA II 275).
29
On the sudden occurrence of 1 u g a 1 - a m here and in the following dynasty see p. 68.
30
W-B 62: 72,000; Berossus: 64,800.
81
Written su-KUR-RUki; W-B 62: su-KUR-LAMki; K 11624: LAM-f KUR-RUki. Berossus has
no dynasty of Shuruppak but assigns the rulers of this city to Larak (see nn. 32 and 34 be-
low). Of these writings only the first, su-KUR-RUki, is correct. LAM+KUR-Rukl is in reality
a phonetic writing of the name of a different city, Lamkurru, older Numkurru (see n. 115
below). Its use to designate Shuruppak (su-KUR-RUki) must be due to confusion of the
similar signs su and LAM (see Albright and Kramer, quoted below). Similarly the writing
su-KUR-LAMki is probably due to confusion of RXJ and LAM, which also resemble each other.
On the problems presented by these writings see Poebel, PBS IV 1, p. 45, n. 1; Albright in
JAOS XLV (1925) 206 f.; Zimmern in ZA n.F. V (1930) 252, n. 5; Kramer in JAOS LII
(1932) 117-19.
32
W-B 62: u b u r - t u - t u ; K 11624: u b u r - 1 [ u - t u] ; Berossus: 'Qriap-nys,
'Apdarw (so Polyhist. ap. Sync; ap. Euseb.: Otiartes). The Ninevite version of the Gil-
games epic (Tablet XI 23) has ubar-tu-tu. The form transmitted from Berossus by Polyhis-
tor, ULTtkpmc, probably represents an older *amAjry\c (according to Zimmern in ZDMG
n.F. I l l 24, n. 3; first suggested by Oppert) or even, utilizing the variant 'ApSan?*,
*cK.nAF>ATrHC-, which may perhaps have derived from original *ou*Aprtu-ruic through
a damaged text *axnAfm^rmC. The form with a, Ubar-Tutu(k), attested by the Gil-
games epic and by Berossus, seems to be the more original, as Ubur-Tutu(k) is naturally
oi.uchicago.edu
WB 33 1 u g a 1 - a m m u 18;60033 i-as
WB 1 1 u g a l34
WB 35 m u - b i 18,600 lb-as
WB 5 uruki-me-es
WB 8 lugal
WB m u <-6i> 241,200 1 b -a 5
WB a-ma-ru b a - u r RA 35 TA 35
WB 40 egir a-ma-ru ba-ur-ra-ta
WB nam-lugal an-ta en-dd-a-ba
WB k i § i ki 37 n a m -1 u g a 1 - 1 a38
WB k i § iki g a (?) - [ . . ] - ft r39
WB 1 u g a 1 - km
WB 45 mu 1,200 i-a5
WB L 2 Su 2 g u 1 - 1 a n i d a b a a n - n a - d a s i k i l40
d r 1
WB L 2 Su 2 mu 960 41 i-aB
explained as a development from Ubar-Tutu(k) by assimilation of a to the preceding b

and following w's. Berossus lists 'Ortapr^s as king of Larak (see n. 31). W-B 62 does not list
him as king but mentions him only as father of an otherwise unknown SU-KUR-LAM who there
appears as father of Zi-u-sud-ra. As Landsberger (see Zimmern in ZDMG n,F. I l l 21, n. 1)
has convincingly suggested, SU-KUR-LAM should be read simply Shuruppak, and this ruler
owes his existence to an old misinterpretation of the phrase z i - u - s u d - r a dumu
SU-KUR-LAM (i.e., S u r u p p a k ) , "Zi-u-sud-ra, the man from Shuruppak," in which the
city name was taken to be a personal name: "Zi-u-sud-ra, the son of Shuruppak." The
sign - g i / - g e which follows the name SU-KUR-LAM in W-B 62:10 (i.e., s u r u p p a k -
k i) is explained by Zimmern, op. cit. p. 20, n. 1, as "Vokalverlangerung." Could it pos-
sibly represent a thoughtless rendering of dictated(!) - k ei ? Langdon's reading A r a d -
g i n and his combination of this with 'ApSarris (OECT II 3) do not carry conviction.
33
Berossus: 28,800; the same figure is assigned to SU-KUR-LAM in W-B 62.
34
The omission of Zi-u-sud-ra is curious. It would be understandable, however, if the
source used by WB was an excerpt from a larger legendary composition (cf. p. 64, n. 119)
in which the antediluvian kings down to Ubar-Tutu(k) were treated together in one sec-
tion of the narrative, that dealing with the founding of antediluvian cities, while Zi-u-sud-ra,
as hero of the Flood story, was dealt with separately in a following section on the Flood.
Such separation could have caused the excerptor to overlook him.
W-B 62 lists z i - U 4 - s u d - r a " as son of the preceding SU-KUR-LAM (see n. 32) and
gives him a reign of 36,000 years. Epic fragment K 11624 correctly has him (as z i - U4 -
s u d - r a ) after Ubar-Tutu(k), as son of that ruler. Berossus lists him after 'firtapr^s
(Ubar-Tutu(k)) as king of Larak (see n. 31) and gives him a reign of 36,000 years. The
name is written Xisuthros (Polyhist. and Abyd. ap. Euseb.), Et<n>u#pos (Polyhist. ap.
Sync), Xicrov&pos (Abyd. ap. Sync), and Xeuntipos (Abyd. ap. Euseb. Praep. ev.) or 2«rn?pos
(Sync) (see Schnabel, Berossos, p. 266, n. 21), which Zimmern (ZDMG n.F. I l l 24) rightly
derives from an original Ztaovdpos = Zisudra, which may be compared with the writing
zi-sud~ra in KAR, No. 434 obv.(?) 7 (cf. Zimmern, op. cit. pp. 21, n. 2, and 24, n. 5) and with
z i - u i - s u d - r a in K 11624. Should K2054 rev. iv 9 also be read z i - s u d - r a ( l )
instead o f z i - s u d - d a a s given in CT XVIII (1904) PI. 30? The reading of the name,
oi.uchicago.edu
33 became king and reigned 18,600 years.

1 king
35 reigned its 18,600 years.
5 cities were they;
8 kings
reigned their 241,200 years.
The Flood swept36 thereover.
40 After the Flood had swept thereover,
when the kingship was lowered from heaven
the kingship was in Kish.
I n Kish G a . .ur(?)
became king
45 a n d reigned 1,200 y e a r s ;
Destroyed! To the heavenly Nidaba is it clear!
reigned 960 y e a r s ;
which appears in a variety of writings in Sumerian and Akkadian texts, has been widely dis-
cussed (see e.g. Poebel, PBS IV 1, pp. 48-50). We have adopted Zi-u-sud-ra as the form
which we consider most likely to be correct.
88
Dittography from b a - u r - r a - t a in the next line.
86
The verb u r —labatu and sapdnu, "iiber etwas hinwegfahren oder "-streichen, wozu
das gleichzeitige Beseitigen aller Unebenheiten der Oberflache als konstitutiv fur die Be-
deutung von sapanu hinzukommt" (Landsberger in ZA n.F. I l l [1927] 216, n. 4). On
u r in connection with a - m a - r u see Langdon, OECT II 9, n. 4.
37
The sign with which the name of the city Kish is always written represents the head
of an ungulate quadruped. Since the Sumerian word for "to run," "to gallop," k a §,
would furnish a good etymology for a word k i § or k a & denoting a swift-running hoofed
animal, we may assume that the sign originally represents an animal of that name. Its
use to express the similar-sounding city name is thus purely phonetic.
38
Here ends the secondary section on the antediluvian period which was added to the
King List. See pp. 55-68 and esp. p. 64, n. 119.
89
We have, with some hesitancy, accepted Langdon's reading g a\ In the photo-
graph the sign shows as w&f'/M. There is a break in the tablet before and after this
sign. Langdon (OECT II 9, n. 5) may be right in assuming that nothing is missing in
the break before GA(?); in the break after GA(?) the photograph seems to show faint rem-
nants of wedges. Langdon's combination of the name with Polyhistor's Evek'sios, Efoqxoios
(see n. 115 below) is untenable (see Zimmern in ZDMG n.F. I l l 30, n. 2). The latter name
corresponds to En-me(r)-kar.
40
Clearly not a personal name but a copyist's note that the passage was damaged
( g u l l - a ) in his original. He piously adds: "To heavenly Nidaba (lit., 'with [-da]
Nidaba of heaven [a n n - a (k) -]') is it clear"; i.e., the goddess of writing herself would
undoubtedly be able to make it out. Remnants of the original name seem to be preserved
in L2, which has r t e1 - [ ] here.
41
L2: « r , which should undoubtedly be restored W, i.e., 1,200; see Poebel in OLZ
XXVII (1924) 263.
oi.uchicago.edu
Col. ii
p 6 L2 Su2 1 [ p a - l a - k f - n a - t i - i m42
L* Su2 [m u 90043 1- a5]
p 6 L2 Su2 [ n a - a n - g i § - l i - i & - m a 44
L2 Su2 [mu 1- a5]
WB P6 Su2 5 b a - b [i (?) - n a46 mu ., 1 -a 5 ]
1 46 r 1
WB P6 Btr-AN-f? - [u m] mu [8]4047 i-a5
48
WB P2 P3 P 6 Sui ka-H-bu-u m
WB P2 P3 SU! mu 96049 i - a5
60 61
WB P2 P3 P 6 Sui q a -1 u - m u (!) mu 84052 i-a5
53 64 55
WB P2 P3 Ps Su, 10 z ti - q a - q i4 - i p
WB P2 Ps Su, mu 90066 i-a5
WB P2 P3 p s <SUl> a -1 a b" mu 60068 i-a6
42
Thus L2; we read the last signs, mm, as 9&W ( - ' t i - i m 1 ) . Su2 has [ p a - l a -
k i ] - n a - i - d i m 7 (cf. its similarly awkward spelling e n - m e - b & r a - g a - e - s i for
En-me(n)-barage-si in ii 2). Pfi: wmmmm, i.e., 'pa 1 -1 [a] - k i - [n a - t i -i m ] . The
name means "reign of righteousness" (with gen. pi. of kiitum).
43
Thus L2. This reading of r<W is preferable; see Poebel in OLZ XXVII 263.
44
Thus L2. Su2: [ ]mmmmmr * vi, i.e., [ n a - a n - g i s j - l i - r i s 1 (?) - m a ; P 5 :
M W M W , i.e., f n a - a n - g i s - l P - t i S - m a ) . The meaning of the name is probably
"May he (a deity?) listen cheerfully." Nangis seems to be the -ivs "adverb" of *nangH,
"to be cheered up," IVi infinitive of nigti, "to be joyous" (AH, p. 446 a). The old form
Uhna(') (cf. Ungnad in MVAG XX 2 [1916] p. 87) instead of later lisme is noteworthy.
45
The reading and meaning of this name are uncertain. WB has f^^l ), i.e., b a -
b [ i ( ? ) - n a ] ; Su2: [ b a - f c i ] - n a ; P 6 : mam, i.e., b a- [i>]i (?) - rn a1 - e . After the
name Su2 gives [ ] - a m , which probably represents [1 u g a 1] - a m . It is unlikely that
this line is original, for there is not room for it in WB, and it is lacking in P6 also.
4$
The reading and meaning of the name are uncertain. WB has according to Langdon's
copy j * - m- ttm/wwMff/Mm. (in photograph ^0*x*»w'»»Mmr*wi* can be seen), i.e., BU-AN-
r
NU1(?)-rtJM1(?). P 5 gives *wwr, i.e., BU-AN-?-UM. Since all the following names are ani-
mal names, we should perhaps expect an animal name here.
47
Preserved only in WB: *» (in photograph %*& can be seen), which probably represents
m (840).
48
I.e., "dog" (kalbum). Note the old form with i, for which Syriac ^V^ and the prop-
er name 2?p may be compared.
49
P 2 : Y/0% i.e., ff (900). WB's figure must be considered closer to the original; see
pp. 19-21.
60
Thus also P6. P 3 : q & - ; Sin: q a - .
61
P 3 and P6 add - u m ; P 2 adds - u m - e . Langdon's transliteration and translation
of WB (OECT II 10) give k a ( = q a ) - l u - m u - m u ; his copy gives q a * 1 u - m u .
A collation shows the latter form to be correct. The name means "lamb" (AH, p. 333;
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COL. ii 1-12 79
Col. ii
1 Pala-kinatim
reigned 900 years;
Nangish-lishma
reigned . . years;
5 Bahina(?) reigned . . years;
Bu-AN-?-um reigned 840 years;
Kalibum
reigned 960 years;
Qalumu (m) reigned 840 years;
10 Zuqaqlp
reigned 900 years;
Atab reigned 600 years;
Poebel, PBS IV 1, p. 111. Poebel's argument for reading this word qalumu is supported
by the writing with q a - in Sui. On this name ef. also Giiterbock in ZA n.F. VIII [1934] 5).
52
P 3 : 900. In Sui m must also represent 900 (Hr). P2's m is ambiguous. Since WB
seems to be the most reliable source for this section as a whole (see pp. 15-23), it is likely
that it preserves the better text.
53
P 3 and P 5 : z u - ; Sui and P 2 with If are ambiguous.
54
Thus also P2, P3, and Sui. P 5 has - q i - .
55
P 2 adds - e . The name means "scorpion" (see Poebel, PBS IV 1, p. I l l ; Giiter-
bock in ZA n.F. VIII 5). The reading Zuqaqlp is proved correct by the fact that it fur-
nishes a parallel to the other animal names in the section. There is nothing to support
Langdon's view (OECT II 10, n. 7) that we should read d u g 4 - g a - g i 4 - i b .
The place of Zuqaqlp and his predecessor Qalumu(m) is different in WB from what it
is in the other lists, where these rulers appear farther down. As shown above (pp. 19-23),
the order of WB is here original.
56
Thus also P 2 : mw, which can only be WW (900). P 3 has—according to Poebel's
copy— msh9 but the photograph in PBS V, PL XCI, seems to have « , i.e., tffW (900).
Sui has m, i.e., I* (600), which has evidently developed from F? (900) through rTO in a
damaged text (see p. 17).
57
P 2 has m- m, P6 a" - m, both of which should obviously be restored r£ - b a 1 . This
agrees with P3, which has [ ] - b a . Sui omits the name. On the original form of the
passage see pp. 21-22. In passing through a damaged text the words m a § - d a d u m u
were lost, so that d - t a b - b a came to figure as name of the successor of k-1ab,
a stage in the development of the text now represented by WB. In the other branch
of the tradition, the B branch, the & -1 a b passage was then damaged once more, as wit-
nessed by the entire omission of the line &-1 a b m u 600 i - a5 in Sui, the blank line left
for m u 600 i - a5 in P2, and the form & - b a (!) in P2, P3, and P5, which is clearly an in-
correct restoration based on a damaged a - [t a b] and made under influence of the fol-
lowing a - t a b - b a .
68
P2 leaves a blank space for the figure and 1 - a8. Sui omits the line entirely.
oi.uchicago.edu
WB P 2 Pa P S Sui 13 <m a 8 - d a d u m u > 6 9 a - 1 a b - b a60 m u

84061 i - a 6
WB P2 P, P« Sui ar-wi-u-um62 dumu ma§63-da-ke4
WB Ps P» Sui 15 m u 72086 i - a6
66 67
WB P2 P, P 6 Sui e-ta -na si pa 1 il an-§6
WB P2 P 3 P* Sui b a68 - e u - d 66»
WB P2 Ps Sut 1 u kur-kur" m u - u n - g i71 - n a72
WB P2 Sui » l u g a l 4 m ' m u 1,56076 (var. 1,500?)
4
i-as
WB P 2 Ps SU! K 20 b a - l i - i b 7 6
WB P 2 P 6 Sui K dumu e - 1 a" - n a - k e4
WB P2 Sui mu 40078 (var. 410) i - a6
WB P 2 P s SU! K e n - m e79 - n u n - n a80 m u 66081 1 -a 5
59
On this emendation see pp. 21 f.
60
Not a new name but merely the genitive of the preceding a - 1 a b , as shown in de-
tail on pp. 21 f.
M
Thus also P 8 : r«r, which, can be restored as Wr (840). Sui's Hr (720) is clearly
secondary and derived from W (840) through Fff in a damaged text. P 2 leaves a blank
line here.
62
P 5 : a r - w i - t i ; P 2 : a r - w i ; Sui: a r - w i - u m ; P s : a r - b u - u m. All of these
forms would seem to derive from that given by WB, a r - w i - u - u m , through dam-
aged texts; cf. p. 17.
88
P 3 adds -EN-.
94
The addition of -EN- in P 3 changes the meaning to "a plebeian" (see p. 17, n. 36).
That this variant is secondary is clearly shown by the close relationship in meaning be-
tween m a § - d a = sabitum, "gazelle," and arw&um, "male gazelle." The form of the
entry also excludes a rendering "plebeian." A detailed discussion of the passage is given
above (pp. 17 f.).
w
Thus also P2. In P, *w should be restored m (720).
• Thus also P2, P,, and P*. Sui has - 1 a - .
*7 This ruler appears as the chief character of the Etana epic (Langdon, "The legend of
Etana and the eagle," Babyloniaca XII [1931] 1-56). He is also mentioned in the frag-
ment K 14788 (CT XXXIV [1914] PI. 18) and in the Gilgames epic (R. C. Thompson,
The Epic of Gilgamish [Oxford, 1930] PL 30:49), where he appears as resident in Hades.
An omen concerning him is KAR, No. 434 obv.(?) 9. The text Ki 1904r-10-9, 8 7 + K 5119
(O. R. Gurney, "A bilingual text concerning Etana," JRAS, 1935, pp. 459-66), which
deals with the seven apkallu's, refers to him in 11. 3-4: [ e - t a - n a s i p a] 1 u* a n - § &
b a - a n - £ / [ e - t a - n a r&u] So ana hatni* i-lu[-<u], "Etana, a shepherd, the one who ascend-
ed to heaven." Most of the references to Etana have been collected by Guterbock in ZA
n.F. VIII 22.
A Sumerian etymology of the name Etana, "Ascender (e d) of heaven (a n n - a),"
was proposed by Poebel, PBS IV 1, p. 112. It has the support of the writing d£ - 1 a - n a
jn Lutz, Selected Sumerian and Babylonian Texts (PBS I 2 [1919]) No. 112 ii 67: z i de -
t a - n a a n - & k u r . . - g a fc 6 < - p a } , "The life of Etana, (the one) ascending heav-
en, the . . . . of the foreign lands, be invoked." (Cf. Gurney, op. cit. p. 462, n. 1. Gurney's
oi.uchicago.edu
13 (Mashda, son) of Atab, 60 reigned 840 years;
ArwPum, son of Mashda, 64

15 reigned 720 years;
Etana, a shepherd, the one who to heaven
ascended,
the one who consolidated all73 lands,
became king and reigned 1,560 (var. 1,500?) years;
20 Balih,
son of Etana,
reigned 400 (var. 410) years;
En-me(n)-nunna reigned 660 years;
explanation of the name as an imperative of & is not probable.) We do not yet, however,
consider this etymology sufficiently well established to adopt a form Ed-ana(k), but keep
provisionally the rendering Etana.
68
Thus also P3, Sui, and perhaps P6, where the sign is damaged. P 2 has ri (?) - i bC?)1 - .
69
P2 and P 3 : - d a .
70 72
P2 and Sui add - r a . Thus also P2 and Sui. P«: - n i .
71 73
Thus also P2. Sui: - g i4 - . Cf. GSG § 142.
74-74 p 2 a n ( j g Ul o m i t . Since the phrase can easily have been omitted by accident (being
overlooked by a copyist or disappearing in a lacuna), whereas it is difficult to imagine why
a copyist should have inserted it if he did not find it in his original, we must assume that
WB preserves the correct text.
78
Thus according to the photograph; Langdon reads 1,500. Sui: ^%%f = OftHF (1,500);
P 2 : #* 4^ff (635), a figure which is no doubt derived from an original 1,500 (r^W) through
a damaged text, difficult to read.
76
Pg and Sui also write b a -1 i - i li • P2 has w a -1 i - i b , which perhaps represents
a mishearing in dictation. K has AN-ILLAD, i.e., Balihu; see E. A. Speiser, Mesopotamian
Origins (Philadelphia, 1930) p. 151. The name occurs also (as ba4i-fyum) on a cylinder
seal seen by Scheil in Aleppo and published by him in MDP VI (1905) 53, No. 11, and
in RA XIII (1916) 11, IV 3.
77
Thus also P 2 and P 6 ; Sui: - [t] a - . For the name as a whole K has KI-MIN, which
refers back to its previous mention.
78
P 2 : 410. Sui's Tft-& also would seem to represent 410 (RF^). It is uncertain which
branch of the tradition preserves the better text.
79
Thus also P2, P5, and Sui; K: - m e n - .
80
Thus also P5, Sui, and K; P2 adds - k e 4 . The name En-me(n)-nunna forms part
of the name of a special type of wasp: d u m u - e n - m e - n u n - n a , "son of En-me(n)-
nunna." See Poebel, PBS IV 1, pp. 113 f.; Landsberger, Die Fauna des alien Mesopota-
mien, p. 41, No. 39a; cf. ibid. p. 132.
81
Thus also Sui; P 2 : 611. As the figure 660 is represented in both the A branch (WB)
and the B branch (Sui) of the tradition, it must be original. The figure of P2, 611, has ob-
oi.uchicago.edu
WB P 2 P 5 Sui K 24 m e82 -1 & m - k i s i ki d u m u e n - m e -

n u n - n a - k e483
WB P 2 PB SUI K 25 m u 900 i-a6
WB P 2 PB Sui b a r - s a l - n u n - n a 84dumu en-me-
n u n - n a84
WB P 2 PB Sui mu 1,200 i -a5
85
WB P 2 PB SU! s u m li g s a - m u g 8 5 d u m u bar-
sal-nun-na86
WB P 2 PB mu 140 i - as
8 7 88
WB P 2 PB 30 t i - i z - k k r dumu s u m i i g sa-mug88
WB PB mu 305(!)89 i-a5
90 90
WB PB il-ku-u mu 900 i-a5
WB PB il-ta-sa-du-um
WB PB Su2 mu 1,20091 i - a5
WB PB Su2 35 e n - m e - e n - b a r a - g e 9 3 - s i (!)94
92
95
WB PB Su2 1u ma-da e l a m ki - m a
viously developed from an original ^ (660) through a text in which the stylus of the
scribe slipped so that it left a double imprint of the < -wedge: r<T.
82
Thus also P2, P6, and Sui; K: m e5 - .
83
Thus also P 2 ; Ps omits - k e 4 . K writes KI-MIN, referring back to En-me(n)-nunna
in the preceding line.
8I 84
~ P 5 omits; P2 adds - k e4. Since d u m u e n - r a e - n u n - n a f - k e j appears
in both branches of the tradition (A branch: WB; B branch: P2), it must be original, and
the omission in P6 is therefore secondary.
85-85The name has been variously read: m e s - z a - m u g (Poebel, PBS IV 1, pp. 74
and 82; Zimmern in ZDMG n.F. I l l 30; Legrain, PBS XIII 11), t u p - z a - a h (Lang-
don, OECT II 10), Dubzah(?) (Weidner in Meissner, Babylonien unci Assyrien II [Hei-
delberg, 1925] 441), m e s - z a - a h (Barton, RISA, p. 348), etc. The last sign of the
name is without any doubt m u g , as read by Poebel, Zimmern, and Legrain. A com-
parison of our sign as it is written in WB and P5 with the way a b / i Ij is written in these
texts (a few lines higher up in the name Balih) establishes this beyond question. More
difficult is the first sign; WB has a clear DUP (REC, NO. 385), while P5 has W, which
may be vu (REC, No. 79) or a simplified form of DUP. Since the scribes of the period from
which our texts date did not distinguish clearly among the signs UM (REC, No. 79), MES
(REC, No. 363), and DUP (REC, No. 385) (see SL, No. 138.1; and note that WB uses
REC, No. 385, i.e., DUP, in such names as Gilgames, Mes-Anne~pada, and Mes-kiag-nun-
na(k), where a reading m e s is unquestionable), no safe conclusions as to the reading
can be drawn from the form of the sign alone. A reliable clue to the reading is given, how-
ever, by the last two signs of the name: z a - m u g or s a - m u g corresponds so close-
ly to the value s u m t i g / s a m d - g of REC, No. 385 (DUP), that there can be no doubt
that the combination represents an old gloss which has got down into the text (for simi-
lar cases see nn. 159 and 291 below). We should therefore read s u m i i g and restore the
older form of the line as s u m u g«*-"»««. In the translation we have rendered this name
oi.uchicago.edu
24 Melam-Kishi(k), son of En-me(n)-nunna,

Bar-sal-nunna, son of En-me(n)-nunna,

Samug,85 son of Bar-sal-nunna,
reigned 140 years;

30 Tizkar, son of Samug,
reigned 305 years;
IlkiP reigned 900 years;
Ilta-sadum
35 En-me(n)-barage-si,
the one who carried away as spoil
Samug, since Sumug is obviously a later pronunciation, derived from Samug by assimila-
tion of a to the following m and u.
86
P 2 and P5 add - k e 4 .
87
Akkadian. Probably abbreviation of a longer, theophorous name such as Tizkar-
d
Samasf "Be mindful of Shamash." Cf. the names with this element listed in Chiera, PBS
XI 1, p. 50, Nos. 20-21.
88-88
P5 adds - k e«. P2: d u m u b a r - s a l - n u n - n a - k e 4 , which is clearly dit-
tography from s u m u g s & - m u g d u m u b a r - s a l - n u n - n a - k e « just above.
That d u m u s u m u g sa-mug(-ke4) is the earlier form of the text is shown also
by the fact that it appears in both branches of the tradition (A branch: WB; B branch: P 5 ).
89
Thus according to the photograph; Langdon reads 306.
90 90
- P6 has [i] 1 - k u - u m - e . Which is correct?
91
Thus also S112: #&, which can only represent an original RK (1,200).
92
P 5 and Su3 omit.
93
Thus also P 5 ; Su2: - g a - e - . On the latter form cf. PBS X 2, No. 5 rev. 9: e n -
m e - rb a r a - g a (?) - e1 - s i , and note the similarly artificial spelling [k i] - 11 a - i - d i 1117
for k t n a t i m in Su2 (n. 42 above).
94
- s i is clear in WB and in the parallel texts. Langdon's reading - g u r - (OECT
II 11) is erroneous (cf. Zimmern in ZDMG n.F. I l l 30, n. 16; Scheil in RA XXXI [1934]
161). En-me(n)-barage-si is mentioned as father of Aka in the tale of Gilgames' war on
Kish (PBS X 2, No. 5 rev. 9; cf. p. 145, n. 20).
95 95
- Su2 omits. P 3 with [ ] «i5t u k [u 1] - b i i [b] - r t a1 - a [n - g u r] and PB with
lii m a - d a e l a [m]rkil - [m] a giSrt u k u l1 - rb i1 fi b1 - r t a1 - a n - rg ii r1 agree with
WB except that Pa has i b - instead of l b - . The omission of this note in Su2 must be
secondary, for it is attested both in the A branch (WB) and in the B branch (P3 and P5) of
the tradition and is accordingly an original feature. Probably Su2 derives through a text
damaged at this place.
oi.uchicago.edu
gi§
WB P 3 P 6 SU2 37 t ukul-bi fb-ta-an-gi'ir95
97 97
WB P 3 PS Su2 lugal-&m mu 90098 l-a 6
WB Pa Su2 a5-k&"
WB Ps Su2 40 d u m u e n - m e - e n100 - b a r a - g e101 -
si(!) l 0 2 -ke 4 1 0 3
WB Pa Su2 mu 625 l-a 5 104
WB Pa Su2 23 l u g a l
105
WB Pa Su2 mu-bi 24,510 itu 3
WB Pa Su2 u4 3 u4 | 105 ib-a5
WB Pa Su2 45 k i s i k i gi§
tukul ba-an106-sig
107
WB P2 Su2 nam-lugal-bi 6 - a n - n a - § e107
WB P 2 Su2 ba- tum
WB P2 6-an-na-ka
Col. iii
WB P2 [mes-ki]-ag109-ga-[§e-er]
d
WB P2 [dumu] utu en-a[m]
WB P2 [1 u g a 1] - a m mu 32[4]no i - a6
98
This seems the best translation; g t i r = saldlu, "to take away as spoil" (§L, No.
362.20). Langdon's translation (OECT II 11), "with his weapon subdued" (followed by
Zimmern in ZDMG n.F. I l l 30, n. 16; by Barton, RISA, p. 349; and—with slight change-
by Giiterbock in ZA n.F. VIII 3: "dessen Waffe das elamische Land unterwarf"), does not
take into account the fact that - b i can refer only to a neuter or to a collective;
8 " t u k u l - b i must therefore be the weapons of Elam and cannot be the weapon of
En-me (n) -barage-si.
97-97 Thus apparently also P 6 . Su2 and—as is probable from the spacing—P3 omit. Since
l u g a l - a m is attested in both branches of the tradition (A branch: WB; B branch:
P5), it must be original and the omission in Sua and P3 secondary.
98
Thus also Su2, Pa, and P6. The latter has » * , which should probably be restored
OFff* (900).
99
Thus also Su2; P3 omits - k a . Since - k a is found in both branches of the tradition
(A branch: WB; B branch: Su2), it must be original and the omission in P 3 secondary.
Aka (written a5 - k a) is mentioned in the story of Gilgames' war on Kish (see p. 145, n. 20)
and in Chronicle Weidner (Giiterbock in ZA n.F. VIII 48, 1. 31', where it is spelled ma6 -
k a) . The name is frequent in the Fara texts (Deimel, Die Inschriften von Fara III [WVDOG
XLV (Leipzig, 1924)] 20*), where it is spelled a k , i.e., a k (a).
100 102
P8 and Su2 omit. - s i - is certain; see n. 94 above.
101 103
Su2: - g a - e - ; see n. 93 above. Su2 omits.
104
P 3 adds here a summary of the reigns of En-me(n)-barage-si and his son:
f 1,5125(?) m [u d u m u - d u m u ] 1,525(?) years the family
en-me-b[ara-ge-si] of En-me (n)-barage-si.
Since other texts both in the A branch (WB) and in the B branch (Su2) do not show this
total, we must consider it a secondary feature peculiar to Ps. Cf. the similar case below
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COLS, ii 37—iii 3 85
37 the weapons of the land of Elam,96

became king and reigned 900 years;
Aka,
40 son of En-me(n)-barage-si,
reigned 625 years.

23 kings
reigned its 24,510 years, 3 months,
and 3§ days.
45 Kish was smitten with weapons;
its kingship to E-Anna(k)108
was carried.
In E-Anna(k)
Col. iii
Mes-kiag-gasher,
son of Utu, became high priest
and king and reigned 324 years.
(p. 113, n. 255). Apparently the scribe of P 3 (or of one of its predecessors) had a special pre-
dilection for such "family" totals.
io5-io5 p3>8 $gg[ ] a i s o c a n be thus restored. Su2 has $$<>mmr m u ^ i t u
TTT U4 ft4-, which most likely derives from the figures of WB through a broken text:
m u - b i £g&R&mK« i t u ttr m M u< ]*-. Our scribe miscopied W as wr and—being
used to summaries of the form m u - b i x m u y i t u z 114—read as indicated by the
horizontal braces and "corrected" the text by inserting a m u after what he considered
the figure for the years. Scheil in RA XXXI 161 has already seen that the difference
between the two figures "s'explique naturellement par des lapsus de copiste."
106
We must assume that the older form b a - s i g , retained in both main branches
of the tradition (A branch: J; B branch: Li, Sui, and Su3+4), was the one used in the
original (see p. 131).
107-107 T h u s also P 2 ; Su2: 6 - a n - n a ki - § e .
108
As first pointed out by Poebel (PBS IV 1, p. 115), the phrase presupposes that only
the temple precinct E-Anna(k) existed at the time. The city Uruk was built under En-me(r)-
kar (iii 7-9).
109
P2: - i n - .
«° m*W,. Since the sum of the other reigns of the dynasty as given in WB is 324
years less than the total given for the dynasty as a whole, our figure should be restored as
mm) (324; cf. Zimmern in ZDMG n.F. I l l 31, n. 3). A comparison of WB and Su2 sug-
gests that Su2 had the same figures as WB for the first ten rulers (see n. 143 below) and
therefore also 324 years for Mes-kiag-gasher. P2 has 325. Since WB and Su2 seem to agree
on 324, this figure is represented both in the A branch (WB) and in the B branch (Su2) of
the tradition and may therefore be considered original.
oi.uchicago.edu
m
WB P2 4 [mes]-ki-dgU2-ga-[§e-er]
W B P2 5 ab-ba ba-an-tu
WB P2 bur-sag-§e ba113-enm
WB P2 en-me-kar11& dumu mes-ki-
&[g l l 6 -ga-§e-er] 1 1 7
WB P2 1 u g a 1 u n uki - g a Id u[nu gki]118
WB P2 m u - u n119 - d OL - a
WB P2 10 1 u g a 1 - a m
WB P2 Sui mu 420121 i-a B
111-ni This note is probably an early addition; see p. 143, n. 14.
112
P2: - i n - . '
113
P 2 adds fT, i.e., b a - V - e u • The plene writing serves to express hiatus.
114
a b - b a b a - a n - t u can hardly mean "penetrated to the sea" (Langdon, OECT
II 11) but must be "entered the sea,1' "went into the sea" (Zimmern in ZDMG n.F. I l l 31,
n. 2: "zog in das Meer hinein"; Smith, EHA, p. 33: "entered the sea"; Guterbock in
ZA n.F. VIII 3: "drang in das Meer ein"); and the contrast between t u , "to go in,"
and en , "to go out," suggests that we should translate the following J j u r - s a g - S e
b a - en as "he came out toward the mountains." The usual rendering, "he ascended the
mountains" (Langdon, loc. cit.: "went up unto the mountains"; Zimmern, loc. tit.: "stieg
auf das Gebirge hinauf"; Smith, he. cit.: "ascended the mountain"; Guterbock, loc. cit:
"stieg auf das Gebirge"), does not bring out this contrast clearly. The rendering "came
out" is also supported by the fact that the journey of Mes-kiag-gasher, the "son of the
sun-god," obviously reflects the daily journey of the sun. In the evening the sun goes
down into the sea in the west. During the night it travels underground, and in the morning
it comes out to the mountains in the east. Crossing over them, it then appears again to
the world.
118
P 2 : e n - m e - e r - r f t - k d r . En-me(r)-kar is mentioned outside the King List
(1) in the Lugal-banda epic (SEM, p. 1 and texts listed there), (2) in the epic of En-me(r)-
kar and Nigi of Lamkurru (SEM, pp. 1 f. and texts listed there. Nigi is mentioned in
No. 16 obv. iii 18: n i - g i e n - a m m e s - b i , "Nigi, the high priest, is its hero," and
ibid. 1. 23: n i - g i e n l a m - k u r - r u k i , "Nigi, the high priest of Lamkurru." That
the city name is to be read phonetically, l a m - k u r - r u , is proved by the variant writ-
ing n u - u m - k u r - r u in SEM, No. 19 obv. ii 18, for Numkurru is merely a phonetic
variant of Lamkurru: Numkurru>Lamkurru; see Poebel, GSG §64), (3) in the legend
of the king of Kutha (CT XIII, PL 41 obv. 2), and (4) in Chronicle Weidner (ZA n.F.
VIII 48, 1. 32', 49, 1. 1, and 51). (5) The text dealing with the seven apkallu'a (Ki 1904-
10-9, 87 [published by T. J. Meek in AJSL XXXV (1928/29) 138] obv. 5-6+K5119
[published by Gurney in JRAS, 1935, pp. 463-65] obv. 10-13; cf. Langdon, Babylonian
Penitential Psalms [OECT VI (Paris, 1927)] p. 32, and Guterbock in ZA n.F. VIII 9 f.)
mentions the apkallu of En-me(r)-kar: [. . - u ] g - g a l - g i m . This note is of consid-
erable interest and probably refers to the traditions contained in the epic of Lugal-banda
and that of En-me(r)-kar and Nigi. It should be read:
[. . - t i ] g - g a l - g i m n u n - m e d e n - m e - k ^ r d i n a n n a § a e - a n - n a - k e 4
a n - t a en-de
[urn] ap~kal MIN M Hl-tar ih-iu sam6e ana ki-rib a-a-ak-ki li-se-ri-da
[Footnote 115. continued on facing page]
oi.uchicago.edu
C R I T I C A L E D I T I O N O F T H E T E X T , C O L . iii 4-11 87
4 Mes-kiag-gasher
5 went into the sea
and came out (from it) to the mountains. 114
En-me(r)-kar, son of Mes-kiag-gasher,
king of Uruk, the one who built

Uruk,120
10 became king
and reigned 420 years;
". .uggalgim, apkallu of En-me(r)-kar, (him) who made Inanna(k) descend from heav-
en into E-Anna(k)."
The name En-me(r)-kar is written e n - m e - e r - k d r in the Lugal-banda epic and
in the Nigi epic, but appears as e n - m e - k & r in Chronicle Weidner, in CT XIII, PL
41 obv. 2, and in the apkallu text. E n - m e - k d r represents the older orthography, in
which amissible consonants were not expressed {cf. AS No. 6, pp. 17 f.), whereas e n -
m e - e r - k & r represents the younger, more explicit orthography. The name is thus
En-me(r)-kar. Whether the form given by P2, e n - m e - e r - r u - k & r , is intentional
or the wedge >-~ (r u) is merely a slip of the scribe is uncertain.
In post-Babylonian sources En-me(r)-kar is mentioned (1) as grandfather of Gilgames
by Aelian, De natura animalium xii 21, where the name Xtwjxopos—as first suggested by
von Gutschmid (see Zimmern in KAT\ p. 565, h. 3, and in ZDMG n.F. I l l 31, n. 4)—
should be emended to Efojxopos (cf. Sayce, quoted by Langdon, OECT II 12, n. 3). Aelian's
story probably derives ultimately from Berossus. (2) In the usual excerpts of Berossus En-
me(r)~kar appears in the forms Efojxotos (Polyhist. ap. Sync.) and Evek'sios, Eveokhos
(Polyhist. ap. Euseb.), i.e., *Einjxopos (von Gutschmid, quoted in KATZ, p. 565, n. 3;
Langdon, OECT II 9, n. 5; Zimmern in ZDMG n.F. I l l 30, n. 2, and 31, n. 4), which were
first identified with the name En-me(r)-kar by Schnabel (see Zimmern in ZDMG n.F. Ill
31, n. 4); cf. n. 39 above. (3) In the scholia of Theodore bar K6nl, as seen by Hugo Gress-
mann in Theologische Literaturzeitung XXX (1905) 586, the name §~©], which precedes
that of Gilgames (- ^ y V *; cf. n. 128 below) in the list of rulers given by that author
(see Martin Lewin, Die Scholien des Theodor bar K6ni zur Patriarchengeschichte (Genesis
XII-L) [Berlin, 1905] p. 25, variant from ms. P), probably represents *Efo?x°P<"> En-me(r)-
kar.
116
P2: - i n - .
117
P2 adds - k e 4 .
118
P2 adds - g a (?) , apparently dittography from the preceding u n u ki - g a . See
Poebel in OLZ XXVII (1924) 262, n. 1.
119
P 2 adds - d a - , which changes the meaning from "(who) built Uruk" to "under
whom Uruk was built." See Poebel, he. cit.
120 p 2 - "f^Q o n e uncier whom Uruk was built."
121
Thus also P2 and Sui. Berossus (according to Polyhist.) gives Evek'sios, Efyxoios
(emend to Ebqxopos) 2,400 years.
oi.uchicago.edu
d
WB P2 Sui 12 lugal-bd,n-da122 si p a
WB P2 Sui mu 1,200123 i - a8
d
WB P2 Sui d u m u - z i124 §U-PE§
WB P2 Sui 15 u r u k i - n i k u 6 - a k i 126
127
WB P2 Sui mu 100 I - a6
d d 128
WB P2 Sui g i 1 g a m e s ( Gi§-BiL-GA-MEs)
122
Thus also P2. Sui omits the determinative. Lugal-banda is the hero of the Sumerian
Lugal-banda epic (SRT, pp. 34 f.; SEM, p. 1; Faikenstein in OLZ XXXVI [1933] 301 f.;
Giiterbock in ZA n.F. VIII13, n. 1) and appears also in the legendary text PBS V, No. 20.
With the Akkadian myth of Zu and the theft of the tablets of destiny Lugal-banda has
nothing to do, nor does this myth form part of the Lugal-banda epic. Berossus seems to
mention Lugal-banda as the second of his postdiluvian kings and son of *Efajxopos (En-
me(r)-kar), for the badly corrupt Xwju<*<Âos (Polyhist. ap. Sync; ap. Euseb.: Chomasbelos,
Khomasbelos) very likely derives from an original Lugal-banda:
* A tu. «r A A & * N A o c-
That Xo>/ia<rj3i?\os should have developed out of a variant of the name Gilgames, *g a - m e § -
b i 1, as Zimmern suggests (ZDMG n.F. I l l 31, n. 4), is not probable, for such a variant
is not likely to be found. Sayee's derivation (in OECT II 9, n. 6) from g u l - l a
d
n i d a b a a n - n a - d a s i k i l (cf. n. 40 above) is also unconvincing.
123
Thus also P 2 and Sui. Berossus (Polyhist. ap. Euseb.) gives Xaâcr/fyXos ( = Lugal-
banda) 2,700 years.
124
As a terrestrial ruler dDumu-zi(d) appears also in the text PBS V, Nos. 20-21, which
deals with an Elamite attack on Babylonia in his time; see Poebel, PBS IV 1, p. 117. Cf.
also n. 22 above.
128
The meaning of STJ-PES is unknown. Perhaps it is to be identified with s u - k u 6 ,
"fisherman," as is generally assumed (cf. Poebel, PBS IV 1, p. 118; Langdon, OECT II 12;
Zimmern in ZDMG n.F. I l l 31, n. 6).
126
Various readings of the city name written gA-Aki or, with transposition of signs,
ki
A-gA have been proposed. Langdon (Babylonian Liturgies [Paris, 1913] p. 115, n. 2) sug-
gested a reading b a - b u r . Most other scholars read § u b a r i (see Poebel, PBS IV 1,
p. 121; Zimmern in ZA n.F. V [1930] 255, n. 1; and Ungnad, Subartu, pp. 28-30 and the
literature quoted there). Of importance for the reading are only two passages: (1) In the
incantation CT XVI (1903) PL 6 v 239-40 A-QAki of the Sumerian line is rendered hu-ba-ri
in the Akkadian translation. (2) In the Tammuz hymn published by K. D. Macmillan in
Beitrage zur Assyriologie und semitischen Sprachwissenschaft V (1906) 674-75 obv. 25-26
A-gAki of the Sumerian line is rendered ku-u%-a-ra in the Akkadian translation, and the
duplicate (G. Reisner, Sumerisch-babylonische Hymnen [Berlin, 1896] No. 80:9) also has
fkutdyua-a-ra. Of these forms, subari and kv?ara, only one can be correct. A comparison
with the old writing itself, gA-Aki/A-gAki, shows clearly that kuPara should be preferred;
for gA-Aki can be read k Ue -a k i , and, since the amissible consonants were not expressed in
older Sumerian orthography, this is a perfectly correct spelling of a name pronounced
Ku3a(r) (cf. my remarks on non-expression of final -k in AS No. 6, pp. 17 f., also n. 115 above
on En-me-kar/En-me(r)-kar) abbreviated from older KuDar(a). The writing su-ba-ri in CT
XVI, PL 6, must be a corruption offcw(!)-6a-n,which would represent a variant pronuncia-
tion (kv?ara>*kuwara>kubara). The signs su and ku are so similar that a misreading can
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. iii 12-17 89
12 divine Lugal-banda, a shepherd,

divine Dumu-zi(d), a . . . ,125—
15 his city (was) KuDa(ra)—
reigned 100 years;
divine Gilgames—
easily have happened. On the situation of the city and its relations with Eridu(g) see n. 5
above. The connection of Dumu-zi(d) with KiPaCra) has been dealt with in detail by
Poebel, PBS IV 1, p. 121.
127
Thus also P2 and, as it seems, Sui, where W4 can be restored !& (100).
128
The hero of the Gilgames epic (see the literature quoted in Albert Schott, Das Gilr
gamesch-Epos [Leipzig, (1934)] pp. 5-10). A separate tale, the last part of which was incorpo-
rated in the Gilgames epic as the twelfth tablet, has been edited (first half only) by Kramer
in Gilgamesh and the ffuluppu-Tree (AS No. 10 [1938]). See the texts there quoted and add
Langdon, BE XXXI, No. 35, and Radau in Hilprecht Anniversary Volume (Leipzig, 1909)
Pis. 13-14, No. 11, which continue the story and give the beginning of Enkidu's account
of Hades. Another epic text deals with Gilgames' war with Kish through which he liberated
Uruk (see p. 145, n. 20). The "History of Ebmal" (PBS V, Nos. 6-7 +PBSXIII, No. 48; see
Poebel, PBS IV 1, pp. 143-47, and Giiterbock in ZA n.F. VIII 7. On the reading Ebmal
instead of Tummal see Poebel in OLZ XXVII [1924] 263, n. 4, and cf. Landsberger in OLZ
XXXIV [1931] 129) mentions him as builder of the Giparu (written g u (gi) - b u r - r a) of
Enlil in Nippur. A late ruler of Uruk, Anam, states that Gilgames built the city wall of Uruk
(SAK, p. 222, No. 2 b). As protecting genius Gilgames is mentioned by Utu-hegal (RA IX
[1912] 115 iii 1-3). Ur-Nammu(k) is his "big brother" (TRS I, No. 12:112). He is Shulgi's
"beloved brother" (PBS X 2, No. 6 rev. i 16) or "brother and friend" (Chiera, Sumerian
Texts of Varied Contents [01P XVI (1934)] No. 51 rev. 35). A prayer to him is given
in Paul Haupt, Das babylonische Nimrodepos (Leipzig, 1884) No. 53. A collection of Gil-
games omens has been treated by Zimmern in ZA XXIV (1910) 166-71. They all seem to
derive from the epic (Giiterbock in ZA n.F. VIII 9). He is mentioned in MaqM, Tablet I,
and an image of him in R..F. Harper, Assyrian and Babylonian Letters I (London and Chi-
cago, 1892) No. 56 rev. 5. The name of a certain drug contains his name (KAR, No. 186
rev. 10; see Thompson, The Epic of Gilgamish [1930] p. 10). Aelian (De natura animalium
xii 21) relates a curious story of how En-me(r)-kar (Xevtjxopot), having been told that his
daughter would bear a son who would seize the kingdom from him, locked the girl in a
tower. But the girl bore a son by an obscure man, and when the guards threw the child
down from the tower an eagle carried him to a garden where the gardener found him and
brought him up. This child was Gilgames. The list of ancient kings contained in the scholia
of Theodore bar K6nt (Lewin, Die Scholien des Theodor bar K6n%) mentions Gilgames
(s^pnVi^Sn,,,, vars. uocuJUao^. and a g a ^ , ^ , , [pp. 2 and 25], reflecting an original
, w ^ V *) as the last of the ten kings from Peleg to Abraham and contemporaneous with
the latter (p. 2,11. 8-9).
The orthography of the name changes considerably in the various periods of Babylonian
history. The oldest occurrences of the name known to us date from the time of Lugal-anda
and give the writing dGis-BiL-GiN-MES (Allotte de la Fuye in RA VI [1904-7] 124). From
Sargonic times down to the period of Isin and Larsa the standard form becomes dGis-BiL-
GA-MES (Sargonic period: Ur-dGis-BiL-GA-MES, UEII316, U 11418; Utu-hegal: dGis-rBiL1(?)-
GA-MES, RA IX [1912] 113 iii 1). This writing also occurs in the inscription of Anam
oi.uchicago.edu
WB P 2 Sui 18 a b - b a 1 2 9 - n i 1 £ l 130 -1 a
WB P 2 Sui en k u l - a b - b a - k e4132
WB P 2 Sui 20 m u 126133 i - a6
134 d 134
WB Sui u r - nun -gal (var. u r -1 u g a 1)
d
WB Sui dumu g i l g a m e s (dGis-BiL-GA-MEs)
WB mu 30 i - a5
WB u~dtil-kalam-ma
WB 25 d u m u u r -135 d n u n - g a l - k e4135 (or
ur-lugal-ke4)
WB mu 15 i - a5
136
WB l a - b a - a <b? (old lacuna?)>m 136
WB mu 9 i - a5
WB e n - n u n - d i r a(!) 137 - a n - n a
WB 30 m u 8 i - as
(SAK, p. 222, No. 2 6) and in the Sumerian fragments of the Gilgames epic (e.g. SEM,
No. 22 obv. 8, No. 26 rev. iii 11; RA XXX [1933] 128-29,11. 8 and 41) which belong to the
Isin-Larsa period. Other examples from this period are frequent. A writing dGis-BiL-GA-MES
is found in the King List and in the Sumerian version of the epic, PBS X 2, No. 5, and Zim-
mern, Sumerische Kultlieder aus altbabylonischer Zeit, 2. Reihe (1913) No. 196, from ap-
proximately the period of the 1st dynasty of Babylon. The Akkadian versions from Old
Babylonian times have the abbreviated form dGis (Meissner, Ein altbabylonisches Fragment
des Gilgamosepos [MVAG VII 1 (1902)]; M. Jastrow and A. T. Clay, An Old Babylonian
Version of the Gilgamesh Epic ["Yale Oriental Series. Researches" IV 3 (New Haven, Conn.,
1920)]). In a personal name in a document from Elam (Scheil, MDP XXII, No. 41:1:
puzur-Agis-ga-mas) the name is written dGis-GA-MAs; later forms are dGi§-GiM-MAS (Hittite
texts: see J. Friedrich in ZA n.F. V 32 f.), dGAL-GA-Mi-su-UL (Hurrian texts: F. Hrozny',
Keilschrifttexte aus Boghazkoi, Heft 6 [WVDOG XXXVI 2 (1921)] No. 33 i 8; Ungnad in
ZA n.F. I [1924] 135); and in Assyrian times we find dGis-GfN-MAS (Ninevite version of
epic), Gis-GfN-MAS-si(?) (CT XII [1901] PI. 50, K 4359 obv. 17; cf. Friedrich in ZA n.F.
V 33, n. 2), GI-IL-GA-ME-ES (T. G. Pinches in Babylonian and Oriental Record IV [1890] 264),
KAL~GA-IMIN (syllabary, CT XVIII [1904] PL 30 iv 6). Post-Assyrian forms are TtXyafjios
(Aelian, De natura animalium xii 21) and - -M^Vn V * (Theodore bar K6nf; Lewin, loc.
cit.). Further literature may be found in Poebel, PBS IV 1, p. 123; Friedrich in ZA n.F.
V 32 f.; Thompson, The Epic of Gilgamish (1930) pp. 8 f.
129
Thus also P2. Sui omits.
130
Thus in all probability P2 also. The photograph in PBS V, PI. XC, seems to have
W(?) — rl i I1. Poebel's copy gives ffJ> = A-r. A The conclusions drawn by W. Fortsch
in OLZ XVIII (1915) 367-69, based on this reading, should therefore be abandoned.
131
Thus with R. Campbell Thompson, who in this 1 i 1 -1 k sees Sumerian 1 i 1 -1 d,
Akkadian Ufa, "the demon equivalent to a male vampire. There are four demons of this
class—the idlu lilt, the ardat Mi, the UliX, and the lilitu. The ardat lili is well known as the
female vampire or succuba who visits men by night and bears him[!] ghostly children: the
idlu lilt must be her male counterpart who can visit women and beget offspring by them,
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. iii 18-30 91
18 his father (was) a lilM-demon1*

a high priest of Kullab,
Ur-Nungal(ak) (var. Ur-lugal(ak)),
son of divine Gilgames,
reigned 30 years;
Utul-kalamma(k),
25 son of Ur-Nungal(ak) (or Ur-lugal(ak)),
reigned 15 years;
Laba. . . JR 136
reigned 9 years;
En-nun-dara-Anna (k)
30 reigned 8 years;
just as demigods are created" (The Epic of Gilgamish [1930] p. 9). Thompson rightly com-
pares the "obscure man" in the story of Aelian (see n. 128 above) and the statement in the
epic that Gilgames was 2/3 god and 1/3 mortal, which seems to show that neither of his
parents was fully mortal. Langdon (OECT II 12, n. 3) translates 1 i 1 -1 & as "fool, im-
becile" and considers it possible "that lil-la is really a title of Tammuz who in the Louvre
hymn, RA. 19, 175-185 is called mu-lu-lil, 'the fool god/ and that Tammuz is really the
father of Gilgamis here." This combination seems doubtful. Gilgames is generally consid-
ered son of the goddess Ninsun (thus in the Utu-hegal inscription, RA IX 113, and in the
Ninevite version of the epic; see Poebel, PBS IV 1, pp. 124 f., and in OLZ XVII [1914]
4-6), who was the wife of Lugal-banda (Deimel, Pantheon babylonicum [Romae, 1914]
No. 2701).
132
A quarter of Uruk (cf. Thureau-Dangin in RA IX 119), originally probably a separate
city. It seems likely that it was the quarter around the main temple, E-Anna(k) (see Zim-
mern in ZA n.F. V 258, where E-Anna(k) is styled e m e - g a l k u 11 a b k i - a , "Tem-
pel, mit den grossen Ordnungen, in Kullab").
133
Thus also P2. Sui has frffi;but ] is in all probability merely due to the stylus having
slipped, so that we can read here also IT iff (126).
134-134 g U l . . l u g a 1. In the "History of Ebmal" also (PBS V, No. 6 obv. 5) the name
appears as [u] r -1 u g a 1. The variant is clearly due to mishearing when the text was
dictated to a copyist. A similar mistake, which must likewise be due to mishearing in dic-
tation, is mentioned in n. 146 below. We cannot say which form is original.
las-las gee n . 1 3 4.
136-136 Langdon's rendering of the third sign as § e (%k) and his spacing of the signs in
his copy are not correct. The photograph shows J ^ £ T ^ JB>. The third sign looks like
an incomplete a fc ; and, since the large blank after it suggests a lacuna in the scribe's
original, our line may render &gKt
137
According to the photograph the sign is <$5E=*, i.e., d a r a (cf. Charles Fossey,
Manuel d'assyriokgie II 1 [Paris, 1926] No. 7919).
oi.uchicago.edu
92 THE SUMEBIAN KINO LIST
WB 31 ^ M E S H ? ) - ^ si mug 1 3 8 m u 36 i - a 5
139
WB Su2 m e-Um-an-na*39
WB Su2 mu 6 140 i-a6
WB Su2 l u g a l - k i - t u n 3 ( ? ) 1 4 1 m u 36 142 i - a 5
WB Su2 35 12 l u g a l
WB Su2 mu-bi 2,310 143 fb-a5
kl gi§
WB Su2 unug t ukul ba-an-sig
WB P2 Su2 nam-lugal-bi u r fki - § 6
WB P2 b a-td m
WB P2 Su2 u r iki - m a
WB P2 Su2 40 m e s - a n - n - 6 - p & - d a 1 4 4
WB P2 l u g a l - & m m u 14580 (emend to 80-x?)
i-a*
41a (a-an-n6-pd-da
las-lasLangdon reads s u & u £ - k e * - d 6 , but a royal name meaning "May he de-
stroy the foundation" is not very likely. The photograph seems to have t£f for the
first sign, which looks more like FMEST or 'DTJP1 than like s u b u §. We therefore read
provisionally 'MES^J-QE s i m u g , "MES(?)-£E, a smith."
1*9-139 gU2 has here *+< *%, i.e., t i - k u . It is not impossible that this has developed
out of an original m e - l & m - a n - n a through a badly damaged version [m e -1 a m -]
****?, in which ** (a n) and ^ (the front part of n a) were read together as one sign,
HTK (t i), and separated from ** (the remainder of n a), which was read as k u . That the
original of Su2 was indeed seriously damaged at this place can be seen from the fact that
its formula x m u 1 - a5 changes to m u x i - a6 just after the reign of Ti-Ktr, which can
only mean that here the copyist had to use a different source to fill in a lacuna in his original.
140
Su2 has 900! See n. 143 below.
141
Photograph: &£=r?
142
Su2: W (7), read as 420! See n. 143.
143
This total agrees with the sum of the single reigns if we restore the broken figure
for Mes-kiag-gasher as 324 (see n. 110 above). Su2 has the total 3,588 years. This is cor-
rect if we restore the first ten reigns, now missing in Su2, according to the figures given in
WB and add the very high reigns of the last two rulers which are preserved in Su2: [324 +
420+l,200-flOO+126+30-fl54-9+8+36]+900+420=3,588. There can, of course, be
no doubt that WB's reasonable figures for these two reigns, 6 and 36, which agree with the
historical character of the other reigns in the last half of this dynasty, represent a better
form of the text than the 900 and 420 years given in Su2.
144
This writing is found also in Mes-Anne-pada's seal inscription (UE II, PI. 191,
U 13607), in the seal inscription of his wife (UET I, No. 268), in the inscription of his son
A-Anne-pada (UE I, PL XL, T[ell] 0[beid] 160), and in the list of names in PBS XI 1,
No. 25 obv. In the late text BM 56488 rev. iv 11-16': Utu Id dmes-an-n6-p&-da i-pu-su
naran-na la-qi-it zi-i-ri u$-tal-pitt "The house which the divine Mes-Anne-pada had built
did Nanna, the 'seed-gatherer/ destroy(?r (Meissner in ZA VII [1892] 29; G. Dossin in
RA XXII [1925] 115-17; Giiterbock in ZA n.F. VIII 8, who rightly suggests that the pic-
ture of birds picking up seed in a field underlies the expression zira laqatu), the name is
written with the determinative for god.
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. iii 31-41a 93
31 MES(?)-9E3, a smith, reigned 36 years;

Melam-Anna(k)
reigned 6 years;
Lugal-ki-tun(?) reigned 36 years.
35 12 kings
Uruk was smitten with weapons;
its kingship to Ur
was carried.
InUr
40 Mes-Anne-pada
became king and reigned 80 (emend
to 80-a;?)145 years;
41a <A-Anne-pada,
145-H5 xhis surprisingly high figure probably represents—as pointed out by Gadd in
UE I 128—a conflation of two reigns, that of Mes-Anne-pada himself and that of his son
A-Anne-pada. We may assume that an original text
mes-an-n£-pa-da
l u g a l - a m m u 80—x 1 - as
[a-an-ne-pa-da
dumu] m e s - a n - n 6 - p a - d a
m u x 1 - a«
broken as here indicated became
mes-an-n£-pa-da
l u g a l - a m m u 80 —x l - a 6
mes-an-n£- pa-da
m u x i - as
in tlie next copy and that a still later copyist, who did not realize that the blank space
stood for a lacuna, added up the two reigns which both seemed to be assigned to Mes-Anne-
pada.
A-Anne-pada is known from his own inscriptions (UE 1126 f.), where his name is written
a - a n - n 6 - p a - d a and where he designates himself as son of Mes-Anne-pada. Accord-
ing to the "History of Ebmal" (PBS V,Nos. 6-7+PBS XIII, No. 48), he built the park
(g i r ii2 - m a b) of Enlil in Nippur. His name is there written [a n - n] a - n 6 (PBS V,
No. 6 obv. 9), a n - n a - n 6 (ibid. 1. 10), and n a - a n - n e* (PBS XIII, No. 48 ii 2). (It
has been suggested that in BM 56488 rev. iv 13 quoted above [n. 144] his name appears
as na-an-na. More likely, however, na-an-na there stands for Nanna, the god of Ur; for
laqit zlri is an epithet which it is natural to apply to a deity, not to a mortal.) It is not im-
probable that these short forms of the name derive ultimately from the King List, more
exactly from the point in the tradition of that list when the lacuna mentioned above had
begun to form: a - a n - n e - [ p a - d a ] > a - a n - n 6 , i.e., a 3 a n n e , which by as-
similation of D to the following n would become a n a n n e . On the forms a n - n a - n e *
and n a - a n - n e * see Gadd in Studia Orientalia I (1925) 25f. and UE I 130. Gadd con-
siders them "slightly different echoes of the ill-remembered A-anni-(padda)." Cf. also Poebel
in OLZ XXVII (1924) 263.
oi.uchicago.edu
416 dumu mes-an-ne-pa-da

41c mu x i - a5>145
WB P 2 m e s - k i - a g - 1 4 6 d n a n n a146 (emend to
-nun-na)
147
WB P 2 d umu m e s - a n - n ^ - p ^ - d a147
148 148
WB 1u ga1 -am
WB P 2 U 45 mu 36149 i - a5
Col. iv
P2 Li [e - 1 u - 1 u mu 25 i - a5]
P2 Li [b a -1 u -1 u mu 36 i - a5]
P2 U [4 l u g a l ] 1 4 9 *
P2 U [m u - b i 177150 i b - a5]
ki
P2 Li 5 [u r i ^Hukul ba-an-sig]
P2 [n a m - 1 u g a 1 -hi a - w a - a n ki -se]
P2 [ba-tum]
P2 [a - w a - a nki-na]
t.... lugal-am]
[mu .. i-a*>]
[....]
WB m uu [.
[••. i-a5]
WB ku-ul-[ ]
WB m uu 36 [i-a5]
WTB (P2) 3 [lugal]
WB(P 2 ) m uu--bbi i 3 rr5 6 ii52
3 56 1152
[lb-ah]
WB a-w w aa - aa nnkiki gi
gi'H[tukul
'Htukul ba- a] n - s i g
H6-146 p 2 : - n u n - n a . As shown by an inscription of the ruler himself, U 11675
(Burrows in UE II 321, n. 10): [ r a e s - k | i - d [ g ] - n u n l u g a l u r i k i , P 2 has pre-
served the original text: m e s - k i - a g - n u n - n a . The form of WB, m e s - k i - d g -
d
n a n n a , must be due to mishearing in dictation (for a similar case of mishearing see
n. 134 above). The "History of Ebmal" also (PBS V, No. 6 obv. 10; PBS XIII, No.
48 ii 1) has the erroneous m e s - k i - a g - d n a n n a , whereas the Nippur list of proper
names (Chiera, PBS XI 1, No. 25 obv.) gives both forms, m e s - k i - a g - d[n a n n a ]
and raes-ki-dg-nun-n[a]. For a different solution of the problem see Poebel in
OLZ XXVII 254. We consider the one here offered (cf. also n. 147) more probable.
147-H7 T n u s a i s o p2> T n e "History of Ebmal" (PBS V, No. 6 obv. 10; PBS XIII, No.
48 ii 1-2) has this ruler as son of A-Anne-pada ( a n - n a - n e , n a - a n - n 6 ) . It is prob-
able that the King List originally had a - a n - n e - p a - d a here and got m e s - a n -
n e - p a - d a b y dittography from the lines above. Cf. n. 148.
148-148 p 2 omits, which is obviously correct. The l u g a l - a m of WB must be due to
dittography from m e s - a n - n ^ - p a - d a l u g a l - a m above.
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COLS, iii 416—iv 17 95
416 son of Mes-Anne-pada,

41c reigned x years;)
Mes-kiag-Nanna(k) (emend to Mes-
kiag-mmna(k)),
son of Mes-Anne-pada,
became king and
Col. iv
Elulu reigned 25 years;
Balulu reigned 36 years.
4 kings
reigned its 177 years.
5 Ur was smitten with weapons;
its kingship to Awan151
was carried.
In Awan
. . . . became king
10 and reigned . . years;
reigned . . years;
Kul
reigned 36 years.
15 3 kings
Awan was smitten with weapons;
149
P 2 : 30. Thus, in all probability, Li also, for it has the same total for the dynasty as
has P2. We must consider the text of WB original; for while 4W (36) can easily develop
into -« (30) through a slightly broken text: « « ] , the opposite development would be dif-
ficult to account for.
149a
If the argument of n. 145 is accepted, this line must originally have read 5, not 4,
"o p 2 : ft** ( 1 7 1 ). L l . -mr^ ie f [mj ( m ) > T h i g t o t a l a g r e e g w i t h t h c g u m o f t h e g i n g I e
reigns given in P2. Since Mes-kiag-nunna(k) has there a reign of 30 years whereas WB
gives him 36 years, we must assume that W^B's total was six units higher, i.e., 177 years.
151
Awan was situated in Elam not far from Susa. See Unger and Ebeling in RLA I
324 and the literature quoted there.
152
The figure of WB is damaged. The traces left show in the photograph as Wi&m, which
we read as m^Wh (&5&). Langdon's rendering of the traces in OECT II, PL II, does
not agree with what can be seen in the photograph. P 2 gives in the final summaries (xi
18-21) 356 years for Awan.
oi.uchicago.edu
WB 18 nam-lugal-bi
WB k i § i ki - § 6 [ba-tum]
WB 20 k i § i ki
su8-[ ]
WB lugal-&m
WB mu 201[+x 1 5 3 i-a*]
WB da-da-sig mu [. . i-a5]
WB 154
m 4 - m a - g a l - rla1155 (emend to m k-
g a l - g a l - ' l a 1 ) 1 5 4 [md4afp]{l)m
WB 25 m u 360[+x(?) 1 5 7 i-a5]
m
WB k a - a l - b u -[um ]
WB d u m u m a - g a l - g a l - rl a1 [?]
WB mu 195 l-[aj
WB Sis- e m u 360 i-'a. 1
WB 30 G i + s u B - n u n - n a m u 180 i-a 5
WB 159
i - e n b i - n i ( ? ) - i b ( ? ) -eS 4 (?) r t £ rU59
WB mu 290(?)160 1-as
WB U 161
1 u g a 1 - mu 161
mu 360162 !-a5
WB U 8163 l u g a l
WB U 35 m u - b i 3,195164 (var. 3,792) i[b-a5]
WB U k i §i ki
«i§t u k u l ba-an-sig
WB Li nam-lugal-bi
WB U b a - m a - z i165 k i - § & ba-tiim
183
HHM, i.e., 201+z (x less than 5).
154-154xhe name appears a s m d - g a l - g a l - ' l a 1 below in 1. 27. That this is the cor-
rect form is shown by an omen which refers to this king (A. Boissier, Choix de textes rela-
tifs d la divination assyro-babylonienne I [Geneve, 1905] 47, Nos. 17-18): md(see Boissier,
n. \W)-gal-gal ^malahfiu sd kis-su-tam $pusu$ ina libbi dli-su Jiu-id-la-liX iJdu-uW-su} "Ma-
galgal, the skipper, who exercised emperorship; in the midst of his city a . . . .-snake (hul-
Ulu is a synonym of nis qaqqari [AH, pp. 440 f.], which according to the Gilgames epic,
Tablet X I 296, seems to be a [general or specific?] term for a snake) killed him." The name
m&-gal-gal is obviously identical with m a - g a l - g a l or l u - m d - g a l - g a l
(SL, No. 122.59), a term for a special kind of sailor or skipper.
166
Photograph has #&, i.e., rl a 1 .
168
Since Magalgal in the omen quoted above (n. 154) is designated as m A -1 afr4(wa-
lafyfyu), "skipper," we should probably restore this word in the lacuna after his name.
167
m s u - s i [ ] = 360-f-s (x less than 60).
158
This restoration was suggested by Langdon (OECT II 14).
159-159 The line gives no sense as it stands and is clearly corrupt. We have assumed that
it represents an effort on the part of the scribe to render as exactly as he could a damaged
original which he did not understand: l" e n b ini"ib - e h -1 & r . See the detailed discus-
sion on p. 169 and cf. the parallel cases in cols, ii 28 (n. 85 above) and vii 31 (n. 291
below), where an original gloss has similarly worked down into the line.
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COL. iv 18-38 97
18 its kingship
to Kish was carried.
20 In Kish Su
became king
and reigned 201+x years;
Dadasig reigned . . years;
Mamagalla (emend to Magalgalla),
a.skipper(?),156
25 reigned 360+x(?) years;
Kalbum,
son of Magalgalla,
reigned 195 years;
Sis-e reigned 360 years;
30 GA+suB-nun-na reigned 180 years;
Enbi-Eshtar(?)
reigned 290(?) years;
JfUgalmu reigned 360 years.
8 kings
35 reigned its 3,195 (var. 3,792) years.
Kish was smitten with weapons;
its kingship
to Hamazi166 was carried.
160
The figure, fci* as given in OECT II, PL II, is inconsistent with the Sumerian
numerical system and cannot be correct. In the photograph the figure shows as ?F1£,
which makes quite clear what has happened. The scribe by accident wrote the third •<-
wedge of the upper row so close to the second that the latter was practically blotted out,
and he therefore added a new, clear wedge. The figure is therefore meant to represent 290.
i6i-i6i Langdon (OECT II 6) translates 1 u g a 1 - m u as "a king by name" and sug-
gests that it is "a substitute for a name which our late compiler did not know." L u g a l -
m u could indeed be the remnant of a copyist's note, l u g a l m u n u - t u k , "a king
without name" (cf. vii 28, to which Langdon also refers); but, since l u g a l - m u is a
well attested Sumerian name (see e.g. the instances listed by Schneider in Orientalia No. 23
[1927] p. 156, No. 2165, and cf. also the following names which begin with l u g a l - m u -) ,
it seems simpler to assume that there actually was a king of Kish by that name.
162
Thus apparently Li also, where ^ can be restored as W.
163
Thus apparently Li also, where Wf can be restored as ffff.
164
Li: 3,792. It is impossible to say which is the more correct.
165
Li: - z i - .
166
The city is mentioned in the inscription of Utuk (SAK, p. 160), who styles himself
"defeater of Hamazi" (afN-Sfe b a - m a - z iki)« People from Hamazi are frequently men-
tioned in texts of the Agade period from Nuzi (T. J. Meek, Excavations at Nuzi . . . . III.
Old Akkadian, Sumerian, and Cappadocian Texts from Nuzi ["Harvard Semitic Series" X
oi.uchicago.edu
WB Li 39 l ) a - m a - z im ba-ta-ni-i§
169
39a (%oI4m>
WB Li 40 m u 6§u-§i170 (emend to 6?) m i-a 5 1 7 1
(probably old lacuna here)172
WB(P2) 1 (emend to 1+x?)172 1 u g a 1
WB(P2) m u - b i 6 § u - § i m (emend to 6+x?)
fbm-a5
WB fea-ma-ziki giS
t ukul ba-an-sig
WB Su; nam-lugal-bi unugki-§e b a -
t tim
(Cambridge, Mass., 1935)] Nos. 143:15, 153 iii 25, 154 ii 10, and 155 v 8); the name is in
all cases written fea-ma-ziki. During the period of the 3d dynasty of Ur a certain
Ur-Ishkur(ak) is known as ishakku of Hamazi (Langdon, Tablets from the Archives of Dre-
hem [Paris, 1911] No. 53). The name of the city is in this period most often written h a -
m a - z i k i (Langdon, he. cit; H. de Genouillac, La trouvaille de Drehem [Paris, 1911] No.
69). As fea-m a - z i k i (var. \i a - a m -z iki) it also occurs in the inscription of Arad-
Nanna(k), s u k k a l - m a ^ of Shu-Sin of Ur, who styles himself "ishakku of Hamazi
and Ganhar." The older writing l j a - m a - z i k i has not completely disappeared, how-
ever, for an unpublished text quoted by Forrer in RLA I 231, which seems to date from
shortly after the 3d dynasty of Ur, has it, and a copy of a historical (?) inscription from
the Isin-Larsa period (Chiera, Sumerian Texts of Varied Contents [OIP XVI (1934)] No. 98
obv. 6) also refers to the city as fca-ma-ziki ( m a - d a ^ a - m a - z [i ki ]) . Hamazi
seems to have lost all importance toward the end of the Isin-Larsa period, for as far as I
know it is never referred to in later texts. Since it is mentioned by Arad-Nanna(k) in
close connection with Ganhar and since people from Hamazi appear in texts from Nuzi,
it should without doubt be located in the mountainous region east of Kirkuk, near modern
Sulaimaniyyah.
167
Li: - z i k i - a .
168
Probably abbreviation of a longer name such as *tiata,nis-qabi> "He (a god) prom-
ises to protect" (on -i§ with infinitive see W. von Soden in ZA n.F. VII [1933] 105 fL).
This ruler is mentioned—somewhat unexpectedly—in the famous list of gods AN : dan~nu-
um (cf. Hommel in ZA n.F. II [1925] 82). The section in question reads:
18 dl u m - m a su Lumma, pronounced as
written;
19 % a - 1 a - n i - i [h] [§]u m i n u t u k 6 - k u r - r a - k e 4 Hatanish, pronounced as
written; both manes of
E-kur.
20 de n -111 -1 a - z [i] nu-banda d-kur-ra-ke4 Enlilla(k)-zi(d), laputtu
of E-kur.
21 d u r - d s i n - n a n i m g i r e § - b a r - r a- *kef Ur-Sfna(k),nagiruof Esh-
barra.
Text: CT XXIV (1908) PL 6:18-21. Variants: (1) CT XXV, PL 28, Bu 89-4-26, 77:3,
d
l u m - b a / m a ; (2) ibid. 1. 5 adds - a m after m i n ; (3) in CT XXIV, PL 22, the
explanatory note beginning with m i n runs from 1. 116, which deals with dLumma, down
to 1.117, which deals with d Hatanish. The prototype of AN: Aan-nu-um, AO 5376, published
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. iv 39-44 99
39 In Hamazi Hatanish168
39a (became king and)
40 reigned 6X60 (emend to 6?) years;
(probably old lacuna here)
1 king (emend to 1+x kings?)
reigned its 6X60 (emend to 6+x?) years.
Hamazi was smitten with weapons;

its kingship to Uruk was carried.
by de Genouillac in RA XX (1923) 98 f. and TRS I, Pis. XXV-XXXI, which seems to date

back to the Isin period, also contains our section; obv. ii 4-7 (in RA = 11. 53-56 in TRS)
reads: d l u m - m a d ^ a - t a - n i - i § d e n - l i l - l £ - z i d u r - d s l n .
How Hatanish, a conqueror from Hamazi, came to be included in the official Sumerian
pantheon is fairly clear from the list itself, which styles d Hatanish a herds, literally "ghost"
(u t u k), of E-kur. That the ghosts, the manes, of Hatanish and of Lumma (better known
under his other name, E-Anna(k)-tum; see Poebel, PBS IV 1, p. 166) were worshiped in
the Enlil temple in Nippur can only mean that these kings had erected statues of them-
selves there, providing for offerings to the statues and thereby to their manes after they
died. The passage is thus of some historical importance, since this would seem to imply
that Hatanish actually possessed Nippur. On the following two deities, Enlilla(k)-zi(d)
and Ur-Stna(k), who also seem to be deified human beings, see Landsberger, Der kidtische
Kalender, p. 28, n. 13. Concerning cult of the manes of deceased kings compare the lists of
offerings to statues of Ur-Nanshe(k), En-temena, Lugal-anda, etc. published by Deimel in
Orientalia No. 28 (1928) pp. 25 ff. ITT IV, No. 7310 rev. i 26-28, mentions offerings to a
statue of Gudea. Cf. also Deimel, "Die Listen liber den Ahnenkult aus der Zeit Lugalandas
und Urukaginas," Orientalia No. 2 (1920) pp. 32-51.
169
Li adds [1 u g a 1] - a m , "became king," which is obviously correct. The omission
of this phrase in WB must be due to accident.
170
Thus according to the photograph; since § u - § i is in small script, which shows
it to be a gloss, the original probably read 6. The figure is not preserved in Li. Cf.
n. 173.
171
Thus according to the photograph; the additional i b in Langdon's copy is
erroneous.
"2 Seen. 174.
178
Thus according to the photograph; since s u - § i is written in small script, which
shows it to be a gloss (cf. n. 170), the original probably had 6. P* gives in its final sum-
mary one king and W (7 or 420) years for Hamazi. We have no opinion as to whether the
W of WB or the W of P2 is original.
174
Thus according to the photograph; the i - of Langdon's copy is erroneous. The
collective i b - as, remnant of an earlier form of this total operating with more than one
king, indicates that part of the Hamazi dynasty has been lost in a lacuna sometime in
the course of tradition. Cf. pp. 25 and 50 f. for similar cases in Sui and Sua+i.
oi.uchicago.edu
WB Su3 45 u n u k i - g a m en-§akan176-§a4- an-

n a
WB Sua 1uga1-&m mu 1 § u - & i177 i - as
178
WB Su3 nam-lugal-bi mu 2 &u -§i i - as
WB m u 7 179 i - a5178
(Emend 11. 47-48 to read
47a lugal-ki-ni-se-duj-dui
476 mu x+2 i-a 5
47c lugal-kisal-si
47d dumu lugal-ki-ni-Se-dwj-du7
48a mu x + 7 i-a5
(Pi) 486 3 lugal
48c mu-bi .. £6-a5)
Col. v
m
Sui [unugki gi
Hukul ba-an-sig]
Sui [nam-lugal-bi urih-sd b a - 1 u m]
SU! [u r iki-ma lugal-ki-ni-s e» " d u 7 - d u 7 ]
Sui [lugal- a m mu i-a5]
Sui 5 [lugal- k i s a 1 - s i]
SUi [dumu lu g a 1 -ki-ni-se-dty-d U7]
Sui [mu i-a5]
Sui [....-gi4(?)]
Sui [m u i-a5]
Sui 10 [ k a ( ? ) - k u ]
WB Sui [dumu - g i4] - r k e^
WB Sui [mu i] - ra,1180
181
WB Sui [4] lugal
WB mu-bi [116]181 i b - as
15 [uriki Qi
Hukul ba-an-sig]
Lx [nam-lugal-bi a d a b a k i - § 6 b a - 1 li m]
171 SU3 + 4^ -SLi
176
On this reading of the sign and on the identification of this ruler with En-shakush-An-
na(k) see p. 171.
177
The figure is not preserved in Sus+«.
178-178The text of WB is badly corrupted here: n a m - l u g a l - b i belongs to the
formula for change of dynasty; m u 2 § u - § i (120) 1 - a& and m u 7 1 - as are from formu-
las for single reigns. Apparently the scribe has given a few disconnected passages still read-
able in a much broken original. The damage is old, as shown by Sus+4: [ j - a n - d e ' - a
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COLS, iv 45—v 16 101
45 In Uruk En-shakush-Anna(k)(?)
became king and reigned 1X60 years;

its kingship; reigned 2X60 years;
reigned 7 years.
(Emend 11. 47-48 to read
47a Lugal-kinishe-dudu
476 reigned x+2 years;
47c Lugal-kisal-si,
47d son of Lugal-kinishe-dudu,
48a reigned x+7 years.
486 3 kings
48c reigned its . . years.) 178
Col. v
its kingship to'Ur was carried.
In Ur Lugal-kinishe-dudu
became king and reigned . . years;
5 Lugal-kisal-si,
son of Lugal-kinishe-dudu,
reigned . . years;
gi
reigned . . years;
10 Ka-ku(g),
son of . . . .gi,
reigned . . years.
4 kings
15 Ur was smitten with weapons;
its kingship to Adab was carried.
[ ] - a6 [ 1 u] g a 1, which can be restored [ t i - g u b a] - a n - d 6 - a [.. m u 1] - as

[. . 1 u] g a 1 . . . . , "Somebody who has disappeared reigned . . years. . . kings "
See the detailed discussion of the passage on pp. 169-72.
179
Thus according to the photograph (cf. Langdon's translation).
180-180 xhe restoration of this section is based on the traces left in Sui. See the detailed
discussion of the section on pp. 175 f.
181
Restored on the basis of the dynasty total, which can be reconstructed in P2. See
pp. 173-75.
oi.uchicago.edu
Lx 17 [adabaki m lu gal - a n - n £ - m u -
u n - d il]
Lx [lugal4m mil 90184 i-a5]
(probably old lacuna here)184
WB Lx [1] (emend to 1+x?) 1 8 4 1 u g a 1
WB Li 20 [m u - b i 9]0186 i b186- a5
ki gi§
WB Lx [a d a b a ] t uku1 ba-an-sig
187
WB Li Sui [nam-lugal-b]i m a - e r i ki - § e187
b a - tiim
WB Li Su, [m a - e r i ki ] - § d188 a n - s u d4189
WB Li Su, [1 u g a 1 - k m] m u 30 i-a5
WB L, 25 [? 190 -zi] dumu n - s u d4 - k e 4
WB [mu J7(?)]191 i-ag
WB [ ] -1 u g a l m u 30 i - a5
WB [ ]192-lti-gal mu 20 l-a5
r 1
WB [ ] - b i -MUS4-MA& mu 30 i-a6
WB 30 [ ]-ni mu 9 i - a5
WB W lugal
WB [mur-bi] 136 i193(emend to f 6 ) - a 6
WB [ma-eriki gi
Hukul] ba-an-sig
WB [nam-lugal] - b i
182
Li adds -. a .
183
Two copies of an inscription of this ruler are known (Poebel, Babylonian Legal and
Business Documents . . . . Chiefly from Nippur [BE VI 2 (1909)] No. 130, p. 123 and PL 57;
PBS V, No. 75; transliteration and translation by Guterbock in ZA n.F. VIII [1934]
40-46). They were written in the reigns of Amml-§aduqa and Abi-Eshuh. The text deals
with the building of the temple of Nintu in Adab, 6 - n a m - z u , and of the establish-
ing of offerings and rites of this temple. Guterbock (op. cit. pp. 46 f.) is of the opinion
that this inscription is a late literary composition. This seems possible but is far from cer-
tain. The name is written l u g a l - a n - n a - m u - u n - d u (PBS V, No. 75) and
l u g a l - f a n ^ n ^ - m u - u n - d u (BE VI 2, No. 130), whereas the King List (Li) has
lugal-an-n6-mu-un-du.
« 4 See n. 186.
185
%«, which can be restored r«< (90); Li: 90.
186
Thus in both WB and Li; it would therefore seem to be original. But if the collective
form of the verb is old, we can assume that the dynasty originally had more rulers than
Lugal-Anne-mundu and that the others have been lost in the course of tradition. Langdon
also (OECTII 5) considers it probable that this dynasty once comprised several kings.
187-187 Thus also Li; Sux places the dynasty of Maeri after Akshak and apparently before
Kish III-f IV. Since the order of WB is represented in both the A branch (WB) and the
B branch (Li) of the tradition, we must consider it original. That the city name should be
read m a - e r i k i rather than m a - r i k l has been shown by Thureau-Dangin in RA XXXI
(1934) 83 f.
oi.uchicago.edu
CEITICAL EDITION OF THE TEXT, COL. V 17-34 103
17 In Adab Lugal-Anne-mundu183

(probably old lacuna here)
1 king (emend to 1+x kings?)
20 reigned its 90 years.
Adab was smitten with weapons;
its kingship to Maeri was carried.
To (emend to In) Maeri Ansud

25 . . . .zi, son of Anstid,
reigned 17(?) years;
. . . .-lugal reigned 30 years;
. . . .-1 d - g a 1 reigned 20 years;
. . . .bi-MTJ§4-MA§ reigned 30 years;
30 . . . .ni reigned 9 years.
6 kings
Maeri was smitten with weapons;
its kingship
188
Li: - a ; Siu omits. The -§ e of WB is erroneous, due to dittography of m a - e r i ki -
§ d in the line above.
189
The sign is BU in WB and I*; Sui is too broken to allow a decision. Contrary to
Landsberger (OLZ XXXIV [1931] 127) a variant AN-STJD on which he bases his reading
il-lu does not occur in the King List. It seems, however, to appear elsewhere; for the traces
on the macehead U 11678 can only be those of 'AN-SUD1 l u [ g a l ] (see the photograph,
UE II, PL 183 e, which gives mm^m wwm. Burrows in UE II 322 read ' a n - b u 1
r
l u g a l ; the catalogue, ibid. p. 572, apparently u t u1). In the inscription of this king's
1 rd
daughter (UET I, No. 12) his name is written AN-BU. The same variation between BU and
SUD met with in this name occurs also in z i - u4 - s u d - r a (see Poebel, PBS IV 1,
p. 49). Since the Greek transliterations E«rout?pos, 2i<rovdpos, Xun&pos, etc. show that this
name was pronounced Zi(u)sudra, we must assume that BU had the value s u d4. As
shown by the writing a n - s u d i n U 11678, BU has this value in the name under discussion,
and we have therefore adopted the reading a n - s u d i .
190 Only one sign seems to be missing before - z i. Legrain's copy gives mfo, which could
be remnants of g a z .
191
The figure is not preserved in any of the versions. Our restoration of 17 is arrived
at by subtraction of the reigns preserved from the dynasty total.
192
2&% is preserved in the lacuna before -1 ti - .
193
Error for i b - .
oi.uchicago.edu
WB 35 [kisik%-Ml1M b a -1 a m
WB 8 Su 3 [ki§iki195 • kiL-d]ba-ba6196
WB S Su3 i97[mu10i i i - k u r u n ] - n a197
WB S Su3 [ s u b us 1 9 9 k i s i ] k i mu-un-gi200-na
202
WB S Su 3 [lugal]-&m202 mu 100203 i-a5
WB 40 1 lueal
WB mu-bi 100 i-a5
WB k i § i ki gi
Hukul ba-an-slg
194
P2) La, and Sui also must have had this arrangement. The fact that P2 lists four
dynasties of Kish (xi 5-6) means that it must have counted Ku(g)-Baba as a separate
dynasty. Li has after the dynasty of Akshak a dynasty of Kish beginning with Ku(g)-
Baba's son, Puzur-Sin, just as in WB; here also Ku(g)-Baba must have been counted as
a separate dynasty preceding Akshak. In Sui, finally, the dynasty total giving 7 rulers
(vi 8) shows that it had a Puzur-Sin dynasty as in WB, i.e., that Ku(g)-Baba must have
been listed separately. A different arrangement is found in S and Su3+4, where Ku(g)-Baba
has been moved down after Akshak and appears as first ruler of the dynasty which in the
other versions begins with her son.
Since the arrangement with Ku(g)-Baba as a separate dynasty before Akshak is attested
in both main branches of the tradition (A branch: WB; B branch: P2, Li, and Sui), it should
be original, and this is also shown by a consideration of the variant itself. The arrangement
found in WB, P2, Li, and Sui can very easily have been changed in the course of tradition to
that of S and Su3+4> for it is obvious that any copyist who noted that in his original Ku(g)-
Baba was separated from her son by a complete dynasty would feel a strong urge to correct
this absurdity and move her down to join her son and his successors. While a development
from the arrangement of WB, P2, Li, and Sui to that of S and Su3+4 would thus be very
natural, it is clearly impossible to imagine that any copyist should have changed the plau-
sible-looking arrangement in the latter texts to that of WB etc. It is therefore clear that
WB, P2, La, and Sui are original on this point.
As we shall see below (p. 177), the author of the King List arrived at his curious arrange-
ment because he possessed a synchronism showing that the dynasty of Akshak was roughly
contemporaneous with Ku(g)-Baba; and the quite unbelievable reign which he has assigned
to that ruler, 100 years, was meant to carry her across the 99 years of the Akshak dynasty
down to the time of her son (first realized by Langdon, OECT II 7). See also &e discussion
of this variant on pp. 53 f.
195
S adds - a ; Su3+4 adds - a4 .
196
This queen is mentioned in the list of postdiluvian rulers "not arranged in order"
given in V R 44 i 14, and her name is there translated dba-ba& el-lit, "Baba is bright" (Poebel,
PBS IV 1, pp. 129 f., rejects—hardly with right—this translation and suggests that the
name means "silver of Baba"). An omen (CT XXVIII [1910] PL 6, K 766:2-3) refers
tofM-d6a-6a6 $a mata i-be-lu4) "The woman Ku(g)-Baba who ruled the land." An anecdote
concerning her and Puzur-Nirah of Akshak is related in Chronicle Weidner (see Gtiterbock
in ZA n.F. VIII 51 and 54). We have retained for the name of the goddess dBA-tj the
reading d b a - b a 6 suggested by Jensen in Thureau-Dangin, Les homophones sumtriens
(Paris, 1929) p. 40, although Thureau-Dangin in RA XXXII (1935) 150 has advocated a
reading of the last sign a s b u g . The passages in favor of reading b u 8 , namely phonetic
renderings such as d b a - b u , are all late and can (as stated in Les homophones sum6riens,
p. 40) represent an Akkadianized form BaM, BoM, derived from Sumerian B a b a like
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. V 35-42 105
35 to Kish was carried.

In Kish Ku(g)-Baba,
a barmaid,198
the one who consolidated the foundation of Kish,201
became "king" and reigned 100 years.
40 1 king
Mamti, Mami, from M a m a . The reading b a 6 , on the other hand, has the support of
a phonetic writing dating back to the Isin period, namely SRT, No. 5 rev. 41-43: u r
mu-ti-in-na-sd di-di-de ba-ba g a - b a - b u l - bu 1 - l e - e n - d e-e n
. . . . d b a - b a6 SAL - l a - m a - k e i - e § g a - b a - p i - ^ d l - l e - e n - d ^ - e n , "Go-
ing toward the embrace of the husband, O Baba, we will rejoice; O Baba, on account of
my . . . . we will rejoice," where b a - b a in 1. 41 obviously corresponds to d b a - b ae
in 1. 43.
197-197 Thus both S and Su3+4. It is not impossible that WB had m u i o - k u r u n -
n a, which is a better form (see Poebel, PBS IV 1, p. 129, n. 4.); cf. n. 199.
198
Lit., "a wine-woman." We have adopted Barton's translation (RISA, p. 343) in pref-
erence to the perhaps more exact but also more pedantic terms "a female wine seller"
(Langdon, OECT II 15) and "hostess of a tavern" (Gadd, The Early Dynasties of Sumer
and Akkad, p. 4).
199
Thus S and—to judge from the spacing—almost certainly WB; Su 3+ i: m Uio
s u k u s . As m uio is omitted in both main branches of the tradition (A branch: WB;
B branch: S), the form without it must be nearer the original. On the other hand, since
parallelism with the other "historical notes" seems to call for a relative pronoun here and
since, strictly speaking, the agreement of WB and S only shows us a text preceding the
separation of the two main branches of the tradition, it seems just possible that the earliest
form of the text had 1 ti at this point. This 1 u might then be responsible for the incor-
rect mui°l u - k u r u n - n a instead of m uio - k u r u n - n a in the line above (see n. 197).
200
Thus S also; Su3+4: - g U - .
201
The meaning of the phrase is that Ku(g)-Baba "laid the foundations for the political
and economical strength and importance of the city" (Poebel, PBS IV 1, p. 129).
202-202 g U 3 + 4 omits; g. l u g a 1 - a m .
203
This abnormally long reign represents—as noted by Langdon—an effort to carry
Ku(g)-Baba across the 99 years of the dynasty of Akshak which the author had inserted
between her and her son Puzur-Sln (see n/194 above and p. 177; cf. also pp. 159 f.). The
original figure for Ku(g)-Baba's reign, that assigned to her in the source used by the author
of the King List, is unknown. The emendation of the present figure to 14, suggested by
Peiser in OLZ XV (1912) 154 f. and by Poebel ibid. col. 290, was based (1) on the assump-
tion that an original <ft had been misread as W., which must be considered highly unlikely,
and (2) on a particular explanation of the difference in S between the sum of the single
reigns and the dynasty total which at that time seemed plausible but which other versions
of the King List, found after Peiser and Poebel wrote, have shown to be erroneous. (The true
explanation of this difference is that original W&, i.e., 400, for Ur-Zababa(k) > m»] > **,
i.e., 6; see n. 218 below.) There is thus nothing in favor of this emendation, and it should
be definitively abandoned.
oi.uchicago.edu
204
WB 43 nam-lugal-bi u4 - k u § ti ki 2 0 4 -
m
(sd) ba-tiim
WB S Sui G u 4 - k u § ii ki - § e206 u n - zi
WB S Su! G 45 lugal-^m mu 3[0207 l-a5]
208
WB S Sui G un-da-lu-lu mu 6 (emend to
12) i - a 5
WB S Sui G ur-ur mu 6 i - a5
Col. vi
S Sui G [ p u z u r - d n i r a b209 mu 20 i - a5]
S Sui G [i-su-il mu 24 i - a5]
S Siu [ § u - d s f n d u m u i - s u - i l 2 1 0 m u 7211
i-a5]
S Sui [6 l u g a l ]
Li S Sui 5 [mu-bi 99 212 fb-a5]
ki gi
WB L, S Sui u4-kusii Hukul [ba-an-sig]
WB Li S Sih [n a m] - 1 u g a 1 • b i
WB U S Sui k i s i k i - s e213 b a - t li m
WB Li S Su : k i s i ki 2 H p li z u r - d s i n
WB la S Sus 10 dumu k ft - d b a - b a 6 - k e4215
216 216
WB L, S Sll; lugal-&m mu 25 i-a5
1 d 217
WB L! S 2 nr u r - z a - b a 4 - b a4
204-204 ^he rea rling of the ideogram tJ|j ki as Akshak was proved by a variant in the Nahr
el-Kelb inscription of Nebuchadnezzar (E. Unger and F. H. Weissbach in ZA XXIX
[1914/15] 183). That this reading is original is shown by the "ideogram" itself, which is
clearly an old phonetic writing u4 - k u s u representing a pronunciation a k a s a (k) .
Note also the early passages l u g a l u4 - k u s ti ki - k a ( « 1 u g a 1. . . .k - a (k) ) , "the
king of k" (E-Anna(k)-tum, SAK, p. 20 6 iv 25), and Iti U4 - k u & u k i - k a - k e 4
ki
1 u k i s i - k e4 ( = 1 u . . . .k - a k - e 1 u K i s i - (a) k - e ) , "the man of k (and)
the man of Kish" (En-shakush-Anna(k), PBS IV 1, p. 151, 11. 13 f.), which show that
the name ends in k.
205
- s e" omitted by mistake; the copyist looked at u4 - k u § u ki one line too far down.
206
S and G: - a ; Sui omits. The - s e in WB is dittography from xu - k u § ii ki - § e ,
which appeared in the preceding line in the scribe's original.
207 4%. restore «* (30) after S.
208
Dittography from m u 6 1 - a5 in following line. S has 12, which is correct as shown
by the correspondence of the resulting dynasty total, 99 years, with the 100 years assigned
to Ku(g)-Baba. See p. 177.
209
On this reading see Landsberger, Die Fauna des alten Mesopotamien, pp. 60 f.
210
ThusSui;Sadds-ke4.
211
Thus S; Sui: 24, by dittography from the similar i - § u - i 1 24 m u 1 - a5(or - n a?)
in the preceding line.
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COLS, V 43—vi 12 107
43 its kingship <to) m Akshak204 was carried.
To (emend to In) Akshak Unzi

45 became king and reigned 30 years;
Undalulu reigned 6 (emend to 12) years;
Ur-ur reigned 6 years;
Col. vi
Puzur-Nirah reigned 20 years;
IsM-il reigned 24 years;
Shu-Sin, son of Ishfl-il, reigned 7 years.
6 kings
Akshak was smitten with weapons;
its kingship
to Kish was carried.
In Kish Puzur-Sin,
10 son of Ku(g)-Baba,
Ur-Zababa(k),217
212
Thus Li and S; Sin has WfW = WKW (116), due to its erroneous figure, 24, for the
reign of Shti-Stn (see n. 211).
213
Thus also Li; Sui places the dynasty of Maeri here (see n. 187 above).
214
Li adds - a . In S and SU3+4, where Puzur-Stn is the second king of the dynasty
(see n. 194 above), there are, of course, no introductory formulas with his name.
216
Thus S and Li also; S113+4 omits.
216
Thus Li also. In S and Su3+4, where Puzur-Sin is the second king of the dynasty
(see n. 194 above), there are, of course, no introductory formulas with his name.
217-217 Q n the reading of the divine name dZA-MAL-MAL as dz a - b a< - b a4 see Ungnad
in OLZ XXV (1922) 202 f. and especially Weidner in Archiv fur Keihchriftforschung II
(1924/25) 13. Weidner gives good reasons for doubting the authenticity of the variant
reading d Ilbaba (ibid, n. 7). The evidence lately produced in its favor by Ungnad (OLZ
XL [1937] 733, n. 1, a reference which we owe to Dr. Gelb) seems to us too uncertain to
alter the situation materially. On the reading of the name cf. also the musical instrument
*ihur-za-ba-bi4'ti (vars. [ur~z]a-b[a(7yb]i4u[m]f ur-za-bi-tum, and *iSur-2a-pt[?Hu[?]) men-
tioned in the third tablet of d i r i ~mn~siiaku — watruf which was named after our
king (A. Falkenstein and L. Matous in ZA n.F. VIII 147).
According to the note in 1. 33, Sargon of Agade was originally cupbearer of Ur-Zababa(k).
An anecdote concerning Ur-Zababa(k) and Sargon is related in Chronicle Weidner (Giiter-
bock in ZA n.F. VIII 49 and 52). Ur-Zababa(k) is mentioned in the legend of Sargon and
oi.uchicago.edu
WB Li S 13 [d u m u] p ti z u r - d s 1 n - k e4
WB Li S [m] u 400218 i - a5
219
WB S Sui Su4 15 [s i - m] u - d a r - r a220 219 m u 30221 1 - a 5
222r 1
WB S SU4 u -§i-wa-tar (dumu sl-mu-
222 223
dar-ra-ke4> mu 7 i-a 5 2 2 4
225
WB S Sui e s 4 - t a r 2 2 5 - m u - t i m u ll 226 i - a 5
227
WB S Sui i § - m [e]227-ds a m a § m u 11 l-a 6 2 2 8
229
WB S n a - a n - n i - i a229 z a d i m230 mu
231
7 l-a5
Lugal-zage-si, AO 7673 (Scheil in RA XIII [1916] 175-79; H. de Genouillac, TRS II, No. 73
[PL CXLII a]; cf. Giiterbock, op. tit. p. 37). In later times he was considered a god and ap-
pears as GUD-DtJB of Ninurta in the series AN: dan-nu-um (CT XXIV, PL 8:5; Deimel,
Pantheon babylonicum, No. 1241).
218
S: W (6). Since the dynasty total given in S, 586 years, presupposes that S originally
had the same figure as WB, 400 years, for the reign of Ur-Zababa(k), the present 6 must
have developed from original BM1 (400) through a damaged text in which the tens were
broken away: ftf$]. Why such an extremely high figure was assigned to Ur-Zababa(k) is
uncertain. Perhaps the author had reasons similar to those which mad3 him give Ku(g)-
Baba 100 years (see n. 203 above). Langdon's explanation (OECT II 16, n. 10) that ffi«
originally stood for 6 40/60 is inacceptable, as such writings are never used in King
List.
219-219 Simu-dar appears in this place in S and apparently in Su3+4, where only [ ] - d a r
is left. In Sui Simu-dar is listed as the last king of the dynasty, preceded by a fictitious Shu-
ilishu and by Ishm^-Shamash. The order of WB, S, and Su3+4, which is attested in both
main branches of the tradition (A branch: WB; B branch: S and SU3+4), is of course origi-
nal. As for Sui, there can be little doubt that it derives from a version in which part of the
dynasty was blotted out by a serious lacuna, a lacuna which the copyist tried to restore as
best he could from other broken fragments. See also the following notes.
220
S, Sui, and SU3+4 omit. The final - r a in WB is probably dittography from d u m u
s i - m u - d a r - r a after Ost-watar, which suggests that WB originally contained that
phrase; see n. 222.
221 Thus S also. The figure is not preserved in Su3+4. Sui, which lists Simu-dar as the
last ruler of the dynasty, i.e., in the place which belongs to Nannia, also gives him Nan-
ma's reign of 7 years. We must therefore assume that the lacuna in the original of Sui
(see n. 219) ended just after the name of Nannia: [ n a - a n - n i - i a z a d i m ] 7 m u
1 - a$, so that the copyist has only restored a wrong name here.
222-222g. u - s i - w a - t & r d u m u s i - m u - d a r - r a - k e 4 . Sus+4, which has [sf-
m u] - d a r [.. m u i] - a* [li - s 1 - w a] - 1 £ r [. . m u i] - as, seems to follow WB. Since
the Su texts are derived from badly damaged originals and since there are traces in
WB suggesting that its original had the words d u m u s i - m u - d a r - r a - k e * (n. 220),
it is likely that S here preserves the original form of the text.
223
S: 6. WB's figure is more probably original, since B (6) could very easily develop
put of an original W (7) through a slightly damaged text, as indicated in n. 231 below.
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COL. vi 13-19 109
13 son of Puzur-Sln,
reigned 400 years;
15 Simu-dar219 reigned 30 years;
"0§f-watar<, son of Simu-d&r,> reigned 7 years;
Eshtar-muti reigned 11 years;

Ishme-Shamash reigned 11 years;
Nannia, a stonecutter, reigned 7 years.
224
Su3+4 gives the dynasty total after this line. It must therefore derive through a text
which had a lacuna from here down to the end of the dynasty. As we have seen above (n.
219), Sui also shows evidence of being derived through a text with such a lacuna; so we
have here further proof that Sui and SU3+4 derive from a common ancestor (see pp. 25
and 50 f.).
225-225 g. ^tr j obviously remnants of e h -1 & r .
226
Thus S also; Sur. W , probably remains of <W (17). It is impossible to say how
this figure originated. The passage is very badly corrupted (see n. 219 above). The con-
current testimony of WB and S, representing both main branches of the tradition, of
course identifies 11 as original.
227-227Thus Sui also; S: i - m u - . Since i § - m e - is represented in both branches of
the tradition (A branch: WB; B branch: Sui), it is presumably original.
228
After this line Sui lists a king § u - l - l i - § u who reigned <W (15) years. This
king can be no other than the well known Shu-ilishu of Isin; and, comparing the account
of the Isin dynasty given by Sui, we may perhaps assume that the copyist had a loose,
unplaced fragment which he first used when he restored the lacuna in this section (see
n. 219) but which later on he was able to place and to "join" where it actually belonged,
after Sui viii 18:
That the source of Sui and Su3+4 was in an utterly damaged and broken condition is evi-
denced by many details (see pp. 25 and 50 f. and nn. 219, 221, 224, and 226 above).
229-229g o m i t s _ a n _ (cf# n# 231); Sui restores Simu-dar here (see n. 221).
230
This reading, first pointed out by Thureau-Dangin, La chronologic des dynasties de
Sumer et d'Accad, p. 62, is certain. That the sign is not a b but z a d i m is obvious from
a comparison of 4* in n a - a n - n i - i a z a d i m (WB vi 19) with ^ (a ^ / i fc) in
b a - l i - i b (WB ii 20).
231
Thus Sui also (cf. n. 221); S has tt? (3). It is probable that the latter figure is sec-
ondary and derives from an original W (7) through a damaged text: lW (cf. n. 272). Note
also that some other variants in S (see nn. 223 and 229) seem due to omissions around the
oi.uchicago.edu
WB S Sui Su4 20 7232 l u g a l

WB s Sui Su 4 mu-bi 491 233 1 b - a5
ki i§
WB s Sui Su 4 k i §i « tukul ba-an-sig
WB s Sui Su 4 n a m - l u g a l - b i unugki-§6 ba-tiim
WB s Sui Su 4 u n u k i - g a234 lugal-z&-ge-si
WB s Sui Su 4 25 1 u g a 1 - & m mu 25 l-a5
WB s Sui Su4 1 lugal
WB s Sui Su4 m u <- b i>235 25 i-a6
WB s Sui Su 4 u n u g ki gii
t ukul ba-an-sig
WB s Sui Su 4 nam-lugal-bi
WB s Sui Su 4 30 a - g a - d e 2 3 6 k i - § 6 b a - 1 lim
ki 237
WB s Sui a - g a - d e< > lar-ru-ki-in
WB s . . 238 - b a - n i n u - g i r ii2239
WB Li s qa-§u-du8 u r - z a240 - b a4 - b a4
d
242
WB U s l u g a l a-<ga-> d e k i lii a - g a - d & k i
WB Lj s 35 m u - u n243 - d u - a
WB Li s 1u g a 1 - & m mu 56245 i - a5
WB Li rf-mu-u§ dumu §ar-ru-ki-in
WB Lj m u 9246 i - a5
middles of the lines. All these variants will be explained if we assume a single lacuna, a
slightly widened crack, in some predecessor of S:
•§i-wa-tilr
232
Thus Sui also; S, which counts Ku(g)-Baba as a member of this dynasty, has 8.
233
S: 586. This text counts Ku(g)-Baba (100 years) as a member of the dynasty, and
the total is correct if the present figure Iff (6) for Ur-Zababa(k) is restored to its original
form ?R#] (400); see n. 218. The total of Sui, 485 years, stands apart and is probably
the sum of the single reigns after the scribe had restored the lacuna found in his original
(cf. nn. 219 and 221). It is thus altogether secondary.
234
Thus S and Sui also; Su3+4: - W .
236
Erroneously omitted in WB, probably by dittography of the similar m u 25 in the
preceding line. S and Sui correctly: - b i .
238
Thus according to the photograph. The SAR which Langdon's copy shows here is erro-
neous, a modern dittography from the following line: a - g a - d & sar-ru-ki-in.
oi.uchicago.edu
CKITICAL EDITION OF THE T E X T , COL. vi 20-38 111
20 7 kings
In Uruk Lugal-zage-si
25 became king and reigned 25 years.
1 king
its kingship
30 to Agade was carried.
In Agade Sharru(m)-kin—
his . . . . was a date-grower—
cupbearer of Ur-Zababa(k), 241
king of Agade, the one who
35 built244 Agade,
Rlmush, son of Sharru(m)-kin,
reigned 9 years;
237
S adds ki - a , which is correct.
238
According to the photograph WB has pzy. S has (according to the photograph in
RA IX [1912] opp. p. 68) *££. It may be definitely stated that none of the readings thus
far proposed is compatible with the sign as given by the photographs. These proposals are:
Poebel, PBS IV 1, p. 130, n. 10: a b - b a - n i , "his father"; Thureau-Dangin, La
chronologie des dynasties de Sumer et d'Accad, p. 62: s a r - r u - k i - i n - l u - b a - n i ,
"qu'un roi legitime soit cre6" (proposed before WB had become known); Langdon, OECT
II 17: § a r - r u - k i - i n - l - l u - b a - n i ; Zimmern in ZDMG n.F. I l l 33, n. 10: i -
d i b - b a - n i , "dessen Pflegevater(?), dessen Aufnehmer(?)"; Guterboek in ZA n.F.
VIII 4; ( ! ) - d i b - b a - n i , "dessen 'Aufnehmer.' "
239
On this reading of GIS-SAR in our word see Poebel in AJSL LI (1934/35) 172.
240
Thus Li and S. The vertical wedge appearing between z a and b a* in Langdon's copy of
WB is—as shown by the photograph—part of an erased b a* over which z a is now written.
241
Listed as a ruler of Kish in 1. 12 above. On the reading of the name see n. 217.
242
Omitted in WB by mistake; preserved in Li and S.
243
Li adds - d a - ; S is broken but seems to agree with Li.
244
Var. in Li and S(?): "the one under whom Agade was built" (cf. n. 119 above).
245
Li: 55. P 2 , P3, S, Sui, and Su3+4 seem to have followed Li (see pp. 23-26). The figure
of WB, 56, is more probably original than 55, since the latter can so easily have developed
from the former through a damaged text, -$K$i, and since WB as a whole preserves the bet-
ter text in this section (see pp. 26 f.).
246
Li: 15. P2, P3, S, Sui, and Su3+4 seem to have followed Li (see pp. 23-26). The figure
of WB, 9, is original. The figure 15 seems to be due to dittography from the almost identical
d u m u S a r - r u - k i - i n m u 15 I - as in 11. 41-42 (see p. 27).
oi.uchicago.edu
WB U 39 m a - n i - i s - 247t i - i §247 - § u
WB Li 40 s e § - g a 1 r i - m u - r u s1248 - u S
WB La dumu sar-ru-ki-in
WB Li mu 15250 [i-a 5 ]
rd1
WB Li Su3 n a - r a - a m - [s2n]
WB Li Su3 dumu m a - [ra- i s - t i - i s - su]
WB P 3 Li Su3 45 m u [S7(?)251 i-a5]
252
WB P 3 Li S Su3 §[ar- ka-H-§ar-ri]252
P 3 Li Su3 [dumu n a - r a - a m - ds in]
P3Lj Su3 [m u 25 254 i - a 5 ] 255
Col. vii
256
WB P 3 U S Su 3 [a-ba4m lu]gal257 a-ba-im
256
n u 1 ugal
WB P 3 Li S Su 3 [258i259 (var. i r) - g i4260] - g i4261 1 uga 1
247-247LI: - t e - . This king's name in his own inscriptions is always written ma-an-
is-tu-su, which represents man istu-su, "Who is with him?" (on -su written -su in this
period cf. e.g. Ungnad in MVAG XX 2 [1915] p. 14). At the time of the 3d dynasty of Ur
the form iste of the preposition istu/iste became predominant, so our name changes to Man-
ishteshu (man iste-su); see W. von Soden in ZA n.F. VII (1933) 138 and n. 1.
248 WB is damaged at this point. Langdon read the traces as - § u" - ; but the photo-
graph shows '#*T, i.e., - ru s1 - , so it is apparently a case of simple dittography. It may
even be that the scribe erased this first - u s - because he was not satisfied with the spac-
ing and wrote a second one which filled out the line. Li has correctly r i - m u - u § .
249
The meaning of the name, "Who is with him?" (see n. 247), might suggest that
Rfmush and Man-ishtushu were born as twins.
250
Li: 7. P2, P5, Sui, and Su3+4 seem to have followed Li (see pp. 23-26). The figure
of WB, which caused the wrong figure for Rfmush in Li (see n. 246), must be the original.
251
L! has 56, agreeing with P3's %*TO, which also can be restored as £WffJ (56). In WB
the figure for Naram-Sin and likewise the figures for Shar-kali-sharrI and for the inter-
regnum are missing; and if we restore them from the texts of the B group as 56, 25 (this
seems better than 24; see n. 254 below), and 3 we get a much higher sum of single reigns than
that indicated in WB's total, 181 years. One of these figures must therefore be too high.
Using only the lower two figures, 25 for Shar-kali-sharri and 3 for the interregnum, we have
37 years left for Naram-Stn. This figure is indeed very probable; for, looking more closely
at the figure 56 given for this ruler in the B group, we see that it appears just above the sign
S^R of § a r - k a - l i - § a r - r i , which begins with two vertical rows of -(-wedges: $ft\
A copyist could therefore easily misread the uppermost of these wedges as part of the nu-
meral above: ^ ^ . Thus the B figure 56 may derive from ah original 36, which is very
near to the 37 years left for Naram-Stn by the total of WB. We have already noted many
examples of loss of a final unit by transmission through a slightly damaged text.
252-252 Thus in P3, Li, and probably S (S a r - k [a -1 i - § a r - r i]) . Su3 has >- in-
stead of - k a -1 i - , probably because it descended through an ancestor in which only
one wedge of these signs remained, and - r i instead of - r i .
oi.uchicago.edu
CRITICAL EDITION OP THE TEXT, COLS, vi 39—vii 2 113
39 Man-ishtushu,
40 the older brother 249 of Rfmush,
son of Sharru(m)-ktn,
reigned 15 years;
Naram-Sin,
son of Man-ishtushu,
45 reigned 37(?) years;
Shar-kali-sharri,
son of Naram-Sin,253
reigned 25 years.
Col. vii
Who was king? Who was not king?
Was Igigi (var. Irgigi) king?

253
In Shar-kali-sharri \s inscription given in SAK, p. 164 d, the line TUR-DA-TI den-lil
(i 2), which is generally read mar da-ti-den-lil, "son of Dati-Enlil" (doubted by Poebel,
PBS IV 1, p. 132, n. 4; rejected by Weissbach in RLA II 196), is—as first seen by Hrozny
in WZKM X X V I (1912) 151, n. 2—in all probability an ideographic writing litti" den-lil,
"offspring of Enlil" (SL, No. 144.48 0: TVR-DA =lit-tum)t a title which is closely paralleled
by the title li-tum dda-gdn, "offspring of Dagan," taken by Hammurabi (Code of Hammu-
rabi iv 27). This passage, therefore, should not be quoted—as is sometimes done—to prove
the unreliability of the King List.
254
Thus Li and Su 3 ; P3 has 24. Since it is more likely t h a t an original 25 lost a final unit
in transmission than that there was an accretion of one unit, we may assume t h a t the figure
25 attested by Li and Su 3 is correct.
255
P 3 inserts here a subtotal for the family of Sargon: y/M$f m u [ s]ar-ru-ki-
n i m , which may be restored: [TO«W m u [ d u m u - d u m u s ] a r - r u - k i - n i m , "157
years the family of Sargon" (cf. Poebel, PBS IV 1, pp. 80 and 132 f.). The copyist of
P3 seems to have had a predilection for such subtotals (see n. 104 above).
256-266Thus also S. Su 3 gives a - b a l u g a l a - b a l u g a l - a m , which is clearly
a corrupt form of the text found in WB and S. Li and P 3 have the line in Akkadian: ma-nu-
um sar ma-nu-um la sdr. See the discussion of this variant on pp. 52 f.
257
Thus according to the photograph.
258
S, which writes the names of these four kings on two lines, adds a vertical wedge in
front of each name, perhaps to indicate that they were arranged on four lines in its original,
as is the case in the other versions.
259
Thus S. P 3 gives i - ; Li and Su3 give i r - . It is obvious t h a t i r - is a misreading
of the very similar sign i - , or vice versa, and that i - is just a variant writing of
i - . Whether 1 - or - i r - is original must remain uncertain; cf. the discussion of the sec-
tion as a whole on pp. 52 f.
260
Thus P 3 , S, and Su 3 ; Li: - g i5 - .
261
Thus P 3 , S, and Su 3 also; U: - g [i 5 ].
oi.uchicago.edu
WB P, Li S Su s 3 P68!! a - n u262 - u m2M] lugal

WB P 8 Li s SUs pH-mi] lugal
WB P, s Su3 5 [*«eM»-lu-lu] lugal
WB P 3 s Sui Sua *> 7 [4-bi] lugal
WB P s s Sux [mu 3] f b - aB267
WB P 3 s <SUl> [d u - d u m] u 21 1- a6
WB s <Su,> [§u-dur-ul] d u mu d u - d u - k e4
WB s <Sut> 10 [mu] 15 1- a5
WB(P 2 ) s Sin Su 4 ll 2 6 8 lugal
lgl269
WB(P 2 ) s Sui Su 4 mu-bi
ki iS
f b-as
WB s Su, Su 4 a-ga-d& '* t u k u 1 ba - a n - s i g
WB s Sui Su 4 nam-lugal-bi u n u g k i -fid b a - t u m
270
WB s Sui Su 4 15 u n u k ! < - g a > ! !7° ur-ni gin lugal-
am 2 7 1
7272
WB s Sui Su 4 mu
gi5
l-aB
WB s Sux Su4 u r- gigir dumu u r - n i g 1 n273
6274
WB s S Ul Su4 mu
276
i - a6
WB s <SUl><Su4> kud-da mu 6 l-a5
WB S <Sui><Su4> 20 p l i z u r - i - l i mu 5 i-a 5 275
282
Thus Li, S, and Su3; P 8 - n i .
288
Thus S and Su3. Li is broken here; Pa omits.
264
Nanum occurs in this place in Li and Sua; P3 and S have the order Irgigi, Imi, Nanum,
Elulu. Which is original? See also the discussion of this section on pp. 52 f.
266
In this place in L* and SU3; in P 8 and S, before Nanum (see n. 264).
286
Thus P 3 ; S and Su3: i - .
2868
To this king belongs probably the inscribed dagger published by Selim J. Levy in
AOF X (1935-36) 281: IA4U-UI DAN Mr a-ga-dhki. Gelb, who comments on this name in
AJSL LIII (1936/37) 38, adduces evidence for a value t* for LI and proposes a reading
ix-lu-ul-dan. In view of the variant writings of the name of the e n s i (k) of Lagash en-
M-tar~zi and en-e-tar-zi (cf. Poebel in ZA n.F. IV [1929] 82) a reading e^-lu-ul-dan should
perhaps also be taken into consideration. In either case Elulu of the King List may be
considered an abbreviated form of the name.
267-267 -phis seems the best restoration of WB; it fits also Sui and Su3+4. P 3 and S have
4 - b i 3 m u i b ~a6 (S: i n - a 6 ) . Cf. pp. 52 f.
288
S: 12; Sui has m, and Su3+4 has fff, both of which probably represent w (9). The figure
of WB must be considered original; see p. 25, where the problems which these figures raise
are discussed in detail.
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. vii 3-20 115
3 Was Nanum284 king?

Was Imi265 king?
5 Was Elulu266* king?
Their tetrad was king
and reigned 3 years!
Dudu reigned 21 years;
Shu-Durul, son of Dudu,
10 reigned 15 years.
11 kings
Agade was smitten with weapons;
15 In Uruk Ur-nigin(ak) became king

Ur-gigir(ak), son of Ur-nigin(ak),
reigned 6 years;
Kudda reigned 6 years;
20 Puzur-ili reigned 5 years;
269
S: 197. With S agreed P 2 (in final summary mw = 197) and in all probability Li also.
Sin: 161; S113+4: 177. We must consider the 181 years of WB original; see pp. 23-28, where
the problems raised by these figures have been discussed in detail.
27o~27o Thus according to the photograph; the - § h given in Langdon's copy is not in
the text. S: u n u k i - g a ; Sus+4: u n u g k i - a i ; Sui omits. The correct form is obvi-
ously u n u ki - g a .
271
Thus S and Su3+4 also. Sui erroneously: - a n , the copyist having skipped the A of
A-AN = am.
272
S: 3; Sui: 30; Sus+4: 15. Sui and Su8+i derive in this section from a common, very
much damaged, original (see pp. 25 and 50 f., also nn. 219, 221, 224, 226, and 228 above)
which the scribes have tried to restore. Their evidence is accordingly of little value. As for
the figures given by WB and S, 7 and 3, it seems probable that 7 is original; for while w,
through a broken text, W, can easily develop into **, 3 (cf. n. 231 above), the opposite
development would be difficult to account for.
278
Thus according to the photograph. With WB agree Su! and SU3+4; S adds - k e < .
274
Thus S also; Sui: 15; Su»+4: 7. Since 6 is attested in both main branches of the
tradition (A branch: WB; B branch: S), it must be original. Sui and Sus+4 derive through
a badly broken text (see n. 272 and passages quoted there). How the scribes arrived at the
figures 15 and 7 cannot be determined with certainty.
276-276 These lines occur only in WB and S; Sut and Sug+4 omit them, which can only
mean that their common original had a lacuna here (see p. 51).
oi.uchicago.edu
WB S Sui Su 4 21 276u r - d u t u276 m mu 6278 i-a5

279
WB S Su 4 5 lugal
WB P4 S Su 4 mu-bi 30280 i b - a5
WB P4 S Su 4 u n u gki gi§
t ukul ba-an-sig
WB P4 S Su4 .25 n a m -1 u g a 1 - b i
281
WB P4 S Su4 k i - s u - l u - u b 4 < - g a r > 2 8 1 gu-tu 2 8 2 -
u m283 ( k i -se> b a - t li m
284
WB<P4>LX Su4 k i - s u -1 u - u b4 <- g a r> gu-tu-
u mki 284 .
285
WB<P 4 >L! G Su 4 lugal mu nu-tuk"8
276-276 Thus S and Su3+4 also; Sui has an otherwise unknown ruler, l u g a l - m e - l a m .
It seems probable that the common ancestor of Sui and Su3+4 served as original for Su3+4
when the lacuna mentioned in n. 275 reached to just above Ur-Utu(k) and for Sux somewhat
later, when the name of this ruler also had disappeared. The curious l u g a l - m e -
l d m restored here by the scribe may represent an original l u g a l - r e 1 - n e ( = = l d m ) ;
see p. 51.
277
Sui and Su3+4 add d u m u u r - «Vag i g i r ; S agrees with WB. Since the text of
WB is represented in both main branches of the tradition (A branch: WB; B branch: S),
we must consider it original. The addition in the related sources Sui and Su3+4 is clearly
free invention in continuation of u r - «i5g i g i r d u m u u r - n i g i n , made after
the intervening rulers Kudda and Puzur-ili had disappeared.
278
Thus S also; Sui: 7; Su3+4: 25. Since 6 is attested in both main branches of the tra-
dition (A branch: WTB; B branch: S), it must be original. How the variants in Sui and
Su3+4 originated cannot be determined with certainty.
279
Thus S also. Su3+4 has 3, which is correct for it, inasmuch as the rulers Kudda and
Puzur-ili have been lost in this version (see n. 275 above and p. 51). The figure 5 attested
by both WB and S must be original.
280
S: 26; P4 also, which preserves 4/V^, should obviously be restored as -«m (26). The
figure 26 equals the sum of individual reigns preserved in S, where Ur-nigin(ak) is listed
with 3 years instead of 7 as in WB. Since 7, as we have seen above (n. 272), is probably
more correct than 3, the total of WB should be preferred. Su3+4 gives 47 years as total,
which agrees with its single reigns: 15-f-7 +25 = 47. As the total is dependent on secondary
and badly corrupted individual reigns, it can be ignored.
281-281 p 4 : k i - s u - l u - u b - [ ] ; S: k i - s u - l u - u b - g a r ; Su3+4: [m] a - d a . We
consider the omission of - g a r in WB accidental, since this ms. uses the correct form of
the word in col. viii 1.
282
P 4 and S: - 1 i - ; Su3+4: - t u - .
283
S and Su3+4 add ki - s e ; P4 is broken here.
284-284 L I : k i - s u - l u - u b - g a r g [ u - ] ; P4 omits; Su3+4: [ m a - d ] a gu-tu
oi.uchicago.edu
CRITICAL EDITION OP THE T E X T , COL. vii 21-28 117
21 Ur-Utu(k) reigned 6 years.

5 kings
25 its kingship
<to> the horde of Gutium was earried.
In the horde of Gutium
A king without name!
285-285Thus probably G also; restore [ . . . . ] ' n u - t u k 1 . Li: l u g a l n u - u b -

t u k ; Sug+4: [ ] > - a n - d 6 ; P 4 omits. The text*of WB, l u g a l m u n u - t u k , "a
king without name," is clearly a copyist's note that a name has been obliterated by a
lacuna in this place. The same statement, only differently worded, is presented by Su3+4:
[ti -] $> - a n - d e" , which apparently reflects a broken original ti - *K b a ] - a n - d e - [ a ] ,
"somebody who has disappeared" (cf. n. 178 above, referring to another example of use of
this phrase by Su3+4 to indicate an old break).
Completely different in meaning is, however, the text presented b y L i : l u g a l n u -
u b -1 u k , "(the Gutian horde) did not have a king," which states that the Gutians
constituted a disorganized horde with no central authority. That this variant is secondary
is shown by the fact that it is peculiar to a single text in the B branch of the tradition, where-
as the other concept, that a name has been broken away, is found in both the B branch
(Su3+4) and the A branch (WB) and so must be older. It is also proved, however, by the
text of the section itself. WB names 20 kings of the Gutian dynasty. The summary at the
end, however, counts 21 kings both in WB and in Li. This gives excellent sense if our line,
as in WB and Su3+4, is a note that there originally was a name here, making 21 names in
all, but is inexplicable if, as in Li, it is merely a piece of information concerning the organiza-
tion of the Gutians. The text of Li, 1 u g a 1 n u - u b - 1 u k , is in all probability a de-
velopment from that of WB. A slightly damaged version, l u g a l [mu] n u - t u k ,
could become l u g a l n u t u k , "had not a king," which it would be natural to refer
to "the Gutian horde" and to change into the grammatically more correct form l u g a l
n u - u b - t u k (collective).
A suggestion may be offered as to what name may originally have stood in this place
in the King List or in its source. Among the Nippur material in the University Museum
of the University of Pennsylvania, Hilprecht found many years ago a copy of a long in-
scription of a ruler e-er-ri-du-pi-zi-ir (var. en-ri-da-pi-zi-ir), who styled himself da-ntim
Mr gu-ti-im u ki-ib-ra-tim ar-ba-imt "the mighty one, king of Gutium and of the four quar-
ters" (Hilprecht, The Earliest Version of the Babylonian Deluge Story and The Temple Li-
brary of Nippur [Pennsylvania. University. Babylonian Expedition. Series D: Researches
and Treatises V 1 (Philadelphia, 1910)] chap, iv; cf. Poebel, PBS IV 1, p. 134). This title
is—except for the addition of "king of Gutium"—the title of Naram-Stn of Agade and thus
presents the Gutian as heir to Naram-Sfn's kingdom. Since, as we shall see later (pp. 205-7),
the available evidence indicates that the Gutian dynasty is to be reckoned from a conquest
which overthrew the reign of just that ruler, Naram-Sfn, the Gutian Erridupizir, who flaunts
Naram-Stn's titles in his inscription, is indeed very likely to be the anonymous Gutian ruler
who defeated him and thereby laid the foundation for Gutian supremacy.
oi.uchicago.edu
118 THE SUMBBIAN KING LIST
288 287
WB P<iLx G Su4 29 ^ i m - t a - a 1 u g a 1 - k m287 m u
288
3 i-a5
WB P< Li G Su< 30 2 8 9 i n - k i - § u § 2 8 9 m u 6290 (var. 7) i - a 6
2W
WB In G s a . r - l a g a b - l a - g a b 2 W mu292 6 i - a 8
293
WB L, G §u 1 - m e - e 2 9 3 (var. i a - a r - I a - g a -
a§) m u 6294 1-a*
WB G e - l u - l u - m e - e S mu296 6296 (var. 7)
i-a B
WB G i - n i - m a - b a - k i - e § m u 5 i-• a 6
WB G 35 i - g e4 - e §297 - a - u § m u 6298 i- • a g
299
WB U i a - a r -1 a - g a b299 m u 15 i-• a B
m
WB U i - b a -1 e300 mu 3 i-• as
301
WB Li i a - a r -1 a301 <- a n - g a b> m u 3 i-•a5
WB U ^ k u - r u - u m302 mu 1 i-• a B
WB 40 r b a l l * - b i l - k i ( ? ) , D 4 - i n m u 3 1-• a B
m
WB [ l a 4 r a - b u - u m 3 0 5 mu 2 i- • a B
WB i-ra-ru-um mu 2 i- • a B
WB ib-ra-nu-um mu 1 1-• a 5
2w-28« Thus Pi also (see photograph, PBS V, PI. XCI): i m - v$>WA - i m - t V - V .
La's i m - b i - a developed from i m - t a - a through a broken text: i m - p £ * r i - a ;
Sus+4 omits.
287-287Su3+«: [ l u g a l ] - a m ; P 4 broken; Li omits.
288
Thus P 4 also; Li: 5.
289-289p4: i n - k i - [ ] ; I*: i n - g i * - s t i § ; Sus+i*. [ ] T - b a . The concurrent tes-
timony of WB, P4, and Li, which represents both main branches of the tradition, must be
preferred to that of Sus+4. On the name cf. n. 301 below.
290
Li: 7. P4 does not seem to have given the reign of Inkishush; it has . . - d a - [ ]
in the line after the name. The traces in front of - d a show the heads of two vertical
wedges through which runs a horizontal line.
291-291 The fact that KIL has the value 1 a g a b suggests that the following 1 a - g a b was
originally a gloss which got down into the line; cf. the similar cases of Samug (n. 85 above)
and Enbi-Eshtar (n. 159 above). The original form of the line in WB's ancestors was there-
fore probably s a r -1 a g a bla~*ab (cf. p. 207). That the name is Sarlagab is also indicated
by Li, which according to the photograph (PBS XIII, photographic PL II) has #F -
a r - l a - g a - b a , i.e., z a r - a r - l a - g a - b a (not w a - a r - l a - g a - b a as in Le-
grain's copy).
292
298-298LI: i a - a r - l a - g a - a l .
294
Li: m, which can be restored as i l (6).
296
298
G: 7. Which is correct?
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. vii 29-43 119
29 Imta3 became king and reigned 3 years;
30 Inkishush reigned 6 (var. 7) years;

Sarlagab reigned 6 years;
Shulme3 (var. Iarlagash) reigned 6 years;
Elulumesh reigned 6 (var. 7) years;
Inimabakesh reigned 5 years;

35 Igeshaush reigned 6 years;
Iarlagab reigned 15 years;
Ibate reigned 3 years;
Iarla<ngab> reigned 3 years;
Kurum reigned 1 year;
40 Habil-kln(?) reigned 3 years;
Laerabum(?) reigned 2 years;
Irarum reigned 2 years;
Ibranum reigned 1 year;
297
298
G: ¥ , which can be restored as H (6).
299-299LI: [ ] - g d b .
soo-sooLl. [ i - b ] a - t i .
301-301 L I : [ i a - a ] r - Q a1 - a n - g a b (see PBS XIII, photographic PI. II), which is
probably original. WB's i a - a r -1 a may represent a broken predecessor: i a - a r - l a [ -
a n - g a b ] . Dr. T. George Allen calls my attention to the similarity between this name
i a - a r - l a - a n - g a b and i a - a r - l a - g a b in 1.36 above, suggesting that these are
in reality only variant writings of one name i a r l a g a b spoken with nasalized §. For
the existence of a nasalized g in Gutian speaks also the—not quite as close—correspond-
ence between i n - k i - § u s , read perhaps i n - g e* - s" u § (1. 30), and i - g e 4 - e § - a - u §
(1. 35), which may represent one name, i g e s 6 § .
302-302 L i : [ J-bi.
303
Thus according to the photograph, which shows the heads of the upper two vertical
wedges of this sign.
304
Read thus instead of - d i ?
305-305 rpo be identified with the Gutian ruler whose inscription was published by Winckler
in ZA IV (1889) 406. The first line, containing the name, is damaged: °mmwM^. It
was restored by Hilprecht in BE I 1 (1893) p. 13 as La~si(l)-ra(?)-<ib(?), and this restora-
tion has been generally accepted. Dr. Gelb of the Oriental Institute, however, calls our
attention to the facts that the traces which Hilprecht reads as $i(l) look more like x& and
that—while a name Lasirab is otherwise unknown—a name La~e-ra-ab is attested by tablets
of the Agade period from Tell Asmar (As. 31 :T. 30 ix 10 and 29 and As. 31: T.la fragment
4 rev.). We have therefore adopted the reading Laerab as more probable than Lasirab.
oi.uchicago.edu
WB 44 b a - a b - l u m mu 2 1 - a6
WB 45 p t i z u r - d s f n d u m u lja-ab-lum
WB mu 7 i-a5
WB [td(?)]- a r - l a - g a - a n - d a305a mu
7 i-a6
WB [s i] - ru41306 mu 7 i - a5
WB Lx [fc-n-g]a u4 40 l 307 -a 5
WB(P2)L, r 308
Su3 50 2">l l u g a l
WBCP,)^ Su3 [ m u - b i 91]309 u4 40 310
ib-a 5 310
CoL viii
311
WB Li Su3 J k i - s u - l u - u b 4 - g a r3U g [u - t u312 -
' * u mki]
Lx Su3 J la < g i § tukul b a - a n - s i g > 3 1 3
WB L, Su3 J n a m - l u g a l - b i u n u gki - s e [ b a -
tiim]
WB J unuki-ga d
utu-fr6-g[al lugal -
k m]314
315
WB J m u 7 su-si 7 u4 [. . 1 - a5]
(emend to m u 7 i t u 6 u4 15 i - a5)315
WB J 5 1 [lugal]
305a xhis king Iarlaganda is probably to be identified with i a - a r - l a - g a - a n ,
king of Gutium, mentioned in YOS I, No. 13:1-2. The form there given would seem to
represent a shortened form " T a r l a g a n d .
306 @f - r U 4 i < The reconstruction of the name as [si]- 'm1 is suggested by the fact that
a Gutian king by the name Spurn (written s i - u - u m ) is known (Scheil in CR, 1911, pp.
318-37; cf. Poebel, PBS IV 1, pp. 134 f.). The short form of the name met with in WB
may be due to a break in an ancestor of this ms. which swallowed up the last sign: s i -
U4 [- u m ] . This break might also be responsible for the short form of the immediately fol-
lowing name, Tiriga instead of Tirigan.
307
308
mi i.e., W<r. Li and the final summary of P 2 also have 21. This figure is correct,
since the list originally had a name before Imta 3 ; see n. 285 above. A certain mu-a-ti-gu-bi-
$i4n, mentioned as "his king" by a scribe from the city san-rig^ (SAG-PA-KAB-DU) near
Umma (Thureau-Dangin in RA IX [1912] 74-76) is usually considered one of the Gutian
kings. The name—formed with Muati, a name of Nabu of Borsippa—seems, however,
to be compatible with an Akkadian etymology Muati-qu(b)bisin . . . . , "Muati (has heard)
their wail"; and, since no other indications point to his being a Gutian, it is equally possible
that he was merely a local petty king on a par with the kings of Uruk (Uruk IV) etc. in
that period.
309 Thus by addition of the single reigns in WB. The higher totals, 124 years and 40
days in I* and 125 years and 40 days in P2 (final summary), are probably due to misreading
of the final -^ (-es) in the name e - l u - l u - m e - e s or the name i - n i - m a - b a - k i - e s
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , COLS, vii 44—viii 5 121
44 Hablum reigned 2 years;

45 Puzur-Stn, son of Hablum,
reigned 7 years;
Iarlaganda(?) reigned 7 years;
SiDu(m)(?) reigned 7 years;

Tiriga(n) reigned 40 days.
50 21 kings
reigned its 91 years and 40 days.
Col. viii
The horde of Gutium
<was smitten with weapons;)

In Uruk Utu-hegal became king
and reigned years 7 x 6 0 + 7 , days. .

(emend to 7 years, 6 months, 15 days).
5 1 king
as part of the numeral for the reign. Such an error is very easy to make in Li and P2,
where the figures follow directly upon the names ( i - n i - m a - b a - k i - ^ f m u i - a*) ,
but not in WB, where the names are separated from the figures by the word m u (i - n i -
m a - b a - k i - 4 m u V 1 - a6) . Assuming such an error (35 instead of 5 for Inima-
bakesh) to have happened in Li and restoring the missing parts of this text from
WB, we do indeed get the total which it now has, 124 years and 40 days ( 5 + 7 + 6 + ^ +
[6 +35+6+15+3 - ) - 3 + 1 + 3 + 2 + 2 + 1 + 2 + 7 + 7 +7] years+40 days = 124 years and 40
days). P2, which belongs to the same main group of the tradition as Li, may be re-
stored in the same way except that it may have followed G in assigning 7 years rather
than 6 to Elulumesh. This would account for its total being one year higher than that of
Li. The total of Su3+4, 25 years, is likely to be the actual total of the individual reigns pre-
served in a damaged original.
310-310 LX o m i t s #
311-311 L I : k i - s u - l u - i i b - g a r ; S113+4: m a - d a .
312
J and Li: - t i - .
313
Erroneously omitted in WB; preserved in Li, Su3+4, and J.
314
According to the photograph (not Langdon's copy) the space fits this restoration.
J: [1 u g a l ] - a m .
315-315j: m u 7 j t u Q U4 [15 i - a 5 ] , "reigned 7 years, 6 months, and 15 days."
The text of WB, m u 7 s u - s i 7 u 4 [ ] , "years 7X60+7, days . . , " is clearly a cor-
rupt form of the text given by J. Note that the original figure for the months, 6, is pre-
served in the summary.
oi.uchicago.edu
,18
WB J 6 m u - b i 7 § u - § i 6 W [.,l-a B ]
(emend to mu-bi 7 i t u 6 m 15 l-a5)316
WB Su4 J unugki giS
tuku1 b a-an-sig
WB Su4 nam-lugal-bi u r [fki - § &] b a - 1 u m
WB Ps Su4 u r i ki - m a u r - [dn a m m u] 317 1 u g a l318
WB P6 Su4 10 mu 18 [l]-a 6
d319 d 320 d
WB P6 Su4 §ul-gi d u m u ur- nammu-ke4
WB P6 Su4 mu 46321 (emend to 48) i - a5
d 322
WB P6 SU4 b u r 7 - d s t n 3 2 3 d u m u d 324§ u 1 - g i - k e4325
WB P6 Su4 mu 9 326 i-a5
d d 327
WB P6 SU4 15 §u- sfn dumu b u r 7 - s i n 328
d
329
WB P5 SU! Su4 m u 9 i - a5
330
WB P* Sui SU4 J i - b f - ds f n dumu § u - d s i n - k e4
WB P 6 Sm Su4 J m u 24 3 3 1 \ - a5
332
WB P5 Sui Sin J 4 (emend to 5) 1 u g a 1
316-316Thus according to the photograph. J : [ m u - b i 7] i t u 6 U4 15 i - ' a s 1 .

The text of WB, m u - b i 7 l u - s i 6 'm 1 [ ] , " i t s y e a r s 7 X 6 0 + 6 , d a y s . . , " is clear-
ly a corrupt form of the text given by J.
317
P s adds - k e 4 •
318
Ps and S113+4 add - a m . On the omission of - a m in W B see p. 136.
319
Su3+4 also has the determinative; Pg omits it.
320
P 5 and Sua+4 omit.
321
Su3+4: 48; P§: 58. The origin of the variants is fairly clear. As will be noted, the
figures of Su3+4 and PB both show the same number of units, namely eight, while W B has
six. Since m (6) can easily have developed out of an original m (8) by passing through a
slightly broken text: fiW, it is probable t h a t W B ' s source had 8 units in its numeral
but lost two in the course of transmission. This assumption is strongly supported b y t h e
high dynasty total, 108 years, which W B gives. Since t h e present figures for individual
reigns add up to only 106 years, two years less than the total demands, it seems obvious
t h a t the total reflects an older, more correct form of the text in which the two units in the
reign of Shulgi had not yet been lost. We can therefore assume t h a t WB's original had 48
years for Shulgi, which would make it agree with Su3+4.
Comparing W B and Sus+4 with P 6 , we note t h a t P* places the sign for "year," m u ,
directly in front of the figures. Since this sign ends in < -wedges and since the difference
between P 5 's i&m (58) and the 4M (48) found in or behind Su8+4 and W B consists in the ad-
dition of just such a wedge, it seems obvious t h a t the variant in Ps is due to the confu-
sion of -<-wedges which texts of this type would present: ^m. A copyist could easily
come to read one of t h e wedges of m u as part of the numeral and thus get 58 instead
of 48.
The original figure of all three versions is thus ;gfflF (48). This agrees with other
evidence for the length of the reign of Shulgi, for we know 48 full date formulas for that ruler
(see Ungnad in RLA I I 140-43. TJngnad gives 49 formulas, b u t we have no indication t h a t
the third of his formulas, No. 21, "year when t h e n a m - r 1 g -mu of Shulgi was installed
and invested," belongs to Shulgi's reign; it m a y equally well be a formula from the time of
oi.uchicago.edu
CRITICAL EDITION OF THE TEXT, COL. viii 6-19 123
6 reigned its years 7x60+6, days . . (emend

to its 7 years, 6 months, 15 days).
its kingship to Ur was carried.
In Ur Ur-Nammu(k) (became) king
10 and reigned 18 years;
divine Shulgi, son of divine Ur-Nammu(k),
reigned 46 (emend to 48) years;
divine Bur-Sin, son of divine Shulgi,
reigned 9 years;
15 Shu-Sin, son of divine Btir-Sln,
reigned 9 years;
I(b)bl-Sln, son of Shu-Sin,
reigned 24 years,
4 (emend to 5) kings
I(b)bt-Sln). Cf. also Thureau-Dangin in RA VII (1909-10) 184-85, who suggests 48 years
as a possibility, and the discussions of the problem by Langdon (OECT II 20, n. 1) and
Ungnad (op. cit. p. 135). See now also G. G. Hackman, Temple Documents of the Third
Dynasty of Ur from Umma ("Babylonian Inscriptions in the Collection of James B. Nies"
V [London, 1937]) p. 4, who proposes 48 years for Shulgi on much the same grounds as
those we have given above.
322
Ps and Su3+4 omit the determinative.
323
P 6 adds - n a ; thus apparently Su3+4 also: - n [a].
324
Su3+4 also has the determinative; Ps omits it.
328
Thus P* also; Su3+4 probably omits.
326
Thus P6 also; Su3+4: 25. The text of WB and P6 is correct, as shown by existing date
lists; see Ungnad in RLA II 135 and 143-44 (Nos. 68-76).
327
P6 and Sum omit the determinative.
328
Thus Su3+4 also; P 8 adds - n a - k e 4 .
329
P 8 : 7; Su3+4: 16; Sui's <&M = 20+z. The figure of WB is proved correct by the
date lists; see Ungnad in RLA II 135 and 144-45 (Nos. 77-85).
330
Thus Ps and Su8+4 also; Sui adds the determinative for deity.
831
Thus J also; P6 and Sui: 25; Sua+4.' 15. In view of the close relationship between
Sui and Sus+4, it is probable that the -<W (15) of Su3+4 has developed from *1t (25) through
a broken text KK? intermediate between Su and Su3+4. How WB got its 24 is uncertain.
It may represent a damaged original 25; or, more likely, one of WB's ancestors may have
been written before the last year of I(b)bl-Sln and therefore have included only 24 years
for this ruler.
332
Thus, even though 5 kings are listed. Since the easiest way to count the rulers in
such a list is to count the numerals, which stand out from the other signs, it is possible that
the miscounting was due to the identical figures for Btir-Sln and Shu-Sin. J, P«, and Su3+4
all have 5; Su^s W can be restored as 5.
oi.uchicago.edu
WB P5 Sui Su4 J 20 mu-bi 108333 ib-aB

WB P6 SUi SU4 J u r fki 334 _ m a 335 g^t u k u 1 b a - a n - s i g336
WB Ps Sui Su4 J n a m - l u g a l - b i i-si-inki-§e b a - t u m
WB P5 Sui i - s i - i n k i - n a 337i s - b i - 3381 r - r a l u g a l339
WB P6 Sui mu 33 340 i - a6
d341
WB Ps Su t 25 §u-i-lf-§u dumu i§-bi-342lr-
r a - k e4
WB Ps Sui mu 20343 (emend to 10) i -a 6
344 d
WB Ps; P 6 Su, i-din- d a-gan dumu 345§ u .
-!-
lf-§u346
WB P2 P 6 U Su, mu 21 347
WB P2 P 6 L, Stti 348 d
i §-me- da-g[an d u m u 349i - d i n -
d 350
d a - g a n - k e4 ]
WB P2 P 6 Sui 30 mu [20351 1-•a5]
d352 353
WB P2 P 6 Su, l[i-pi-it-e§4-tdr dum.uu i iS-
m e - d d a - g a n - k e4]353
WB P2 P 6 Sui mu [ll 3 5 4 i-a5]
d d 355
WB P2 P 6 Sui ur-[ ninurta mu 28356 i] - a 5
d
WB P6 b u r - d s ! [n d u m u d u r - d n i n u r t] a - k e4
WB P6 35 mu '21 3 5 7 i-a5
d d
WB P6 li-pf-[it- ]en-Hl
333
This figure is right if the present figure for Shulgi is corrected to 48 (see n. 321 above).
P5 has 117, which agrees with its figures for individual reigns. Su3+4 has 123, one more than
the sum of its individual reigns. Sux has W/A (120+#), which should probably be restored to
n trm (123) to correspond to Su3+4.
334
Thus J, P6, and Sui also; Su3+4 omits (cf. n. 335).
335
P 5 and Sui omit; Su3+4 adds k i .
336
Su3+4 adds s u Ij u s rk e1 - [e n - g i - r a] m u - ru n1 - [s i r ] , "the foundation of
Sumer was torn out."
337
Thus P6 also; Sui adds the determinative for deity.
338
Thus Sui also; P5 adds the determinative for deity—a fact which is of interest for the
dating of that ms. (see p. 7, n. 12). For the divine name IR-RA both a reading Irra (i r -
r a) and a writing Era (e - r a) are attested. See Weidner in Archivfur Keilschriftforschung
II (1924/25) 17, n. 4; Dossin in RA XXI (1924) 183; Deimel, SL, No. 50.10 and the litera-
ture there quoted.
338a The name Ishbt-Irra would seem to represent the Akkadianized form of an Amorite
*Jasbt-Irra; cf.the name Ja§bt-ila in Bauer, Die Ostkanaander (Leipzig, 1926) p. 30.
^PsandSuiadd-am.
340
Thus Sui also; P 6 : 32.
341
Sui also has the determinative; P5 omits it.
342
P6 adds the determinative for deity.
oi.uchicago.edu
CRITICAL EDITION OF THE T E X T , C O L . viii 2 0 - 3 6 125

Ur was smitten with weapons;
its kingship to Isin was carried.
In Isin Ishbi-Irra 338a (became) king
25 divine Shu-ilishu, son of Ishbi-Irra,
reigned 20 (emend to 10) years;

I(d)din-Dagan, son of Shu-ilishu,
reigned 21 years;
Ishme-Dagan, son of I(d)din-Dagan,
divine Lipit-Eshtar, son of Ishme-Dagan,
reigned 11 years;
divine Ur-Ninurta(k) reigned 28 years;
divine Bur-Sin, son of divine Ur-Ninurta(k),
divine Lipit-Enlil,
343
P 6 : 10; Sui: 15. As seen by Langdon, OECTII20, n.; 9, the figure 20 in WB is proba-
bly secondary (dittography of the <-wedge). That WB derives from a text which had 10
here, as has P5, is indicated by WB's total, 203 years, which agrees with a reign of 10, not
20, years for Shu-ilishu.
344 Thus P5 also; Sui adds the determinative for deity.
345
Thus P5 also; P2 and Sui add the determinative for deity.
346
P5 adds - k e 4 .
347
Thus P2 and P5 also; Sui: 25.
348
Thus P5 also; P2 and Sui add the determinative for deity.
349
Thus P 6 ; P2 and Sui add the determinative for deity.
350
Thus P 5 ;Sui omits.
361
P 2 : 20 over erasure; P 5 : 20; Sui: «n.
352
Thus according to the photograph. P2 also uses the determinative for deity here;
P5 omits it.
353-353 Thus Ps and Sui; P 2 : d i - d i n - ^ d a - [ g a n ] , by dittography from preced-
ing line.
354
P2 andP 6 : 11.
355
P2 adds d u m u di s k u r - [ ] m u 'f b1 - [ ] rb a l1 (?) [ ].
356
P 6 : 28.
367
<«r, i.e., <NT; P6 also has 21.
oi.uchicago.edu
WB Ps 37 dumu bur-dstn-k e«SM m u 5 i - a5

d
WB P* 83M (var. 7) i - a , t m
l r - r a - i - m i - t i mu
d
WB Ps en-Ml-ba-ni mu 24 i-a 6
d
WB P* 40 z a - a m - b i - i a mu 3 i-ag
WB P6 d|36i - t e - e r - p i 4 - § a
mu 4 i-ag
d
WB P6 ur-du6-kil-ga mu 4 1-ag
d
WB P6 stn-ma-gir mu 11 l-a5
WB 13362 (emend to 14) 1 ug a l
WB 45 mu-bi 203363 f b-a8
8u nu-tir-dnin-§ubur
858
P6 omits.
869
P 6 : 7, probably developed from 8 through a broken text; or P6's text may descend
through an edition from the 7th year of this ruler.
860
P6 adds another line: V(?) - [d i n] - e s4(?) - r t a ri(?) i t u 6 [i n - a 5 ], "I(d)din-
Eshtar(?) reigned 6 months." The identification of this badly damaged name with
that of I(d)din-Eshtar, mentioned in the Tammuz hymn TRS I, No. 8:208, and perhaps in
Zimmern, Sumerische Kvltlieder aus altbabylonischer Zeit, 1. Reihe (1912) No. 26 vii 17, has
been proposed by Carl Frank, Kultlieder aus dem Ischtar-Tamuz-Kreis (Leipzig, 1939)
pp. 105 ff.
oi.uchicago.edu
CBITICAL EDITION OF THE T E X T , COL. viii 37-46 127
37 son of Bur-Sin, reigned 5 years;

divine Irra-imi(t)ti reigned 8 (var. 7) years;
divine Enlil-ban! reigned 24 years;
40 divine Zambia reigned 3 years;
divine Ite'r-pisha reigned 4 years;
divine Ur-Du(l)kuga(k) reigned 4 years;
divine Sin-magir reigned 11 years.
13 (emend to 14) kings
Belonging to Nur-Ninshubur.
381
362
Thus according to the photograph. Actually 14 kings are listed. We can offer no
satisfactory explanation of this error.
363
Thus according to the photograph (as in Langdon's transliteration!). Actually the
reigns listed total 213 year's. The sum 203 is correct if the figure for Shu-iiishu is emended
to 10 as suggested above in n. 343. P5, written later than WB, gives name and reign of
the successor of Stn-magir: d a - ' m ^ - i q - l - l l - S u d u m u d s t n - m a - g i r mu 23
[in- as], "Damiq-ilishu, son of Sin-magir, reigned 23 years." This king closes the 1st dynasty
of Isin. Pfc's total for the dynasty is [16] kings reigning 225 years and 6 months.
oi.uchicago.edu
IV
COMPOSITION
In the preceding sections we have tried to trace the main lines in that proc-
ess of tradition which separates the first edition of the King List from the late
copies of it which are all we have preserved. It is therefore natural to consider
next the problems which center around this first edition, the original of the
King List. When and where was it composed? What were its sources, and
how were they utilized by the author?
DATE
Most of our manuscripts of the King List were written during the second
half of the dynasty of Isin, and the currently accepted view is that the list—
or, since some scholars consider the manuscripts separate compilations, the
"lists"—was composed at that period.1 It has just been shown, however, that
all of our manuscripts must ultimately derive from a single original, and we
have seen that they contain a considerable number of variants due to misread-
ings and errors in copying. So many misreadings and errors would hardly
have developed had not a long period of tradition separated the copies from
the original, and it is therefore necessary to reconsider the question of when
the King List was composed.
Since the King List does not give any direct information as to its date, we
have to rely exclusively upon internal evidence—language, form, fundamental
ideas of the work, etc. It will be practical to consider first those characteristics
which help to establish a lower limit.
1
See e.g. Eduard Meyer, Die dltere Chronohgie Babyloniens, Assyriens und Agyptens (2.
Aufl.; Berlin, 1931) p. 38: "Wie weit es freilich den Gelehrten zu Ende des dritten Jahr-
tausends noch moglich war, den wahren geschichtlichen Zusammenhang einigermassen
festzuhalten oder wiederherzustellen, bleibt fraglich genug"; E. Unger in OLZ XXXVII
(1934) 363: "Man stiitzt sich . . . . mehr auf die sekundaren Konigslisten; diese sind aber
erst in semitisch-altbabyloniseher Zeit, urn 2100 v. Chr., abgefasst und verhalten sich stark
tendenzios"; O. E. Ravn, Babylomke og assyriske Kongers historiske Indskrifter (K^ben-
havn, 1934) p. 27: "Sumererne selv har nedlagt deres opfattelse af vor periode i dynastiske
lister, der kompileredes omkring 2000 f. Kr." Only Sidney Smith, EHA, p. 29, holds a dif-
ferent view. He assumes eorrectly that our present versions depend upon a single original
which must be of earlier date. He does not, however, suggest any definite period for this
original.
128
oi.uchicago.edu
COMPOSITION 129
THE COLLECTIVE i - b - LAL
In summing up the various dynasties five of our manuscripts make use of

the form i - b - LAL (i b - a5) for 3d pi. preterit active, transitive. Only two,
P 5 and S, have instead the form i - n - LAL - e § ( i n - a k - e s ) ; and a single
one, K, uses i - n - LAL - m e - e § (i n - as - m e - e S) . Examining the oc-
currence of these forms in texts outside the King List, we find that they be-
long to different stages in the history of the Sumerian language. The form
generally used in older texts is i - b - LAL,2 whereas i - n - LAL - e § , which
is rare in the older texts, becomes the standard form during the 3d dynasty
of Ur and keeps this position down to the Hammurabi period. Then it is in
its turn supplanted by i - n - LAL - m e - e s . 3
The point during the 3d dynasty of Ur at which i - b - LAL disappears
and i - n - LAL - e § takes its place can be fixed fairly exactly to the reign of
Shu-Sin, as will be seen from the list on page 130 of datable occurrences of the
two forms during the periods of the 3d dynasty of Ur and of Isin-Larsa. 4
As is shown by the list, i - b - LAL disappears during the reign of Shu-
Sin. Why, then, do we find this form in our manuscripts of the King List
written as late as the end of the Isin period, when i - n - LAL - e § was the
standard form? The answer cannot be doubtful: The original of the King List
must have been composed before the reign of Shu-Sin, when i - b - LAL was
still current. From the original this form had then been mechanically handed
down from copy to copy until it reached our manuscripts written in the Isin
period. Only in a few cases have copyists taken offense at the antiquated
form and modernized the text to agree with Isin-Larsa standards. The result
is the form i - n - LAL - e § which we find in P 5 and S. A still later redac-
tion produced the correction to i - n - LAL - m e - e s in K.
From the use of the form i - b - LAL we can thus conclude that the origi-
nal King List was composed earlier than the reign of Shu-Sin of Ur.
2
Cf. Poebel in AJSL L (1933/34) 156, n. 7.
»GSG § 456.
4
The list is based on the following publications: Giustino Boson, Tavoktte cuneiformi
sumere, degli archivi di Drehem e di Djoha, dell'ultima dinastia di Ur (Milano, 1936); Niko-
laus Schneider, Die Drehem- und Djoha-Urkunden der Strassburger Universitats- und Landes-
bibliothek (An. Or. I [Roma, 1931]) and Die Drehem- und Djofyatexte im Kloster Montserrat
(Barcelona) (An. Or. VII [Roma, 1932]); Edward Chiera, Legal and Administrative Documents
from Nippur (PBS VIII 1 [1914]) and Old Babylonian Contracts (PBS VIII 2 [1922]);
Alfred Pohl, Rechts- und Verwaltungsnrkunden der HI. Dynastie von Ur (TMH n.F. I—11
[1937]); F. Pelagaud, {fsd-iilla, textes juridiques de la seconde dynastie d'Our," Babyloniaca
III (1910) 81-132; H. de Genouillac, "Textes juridiques de l'epoque d'Ur," RA VIII (1911)
1-32; F. Thureau-Dangin, "Notes assyriologiques" XXI, RA X (1913) 93-97. It should
give a representative picture of the usage during the period in question.
oi.uchicago.edu
Shulgi
year 42 Captains of four ships "have wrecked" (the ships) f b - b i i l An. Or. I, No. 53:4
year 43 Captains of two ships "have wrecked" (the ships) i - f b - f e u l An. Or. I, No. 58:3
BOr-Sin I
year 1 Officers of the e n s i (k) "have received" §u-ba-ab-ti An. Or. VII, No. 125:22
year 9 Two persons "swore to it" nam-erfm-bi f b - k u5 ITT II, PL 17, No. 920:12-15
Shu-Sin
year 3 Workmen "have brought" im-ttim Pohl, op. tit. No. 311:29
year 4 1 u AL-SE- [e -] n e "have received" &u-ba-ab-ti Boson, op. tit. No. 309 obv. 3
Nine persons "that they have not received" §u-la-ba-ab-ti-a Pohl, op. tit. No. 271 ii 23
Two persons "that they have given" ba-an-si-mu-§a Pohl, op. tit. No. 271 iii 7
Two persons "because they said" mu bl-in-n6-ia- Babyloniaca III, PI. VIII, No. xvii rev. 1-2
The heirs of Dudu "swore" in-pa-de-e§ ITT III, No. 5279 iv 7
year 5 Two persons "have given him" in-na-ab-si Pohl, op. tit. No. 253:11
Two persons "have received" §u-b a-an-ti-e§ Pohl, op. tit. No. 253:13
year 8 Two persons "have received" §u-b a-ti-6& Pohl, op. tit. No. 24:6
I(b)bi-Sin
year 1 Two persons "have received" §u-ba-ti-e§ Pohl, op. tit. No. 34:6
year 2 Three persons "have received" §u-ba-ti-e§ Pohl, op. tit. No. 35:7
Three persons "have received" §u-b a-ti-6§ Pohl, op. tit. No. 36:6
Two persons "have received" §u-b a-1i-e § Pohl, op. tit No. 110:6
Three persons "have received" § u - b a - a n - t i-es" Pohl, op. tit. No. 109:7
Bur-Stn I I
Two persons "bought" i[n]-§i-in-sdm-es PBS VIII, No. 6:17
Two parties "swore" i n - p a-d e-e § PBS VIII, No. 6:21
Irra-imi(t)ti
Two parties "swore" in-pa-de-e§ P B S VIII, No. 20:20
Two parties "who bought" in-§&m-e§-a PBS VIII, No. 103:12
Enlil-bani
Two persons "swore" in-pa-de-es PBS VIII, No. 8 rev. 7
r
Two persons "received" § u 1 - b a - a n - r t i1 - e § PBS VIIIf No. 107:9
Damiq-ilishu
Two parties "swore" in-pa-de-e§ PBS VIII, No. 16:17 and 26
oi.uchicago.edu
COMPOSITION 131
THE DISREGARD OF SUBJECT ELEMENT - n - AND "CAUSATIVE" - n -4ft

The conclusion that the King List was composed before Shu-Sfn's reign
is corroborated by another feature of the list, the treatment of the subject ele-
ment - n - and the "causative" - n - in the verbs. Three manuscripts—P5,
S, and K—give the verb "he reigned" as i n - a 5 . The subject element of
the 3d singular, - n - , is here expressed in writing. All other manuscripts,5
however, ignore this - n - and write i - a5, "he reigned." A similar differ-
ence can be noted in the case of the "causative" - n - , which appears in
b a - a n - s i g , "it was smitten" (WB), b a - a n - k t i r , "it was changed"
(P5), b a - a n - t i i m , "it was carried" (P5), and elsewhere, whereas it is
ignored in the corresponding forms b a - s i g , "it was smitten" (Li, Sui,
Su3+4, and J), b a - k ti r , "it was changed" (S), and b a - 1 ti m , "it was
carried" (WB, P2, la, S, Sui, Su2, SU3+4, and J).
Examining the treatment of subject element - n - and "causative" - n -
in inscriptions other than the King List, we find again that the difference
corresponds to a change in usage which took place during the period of the
3d dynasty of Ur. The crucial point here is the reign of Bur-Sin I. Before
this ruler the "causative" - n - does not appear, and the subject element
- n - is limited to a single case, namely the form i n - d \X - a , "who built."5*
With Bur-Sin I, however, both the subject element - n - and the "causa-
tive" - n - suddenly appear in a number of verbs, and writings expressing
these elements soon become the established norm. The sudden change is well
illustrated by the accompanying table of relevant forms appearing in the his-
torical inscriptions of the 3d dynasty of Ur.
The forms in which the subject element - n - and the "causative" - n -
are not expressed belong to the time before Bur-Sfn I. When we find such
forms in our manuscripts of the King List as late as the end of the Isin dy-
nasty, where they do not belong, we must therefore conclude that they have
been inherited from an original composed when such forms were normal,
that is, before the reign of Bur-Sin.
4a
We keep—for practical reasons—the term "causative - n - " introduced by Poebel
in GSG §§ 521 ff. That it actually is an accusative element has been suggested by Poebel,
ibid. § 526. On its identity with the subject element (subject-accusative element) see Rudolf
Scholtz, Die Struktur der Sumerischen engeren Verbalprdfixe (Inaugural-Dissertation; Bres-
lau, 1931) p. 44.
5
On a few isolated cases of i n - a& in Li see pp. 54 f.
5a
Note also the isolated instance m a - a n - d u u , "he commanded me" (Gudea Cyl.
A ii 15). This form does not show the subject element elsewhere in Gudea's inscriptions.
From a still earlier time there is the exceptional 114 a n - d u KA a n - g £ 1 (Stele of Vul-
tures xvii 5f.; var. m a - d u u* - a - g d 1, ibid. rev. v 28f.) discussed by Poebel in ZA
n.F. II (1925) 3 f.
oi.uchicago.edu
WITHOUT - n • WITH - n •
UR-NAMMU(K)
mu-ba-al "he dug" SAK, p. 188 i i
13; t/JSTI, No. 446
i 4, 8,12, and 16
mu-na-ba-al "he dug for him" UET I, Nos. 42:9,
45:10,46:11
in-du-a "who built ,; SAKf pp. 186 a 4
and g 5, 188 %
i 9 ; UET I, Nos.
33:4 and 50:10
mu-na-du "hebuHtforhim" SAK, pp. 186 b ii 2
and 4, c 12, d ii 4,
and e 8 and 188 h
10, k 10, I 8, and
m rev. 1; UET I,
Nos. 35:8, 36:10,
38:8, 39:7, 40:7,
CO
to 41a 8, 46:8, 47:9
pa5b mu-na-e "he executed UET I, No. 50:11
splendidly for
him"
mu-na-gi4 "he brought back SAK, p . 186 c 13
for him"
mu-ni-gi4 "he brought back *7#!ri, No. 50:14
into it"
mu-na-gub "he planted" UET I, No. 41a 6
KA b f - g i - i n "he confirmed" SAK, p . 188 % i 18
a-mu-na-ru "he presented to UET I, Nos. 32:8,
him" 34:8,48:12,49:9,
52:10
mu-na-tab "he added for SAK, p . 186 6 ii 4
him" (var.)
6b
That PA in this phrase is to be read as p a is indicated by the writing d s u l - p e - e for usual ds u 1 - PA - e in a seal in-
scription published by Scheil in RA XIII (1916) 12, No. 5, and PL II, No. 15. The form s u l p e ' e has naturally developed
from h u 1 p aD e through assimilation of a to the following e.
oi.uchicago.edu
WITHOUT - n - WITH - n -
SHULGI
mu-na-dim "he constructed SAK, p. 194 x 12
for him" and z 9
mu-du "he built" SAK, p. 190 b 7
mu-du-a "who built" SAK, p. 190/rev. 3
mu-na-du "he b u i l t for SAKt p. 192 n rev. 4
him"
mu-na-gu "he brought back SAK, pp. 190 c 8 and
for him" 192 n rev. 2
mu-na-gi-in "he confirmed for SAK, p. 194 t 5;
him" VET \, No. 287:8
a-mu-na-ru "he presented to SAK, pp. 192 q 6,
him" 194 y ii 3, and 196
a'iiGand/'iiS
BUR-SIN I
mu-na-du 'he b u i l t for SAK, pp. 198 c 13 m u - n a - a n - d u "he built for SAK, p. 200 g 17
him" and e 22 and 28 him"
and 200 h 9 and i
rev. 7; UET I,
Nos. 67:8 and
71:26
mi m u - n a - "he prepared for UET I, No. 71:29 pa m u - n a - a n - S "he executed UET I, No. 71:27
n i - d un him" splendidly
for him"
a-mu-na-ru "he presented to UET I, No. 67:10 ki mu-na- "he founded SAK, p. 198 / ii 6
him" an-turn for him"
mu-na-ni-tu "he brought into SAK, p. 198 e24 mu-un-na-ni- "he brought in- SAK, p. 200 i rev.
it for him" in-tu to it for him" 10
SHU-SIN
mu-du "he built" UET I, No. 72:10
mu-du-a "(when) he YOS I, No. 20:23;
built" CT XXXII, No.
103354:23
oi.uchicago.edu
WITHOUT - n - WITH - n -
SHU-SIN—Continued
mu«na-du "he built for SAK, p. 202 d 17; mu-na-an-du "he built for SAK, p. 202 c 16;
him" VET I, Nos. 80: him" OIP XLIII,
17 and 81:14; CT bldg. inscr. No.
XXXII, Nos. 1:18
103353(=PJ3S
XV, No. 43): 13
and 103354:30;
YOS I, No. 20:30;
HRETA, No. 11:
12
bf-in-gi4-a "(when) he FO£I,No.20:26;
turned (some- CT XXXII,
thing) back" No. 103354:26
in-p& "he envisaged" SAK, pp. 200 6 7
and c 5 and 202
d 9; VET I,
Nos. 72:5,80:9,
81:5; CT
XXXII, No.
103354:15;
YOS I, No. 20:
15; HRETA,
No. 11:5
I(B)BI-SIN
in-na-ba "he assigned to VET I, Nos. 88:10, m i - n i - i n - d f b - "who had VET I, No. 289:
him" 96:9, 97:9; PBS ba-a seized" 16 and 45
XIII, No. 5 ii 4
a-mu-na-ru "he presented to VET I, No. 289:29 b a - s i - i n - DU VET I, No. 290:5
him" and 64 b f - i n - g i4 VET I, No. 289
14 and 43
mu-un-gam "he bent" VET I, No. 289
15 and 44
ba-an-tu-ra- "because he VET I, No. 289
k e4 - e s had brought 47
him (into
greatness)"
oi.uchicago.edu
COMPOSITION 135
The evidence from the subject element - n - and the "causative" - n -

thus corroborates and amplifies that from the collective i-b-LAL. The
latter indicated that the original was older than Shu-Sin; now we see that
the original must be older than Shu-Sin's father and predecessor, Bur-Sin.
TRACES OF EARLY REDACTIONS
The dating thus far obtained—earlier than Bur-Sin and Shu-Sin—is con-
firmed and our terminus ante quern moves still higher up when we consider
the evidence furnished by traces of earlier redactions which can be found in
the manuscripts. As we have mentioned before, the concluding sections in our
sources tend to show a certain unevenness of style. The originals which the
scribes used would naturally as a rule be somewhat older than the scribes' own
time, so that it was necessary for them to add a certain amount of material
to their copies to bring the record up to date. In these additions which the
scribes themselves had to compose and which they could not just copy sign by
sign from the original their own stylistic peculiarities are likely to show. By
paying attention to the sudden appearance of such peculiarities one is often
able to identify an addition and thereby to determine where an earlier version
ended. A single example will serve to illustrate this.
In P 5 the subject element - e is used regularly whenever a name is fol-
lowed by a patronymic: N. d u m u P. - k e4 (i.e., P. - (a) k - e) . Toward
the very end of the list, however, from [ l i - p i - i t - d e n - l i l ] d u [ m u
b] u r - ds i n (!) (iv 15) to the last king, d a - m [ i ] - i q - l - l l - S u d u m u
d
s i n - m a - g i r (!) (iv 23), this - k e 4 is omitted (see p. 32, n. 65). The
explanation of the change is clear. A scribe who did not himself use this -.k e4
has followed an original where it was regularly employed; copying sign by
sign he mechanically took over its - k e 4 , but when his original stopped
and he had to write an additional section to bring the list up to date, he was
no longer directly influenced by the original and his own stylistic habits nat-
urally asserted themselves. Therefore we do not find - k e4 used in this sec-
tion. Now if we can identify the section from Lipit-Enlil on as an addition
to bring the list up to date, the older version from which the scribe copied
must have ended with Bur-Sin of Isin, and the natural inference is that it was
written under that ruler.
Observations similar to those here made in P 5 can be made in WB also.
But while P5 tells us only what we already know, that copies of the King List
were being made at the time of Bur-Sin II of Isin, so that the original must
be still older, WB adds important new information. We have discussed earlier
an interesting change in the formula for introducing new dynasties observable
oi.uchicago.edu
in WB (p. 67). The crucial points of this discussion may be repeated and slight-
ly elaborated here. In the main part of this text new dynasties are introduced
by the formula A ki (-a) N. l u g a l - ^ m , "In A N. was (or 'became')
king." The last two dynasties, however, the 3d dynasty of Ur and the dy-
nasty of Isin, are introduced differently: Aki(- a) N. 1 u g a 1, "In A N.
(was) king." The identifying verb -km, "was," is here omitted, and
the nouns to be identified, N. and 1 u g a 1, are merely placed side by side.
To assume that the author of the list, who has carefully written 1 u g a 1 - k m
in dynasty after dynasty, should suddenly and for no reason have changed
his literary style only two dynasties before the end is obviously absurd. The
change in style must indicate a different hand, and we can therefore conclude
that a later copyist has added this section to bring his copy up to date.
Further proof of this conclusion is to be found in the relation of the section
in question to the first part of WB, the antediluvian section. We have seen
above (pp. 55-64) that a variety of indications proves that the antediluvian
section is secondary. Now the scribe who inserted that section had certain
stylistic peculiarities which crop up every time he is not merely copying but
adds on his own. One such peculiarity is his preference for constructions such
as A ki (-a) n a m - l u g a l l - a (written n a m - l u g a l - l a ) , "In A
(was) the kingship," and Aki(- a) N. 1 u g a 1, "In A N. (was) king," in
which the two nouns to be identified are merely placed side by side and the
identifying verb - k m , "was," is omitted. Comparing our change in for-
mula at the end of the list, we see that it consists of the introduction of this
very peculiarity. The man who added the antediluvian section is also re-
sponsible for the last part of the list; his literary peculiarities appear in both
places.
That the closing section of WB was added later can thus be considered
certain. Not only do wefindhere a change of style which shows that a different
han.d is at work, but the character of the change itself links the section as
closely as possible with a known interpolation, the antediluvian section. But
if we can thus detach the section from the beginning of the 3d dynasty of Ur
as a later addition, we have left an original version which came to an end with
Utu-hegal of Uruk and which can therefore be assigned to the reign of that
ruler.
The evidence which we have thus far considered has carried the date at
which the King List can have been composed a considerable way back from
the time to which our present copies belong, the end of the Isin period. The
language used in the list shows that the author must have written before the
middle of the 3d dynasty of Ur, before Shu-Sin and Bur-Sin. Now we have
oi.uchicago.edu
COMPOSITION 137
found evidence of a version from the reign of Utu-hegal, who preceded that
dynasty. Is this at last the original version, or must we seek still farther back?
The answer depends upon whether other features in the King List indicate
this period as the probable time of origin or make a still earlier date likely.
gi5
THE TERM t uk u1 . . . .sig
We have already earlier mentioned that the formula A ki g i i t u k u 1
b a - ( a n - ) s i g . . . . , "the city A was smitten with weapons . . . . , " dom-
inates both of the principal branches of the tradition. This means that it dates
back to a point in the tradition at which these two branches had not yet sep-
arated. We have, of course, no way of proving that this point is identical with
the original; but, inasmuch as our versions are fairly numerous and come
from widely separated parts of Babylonia, it seems reasonable to assume that
a point where this tradition was still but a single stream cannot have been
far from the source. Serious corrections of language and style are accordingly
not likely to have been introduced.
Examining the history of the term gi§ t u k u 1 . . . . s 1 g in Sumerian lit-
erature, we find that the older texts, from E-Anna(k)-tum to Sargon of
Agade, do not use this term at all. The idea "to defeat" is at this time almost
exclusively expressed by GIN-KAR . . . . s i . 6 With Sargon of Agade the first
traces of our phrase begin to appear; I d A ki - a (k) - d a g i § t u k u l e -
d a - s i g , "he fought with the man of the city A," is a common expres-
sion in Sargon's inscriptions. 7 It will be noted that gi§ t u k u 1 . . . . s i g
does not yet serve to express the idea "to defeat," which is still G!N-KAR . . . .
s i ;8 but, construed with - d a , g i § t u k u l . . . . s i g is a term for "to
fight," "to come to blows with." 9
Th6 meaning "to defeat" which gi§ t u k u 1 . . . . s i g has in the King
List occurs fairly late in other inscriptions. It appears first in the inscrip-
tions of Gudea: giH u k u l u r u a n - § a - a n e l a m k i m u - s i g , "he
smote with weapons (i.e., defeated) the cities of Anshan and Elam." 10 After
that, in the period of Isin and Larsa, it is frequent in date formulas from
Larsa, for example Gungunum, year 19: k i - s u - l u - t i b - g a r m & -
6
E-Anna(k)-tum: GfN-KAR b i - s i (e.g. SAK, pp. 18 vii 3 and ix 2; 20 iii 14, 20, 24;
22 vi 20, vii 2, iii 13 and 19; and 24 iv 13 and 15); En-temena: GfN-KAR i - n i - s 1 0SL4iv,
p. 38 iii 14); En-e(n)tar-zi(d): GfN-KAR b i - s i (RA VI [1904-7] 139, AO 4238 obv. iii 4).
7
E.g. Poebel, PBS IV 1, pp. 173-74 i 16-19, i 36-40, and passim.
8
E.g. PBS V, No. 34 obv. i 20-21, i 41-42, and passim: GfN-KAR e - n i - s i .
9
This difference in usage was first pointed out by Thureau-Dangin, La chronologie des
dynasties de Sumer et d'Accad, p. 18.
10
Gudea Statue B (SAK, p. 70) vi 64.
oi.uchicago.edu
a l - g i 4 - a * i § t u k u l b a - a n - s i [ g ] (UET I, No. 265 obv. 13-14; RLA

II 154 b and 156, No. 112); Abl-sare(?): m u ki-su-lu-ub4-gar
1 - s i - i n k i - n a KiSt u k u l b a - a [ n - s l g ] (Thureau-Dangin, La chrono-
logie ... , p. 53, 1. 42; RLA II 151, 1. 42, and 157, No. 129); Sumu-El: m u
k [ i - s ] u - l u - u b 4 - g a r k a - z a l - l u k i gi§ t u k u l ba-an-[slg]
(RLA I I 1 5 1 , 1 . 49, and 157, No. 135; cf. UET I, No. 249); Warad-Sfn, year 2:
\i <ki!->su-lu-\ib-gar [. . . . - t] i - b a - a 1 rs a1 l a r s a m<ki>
iS
« t u k u l b a - s i g (UET I, No. 266 i 14; RLA II 154 a and 160, No. 192).
Since g i § t u k u l . . . . s i g in the meaning "to defeat" thus makes its
first appearance under Gudea and since the King List uses it in this sense,
the term suggests that the King List is to be dated to that ruler or later. Now
we have seen that the lower limit for the date of the King List can be placed
in the reign of Utu-hegal. Gudea, who lived at the end of the Gutian period,
seems to belong to the generation of, or immediately before, Utu-hegal. 10a
We arrive therefore at approximately the time of this ruler as the most likely
period for the composition of the King List.
BASIC IDEAS IN THE KING LIST

The evidence from the term gi§ t u k u l . . . . s i g , pointing to the time
of Utu-hegal as that in which the King List originated, does not stand alone.
Comparison of the King List with the great inscription of Utu-hegal in which
he relates his victory over Gutium 11 shows striking and significant similarities
10a
Reasons for placing Gudea "during the latter years and after the fall of the Gutian
dynasty" have been summarized by Gadd in The Early Dynasties of Sumer and Akkad, p. 32.
To them may be added the following facts: (1) The similarity between the language of
Gudea's inscriptions and that of Ur-Nammu(k)'s and Shulgi's is extremely close. Note e.g.
such a special feature as the omission of the subject element - n - everywhere except in
one form of the verb d u , "to build" ( i n - d u - a , "who built"), where it regularly
occurs. (2) Gudea's wide-reaching trade expeditions would hardly have been possible un-
der the Gutians. As we know from the Utu-hegal inscription (RA IX [1912] 111-20 and X
[1913] 98-100), Gutium had closed the vital trade route passing from the Mediterranean
across Assyria down through Akkad and Sumer to the Persian Gulf: "Both banks of the
Tigris he (i.e., Tirigan) had seized, southward in Sumer he had barred the paths, northward
he had barred the routes, on the highways of the land long grass he had let grow." This
route was reopened by Utu-hegal's victory over Gutium, and it can be only to that event
that Gudea is referring when he says (Statue B v 21-27): "When he (i.e., Gudea) built the
temple of Ningirsu(k) did Ningirsu(k), his beloved king, open up the road from the upper
to the lower sea."
11
RA IX 111-20 and X 98-100. The authenticity of the Utu-hegal inscription, which is
known only from later copies, has recently been questioned by Giiterbock (ZA n.F. VIII
[1934] 14), who is of the opinion that the disposition and style of the text preclude its being
a historical inscription: "Der Verfasser des Textes hat zwar den offiziellen Titel Utu-fcegals
gekannt, der Text ist aber nach Aufbau und Stil offenbar ein literarisches Werk, keine Konigs-
inschrift; von einer solchen ist nicht einmal die aussere Form ubernommen." This view we
oi.uchicago.edu
COMPOSITION 139
in ideology and language. The author of the King List worked, as has often
been mentioned, on the theory that Babylonia was and always had been a
single kingdom. Within the country the capital could change from one city
to another, but there was never more than one king at a time. The ruler of a
city or a province could become king only by defeating the existing capital
cannot share. The Utu-hegal inscription gives a remarkably detailed account of a campaign,
but its disposition (Aufbau) is in no way different from the disposition of accounts of cam-
paigns found in other royal inscriptions. We may compare e.g. the account which En-temena
gives (SAK, pp. 36-38 n i 13-ii 26) of the wars between Lagash and Umma in the time
of Ush:
En-temena Inscription Utu-hegal Inscription
Misdeeds of enemy (i 13-21) Misdeeds of enemy (i 1-14)
Divine orders to fight him (i 22-27) Divine orders to fight him (i 15-23)
Preparations for decisive battle (ii 16-iv 3):
o) Utu-hegal seeks support of his city
god, Inanna(k).
6) Itinerary of march to battlefield: de-
parture from Uruk; purpose of campaign
made known to troops in E-Ishkur(ak); Gu-
tian messengers taken captive in Bar-Ilitab-
ba(k); arrival in Muru; Utu-hegal seeks sup-
port of Ishkur of Muru.
Victorious battle (i 28-31) Victorious battle (iv 4-8)
Results of victory: true boundary Results of victory: capture of Tirigan and
re-established; tribute to be paid by re-establishing of autonomy of Sumer (iv 9
enemy (i 32—ii 26) to end)
As will be seen, the disposition of both narratives is practically the same: (1) misdeeds
of enemy, (2) divine orders to fight, (3) victorious battle, (4) results of victory. That Utu-
hegal, who had to lead his troops far afield to come to grips with the enemy, tells also of the
events which took place during that part of the campaign, whereas the En-temena account
goes directly to the decisive battle, is only natural; for the battle in the En-temena account
took place near Lagash itself, and little could have happened on the way to the battlefield.
Nor does the style of the Utu-hegal inscription differ in any way from that of other his-
torical inscriptions. It is very vivid and uses direct speech a great deal; but a vigorous char-
acterization of the enemy, Gutium, as "a viper of the hills, enemy of the gods, who had car-
ried the kingship of Sumer off to the mountains, who had filled Sumer with evil," etc. is not
much different from En-temena's description of II of Umma, "II, e n s i (k) of Umma, a
robber of fields and lands, always speaking evil," etc., which appears in the inscription just
mentioned (SAK, p. 40 iv 19-23).
There is thus neither in disposition nor in style any basis for separating the Utu-hegal
inscription from other royal inscriptions. Its authenticity is indicated also by its language;
for, although the scribal habits of later copyists show in the use of the subject element - n -
and the "causative" - n - , such features as the use of the old collective i - b - LAL (e.g.
iii 7-8: u r u - n i l u - d i l i - g i m e g i r - r a - n i b a - a b - t i s , "his city as one man
followed behind him") and the use of the phrase GIN-KAR . . . . s i to express the defeat
of the Gutian forces are rarely found in inscriptions of the 3d dynasty of Ur, which followed
Utu-hegal, or in still later periods. Lastly we may call attention to the highly peculiar and
irregular ruling of tablet AO 6018 (RA IX 112 f.) which contains our copy. Cases contain-
ing one line of writing alternate with cases which have two, three, and even twenty-two
lines. This irregularity is not suggested by the content of the text and is indeed explicable
only if it represents the original ruling of a text written between and around the figures on
oi.uchicago.edu
140 THE SUMEEIAN KING LIST
and carrying (b a - 1 ii m) the "kingship" to his own residence. Now the

same view meets us in the Utu-hegal inscription. Like the King List, Utu-hegal
knows only one "kingship," that which has been "carried" (b a -1 ti m)
to the mountains by Gutium. It is only by defeating Gutium that Utu-hegal
can bring it back into the hands of Sumer. Even the characteristic phraseol-
ogy, n a m - l u g a l construed with t ii m , "to carry," is common to
both texts, a point especially significant, as this construction occurs no-
where else.
To the evidence of the phrase Ki§t u k u l . . . . s i g , the ideology of the
King List, and the wording of that ideology is added the fact that the pe-
riod of Utu-hegal in itself constitutes the logical period for such a work as the
King List to have been conceived and carried out. After long subjection under
foreign barbarians Utu-hegal had in the decisive battle near Muru crushed
the Gutian forces completely, restored the old independence, and "brought
back the kingship of Sumer into its own hands."12 This new-won freedom must
have powerfully stimulated Sumerian national feeling and must have awakened
interest in Sumer's glorious past, in the history of that "kingship" which had
now come back again. It would thus naturally inspire the production of such
a work as the King List, which is an effort to trace and present just that
history.
CONCLUSIONS
We have thus seen first of all that the language of the King List shows that
it cannot have been written later than the middle of the 3d dynasty of Ur,
while traces of earlier versions move this limit back to the reign of Utu-hegal,
to whom the earliest traceable version can be dated. Against a still earlier
date, before Utu-hegal, speaks the use of the expression gi§t u k u 1. . . . s i g ,
which appears in the King List as a term for "to defeat" but is not known to
us in this meaning before Gudea, who belongs to the generation of, or im-
mediately before, Utu-hegal. The same conclusion is indicated by a consider-
ation of the ideas on "kingship" underlying the King List. These ideas are
the very same which we meet in Utu-hegaPs own inscription and they are
there expressed in the same characteristic phraseology as in the King List, a
a relief, where the division into cases and columns would necessarily be irregular. That the
Utu-hegal inscription originally was written on a relief is already indicated by the closing
lines of col. iv, which obviously have reference to an accompanying pictorial representation
of a type similar to the relief of Anu-banini (G. Contenau, Manuel d'arMologie orientate II
'[Paris, 1931] 764, Fig. 541); cf. also a seal impression of Ilushu-ilia of Eshnunna(k) (OIP
XLIII, Fig. 100).
™RA IX 113 iv 36.
oi.uchicago.edu
COMPOSITION 141
phraseology met with nowhere else. Finally, we have seen that the time of
Utu-hegal, to which all our evidence points, indeed provides the best imagi-
nable background politically and in ideology for such a work as the King List
to be conceived; for pride in new independence and in the "kingship" which
had been brought back must have furnished a mighty impulse to set forth the
history of this "kingship," that is, to compose our King List. There can thus
be no serious doubt that the composition of the King List should be dated
to the reign of Utu-hegal.
PLACE
While a number of indications in the King List point to when it was written,
very little can be said about where it originated. It is hardly possible to do
more than make a plausible guess.
Since it is obviously easiest to get hold of sources for the history and older
rulers of a city in that city itself, we can perhaps conclude that the city in
which the author of the King List lived must have been one of the three which
are dealt with in greatest detail in his work, namely Kish, Uruk, and Ur.
The second of these, Uruk, was the city of Utu-hegal; and, since the inspiration
for the King List is most probably to be sought in the revival of Sumerian na-
tional pride under this ruler, it seems most natural to look for the origin of
the list where this feeling must have been strongest, in the capital itself from
which the war of independence started. If we must guess, Uruk seems the
most likely place of the three.
SOURCES
GENERAL CHARACTER OF THE SOURCES
An inquiry into the problem of what sources the author of the King List
can have used for his work must start from an examination of that work itself.
By denning the types of material represented and by comparing these with
Sumero-Akkadian literature as a whole, we should be able to determine which
genres within that literature contain such material and can therefore have
supplied it to our author.
The material found in the King List can in general be described as of a
historical character. It divides readily into two distinctive types: (a) infor-
mation concerning names of kings and the number of years each of them
reigned and (b) "notes" added to the names of certain kings giving details
concerning themselves and their exploits. We may consider first the material
found in the "notes."
oi.uchicago.edu
142 THE SUMEEIAN KING LIST
SOUBCES OF THE " N O T E S "
The "notes" appearing in the King List are: 18

a) e - t a - n a s i p a lti an-§& b a - e n - d d lti k u r - k u r m u - u n - g i - n a
l u g a l - & m (ii 16-19)
"Etana, a shepherd, the one who to heaven ascended, the one who consolidated
all lands, became king."
6) e n - m e - e n - b i r a - g e - s i lti m a - d a e l a m k i - m a g i i t u k u 1 - b i
f b - t a - a n - g u r 1 u g a 1 - a m (ii 35-38)
"En-me(n)-barage-si, the one who carried away as spoil the weapons of the land of
Elam, became king."
c) m e s - k i - a * g - g a - § e - e r d u m u d u t u e n 4 m 1 u g a 1 - a m m u 324
1 - a5 m e s - k i - & g - g a - § e - e r ab-ba ba-an-tu ûr-sag-Se
b a - en (ill 1-6)
"Mes-kiag-gasher, son of Utu, became high priest and king and reigned 324 years.
Mes-kiag-gasher went into the sea and came out (from it) to the mountains."
cQen-me-kdr dumu mes-ki-&g-ga-§e-er lugal unuki-ga
ki
lti u n u g m u - u n - d u - a (var.: m u - u n - d a - d u - a ) l u g a l - a m (iii
7-10)
"En-me(r)-kar, son of Mes-kiag-gasher, king of Uruk, the one who built Uruk (var.:
'the one under whom Uruk was built'), became king."
e) d l u g a l - b a n - d a s i p a (iii 12)
"The divine Lugal-banda, a shepherd."
/) dd u m u - z i SU-PES* u r u ki - n i k ue - aki (iii 14-15)
"The divine Dumu-zi(d), a . . ..—his city (was) Ku3a(ra)."
g) d gi 1 g a m e s (dGi&-BiL-GA-MEs) a b - b a - n i H1-1& e n k u l - a b - b a - k e 4
(iii 17-19)
"The divine Gilgames—his father (was) a KJM-demon—a high priest of Kullab."
h) fMEsK?)-^ s i m u g (iii 31)
"MES(?)-H£, a smith."
i) m & - g a 1 - g a 1 - rl a1 [m & -1 a hi (iv 24)
"Magalgalla, a skipper."
j) k u - d b a - b ae m M t i - k u r u n - n a muio s u h u § k i § iki m u - u n - g i - n a
l u g a l - a m (v 36-39)
"Ku(g)-Baba, a barmaid, the one who consolidated the foundation of Kish, be-
came 'king.' "
k) n a - a n - n i - i a z a d i m (vil9)
"Nannia, a stonecutter."
18
For details concerning text and translation see chap. iii.
oi.uchicago.edu
COMPOSITION 143
1) g a r - r u - k i - i n . . - b a - n i n u - g i r i i 2 q a - S u - d u g u r - d z a - b a 4 - b u
l u g a l a - g a - de k i 1 ti a - g a - d e k i m u - u n - d u - a (var.: m u - u n - d a -
d u - a ) l u g a i - a m (vi 31-36)
"Sargon—his . . . . was a date-grower—cupbearer of Ur-Zababa(k), king of Agade,
the one who built Agade (var.: 'the one under whom Agade was built'), became
king."
A glance at these "notes" will show that they correspond closely with one
another in form. Two examples will be sufficient:
d) E n - m e ( r ) - k a r d u m u M e s - k i a g - g a § e r
lugalUnuk-a(k) luUnuk mu-n-du-a
I) S a r r u ( m ) - k i n . . . . l u g a l A g a d e ( - k ) l u A g a d e m u - n - d u - a
and
gr) G i l g a m e s abb-ani lilla en Kulabb-ak-e
I) S a r r u ( m ) - k ! n . . . . b - a n i nugiri(b)lugalAgade(-k)
The correspondence is so pronounced that we can consider it certain that the
"notes" owe their form to the same hand. This hand, further, must have been
that of the author, for they have—with one exception—all been carefully
worked into the general framework of the list.14
In content also the "notes" show close similarity. The information given
in them is of uniform character and can be classified without difficulty:
A. Information concerning origin of ruler
1. Parentage: Gilgames (g), Sargon (I)
2. Native city: Dumu-zi(d) (/)
3. Profession: Etana (a), Lugal-banda (e), Dumu-zi(d) (/), Gilgames (g), MES(?)-
HE (h), Magalgalla (t), Ku(g)-Baba (j), Nannia (k), Sargon (I)
B. Information concerning achievements of ruler
1. Political and military: Etana (a), En-me(n)-barage-si (6), Ku(g)-Baba (j)
2. Founding of capital: En-me(r)-kar (d), Sargon (I)
3. Spectacular adventures: Etana (a), Mes-kiag-gasher (c).
The nature of this material is clear; it is historical information. Comparing
it with the different genres of Sumero-Akkadian literature to see where we
14
The "notes" are as a general rule worked into the formula for introducing new dy-
nasties, Aki(- a) N. l u g a l - a m , a s a relative sentence with 1 ti inserted after N.
The exception mentioned is the "note" to Mes-kiag-gasher, which is added loosely after
the lines which deal with this ruler. On analogy from the other "notes" we should expect
to find
6 - a n - n a - k a m e s - k i - d g - g a - § e - e r d u m u d u t u 1 ti a b - b a b a -
a n - t u k u r " s a g " s e b a - e n e n - a m l u g a l - a m mu 324 I - &&.
In all probability, therefore, this "note" is secondary, inserted by some early copyist of the
list. This insertion must have happened before the A branch and the B branch of the tradi-
tion separated, for the "note" appears in both (in WB and in P2).
oi.uchicago.edu
should seek its sources, we can therefore at once dismiss all genres which rarely
or never contain historical information, such as "private letters," "contracts,"
"administrative documents," and "religious texts".; and as the historical in-
formation which our material gives is of a very special type we can limit the
field even more. Royal inscriptions have at all times been written for the pur-
pose of glorifying the king, and they never contain facts which could detract
even slightly from his dignity. Nor do state papers, treaties with other rulers,
correspondence with officials, etc. ever refer to such facts. It is therefore clear
that the author of the King List cannot have got his information concerning
the lowly birth and original menial occupation of certain rulers from such
sources.15 Furthermore, the legendary character of many statements in his
"notes," for example that Etana ascended to heaven and that Gilgames was
son of a demon, is definitely against a derivation from royal inscriptions or
from the documents in official archives. Since practically every "note" con-
tains information of one of the two types just mentioned, we can with cer-
tainty conclude that the sources used for this part of the King List were
neither royal inscriptions nor state papers, even though these genres carry
historical material.
The objections which can be raised against a derivation from royal inscrip-
tions and from material in official archives hold good for another historical
genre also, namely date lists. From somewhat before the Agade period to the
end of the 1st dynasty of Babylon the years in Babylonia were each named
from some important event of political or religious nature. In such formulas,
however, and in the lists in which they were collected, one would seek as vainly
as in the royal inscriptions for information that a king had started his career
as a shepherd or that his father was a demon. In the periods earlier than the
Agade dynasty and during part of that dynasty as well, other systems of dating
were in use, and we shall presently see how far it is possible to deduce the exist-
ence and probable content of date lists based on them. These lists, however,
must likewise have been compiled for purely practical purposes, and informa-
tion of mythical or anecdotal nature is not likely to have occurred in them.
The eliminations which can thus be made in the "historical" genres of
Sumero-Akkadian literature leave only a single group of texts, the "epic-his-
torical" genre, as a possibility. Examining the material contained in texts of
this genre we do indeed find information of just the type for which we have
been looking. We may mention some examples. The information in the
"notes" concerning lowly birth and original menial occupation of some
15
This argument against derivation from royal inscriptions was first propounded by
Sidney Smith, EHA, p. 29.
oi.uchicago.edu
COMPOSITION 145
rulers can be paralleled from among the legends by the so-called "Birth Legend
of Sargon,"16 which tells how this ruler was found by a date-grower and raised
to become a date-grower himself, and from the stories of the chronicles by
that of Enlil-bant of Isin, who likewise was a date-grower before he became
king.17 The city from which Dumu-zi(d) hailed is mentioned in "note"/- Such
information is also given in the "Birth Legend," which mentions Sargon's na-
tive city, Azupiranu.18 The chronicles which we have abound in information
concerning such political and military achievements as those mentioned in
"notes" a, 6, and h. The Sar tamfyari epic19 and the epic text dealing with
Gilgames' war on Kish20 also have such material. Lastly we may mention the
journey through the darkness and across the waters of death in the Gilgames
epic as a parallel to the spectacular adventures of "notes" a and c.
That the epic-historical genre contains the same kind of material as the
"notes" of the King List cannot be doubted. In many cases the correspond-
ence even goes beyond similarity in kind, and we find the actual material of
the "notes" themselves. Thus the information in the Etana epic that Etana
was carried up to heaven on the back of an eagle is obviously behind the state-
ment in the "note" to Etana: "the one who to heaven ascended"; and the
"note" to Sargon, "his . . . . was a date-grower," is merely a brief reference
to the narrative in the "Birth Legend" of how Sargon was picked up from the
river by the date-grower Aqqi. It therefore seems obvious that we should look
for the source of the "notes" among such texts.
But is that possible? The genre in which we are interested is at present
only represented by fairly late specimens. The earliest copy of an epic yet
found dates from the Isin-Larsa period, and such is the case with the legends
also. The "Birth Legend" is even preserved in a copy as late as Assyrian times,
and from Assyrian and Neo-Babylonian times date all our versions of the
chronicles. The King List, on the other hand, must, if our earlier deductions
16
CT XIII (1901) PL 42; L. W. King, Chronicles concerning Early Babylonian Kings II
87-96.
17
King, op. cit. pp. 12 and 117.
" CT XIII, PI. 42:3. Another Sargon legend, AO 7673 (H. de Genouillae, TRS II, No.
73), also mentioned Sargon's city (obv. 10').
19
See Weidner, Der Zug Sargons von Akkad nach Kleinasien ("Boghazkoi-Studien" Heft 6
[Leipzig, 1922]); Albright, "The Epic of the King of Battle/' JSOR VII (1923) 1-20. A
synopsis of the contents with valuable notes and further literature was given by Giiterbock
in ZA n.F. VIII (1934) 86-91.
20
PBS X 2, No. 5; SEM, No. 29; SRT, No. 38; Fish in John Rylands Library (Man-
chester), Bulletin XIX (1935) 362-72. The text was edited by Witzel in Orientalia n.s.
V (1936) 331-46.
oi.uchicago.edu
are correct, have been composed as early as the time of Utu-hegal. We would
therefore have to assume that our genre—although known in late copies only—
actually is of the same age as, or still older than, the King List.
In view of the extreme conservatism of Sumero-Akkadian literature such
an assumption is indeed very probable. It becomes practically certain when
we look at the material of the genre itself. As is commonly known, stories and
anecdotes about prominent historical persons originate while these persons
and their contributions to history are still fresh in memory and occupy people's
thoughts. So it is with our material. People cannot suddenly have begun to
tell stories and anecdotes about Etana, Lugal-banda, Gilgames, Ku(g)-Baba,
and Sargon when these personages had long ago been forgotten by all but a
few learned scribes; these stories must first have been told when the leading
characters were still familiar to the listeners and held their interest.21
The material of our genre must therefore be old; and when we consider that
narratives of mythological content were committed to writing as early as pre-
Sargonic times22 and that a period of considerable literary activity and de-
veloped literary ability must have immediately preceded Utu-hegal, under
whom the King List was composed, because only such a period could have
produced the famous cylinders of Gudea, there is indeed no reason whatever
to doubt that written versions of epics and legends existed at the time of our
author. Nor can we exclude the possibility that some stories and anecdotes
had already then been collected and arranged chronologically, just as such
material is later found combined in the Chronicle Weidner;23 for the writing
of the King List itself is testimony to the fact that this period had an interest
in the collecting and systematizing of historical material. We must therefore
reckon with the existence of at least three varieties of the epic-historical genre
—"epics," "legends," and "chronicles"—when our author wrote, and we can
accordingly assume that he derived the material for his "notes" from them.
In exactly what form he found this material cannot be determined with
certainty. He may have been a man well read in such texts and have written
21
The author plans to publish an article on the Lugal-banda epic, with detailed discus-
sion of this point.
22
E.g. the very important pre-Sargonic mythological text published by G. A. Barton,
Miscellaneous Babylonian Inscriptions (New Haven, Conn., 1918) No. 1.
23
Published by Giiterbock in ZA n.F. VIII 47 ff. I am indebted to Dr. Weidner, its
discoverer, for a transliteration and translation which enabled me to use this text before its
publication. The redaction which we possess must have been made in circles which had close
connection with E-sag-ila and Babylon, for all the anecdotes point a moral, namely how
dangerous it is to neglect that temple (ibid. pp. 15 f.). It is of course unlikely that the pre-
cursors of this text, with which wre reckon above, served as vehicles for such a tendency.
oi.uchicago.edu
COMPOSITION 147
his "notes" from memory; or he may have known only a single such source, an
early collection of anecdotes such as we have just described, and have got all
his information out of that. This question is, however, of only minor impor-
tance and may well be left open.
SOURCES OF THE MATERIAL WHICH CONSTITUTES THE MAIN BODY OF THE LIST
Having thus found that the "notes" are most likely derived from texts of
the epic-historical genre, we may turn to the material which constitutes the
main body of the list, the names and lengths of reign of the various kings.
Looking as we did with the "notes" through the various genres of Sumero-
Akkadian literature, we find that we can this time discard from the very be-
ginning all but a single genre, the date lists, for in none of the others do we find
that most important feature, the lengths of the reigns.24 This feature, further-
more, the editor of the King List must necessarily have got from his source,
for thefiguresare so obviously based on tradition that they could not possibly
have bieen invented.25 Before we can accept the obvious conclusion that the
author of the King List took his material from date lists we must, however,
pause for a moment to consider that the oldest date lists yet known do not go
back farther than to the time of the 3d dynasty of Ur, while the material con-
tained in the King List covers a span of time reaching far beyond the dynasty
of Agade into pre-Sargonic periods. Is it possible that date lists for these pe-
riods can have existed? If so, how should we then imagine their form and
content? Can they have met the requirements for a source for the King List?
These questions make it necessary to examine the Sumerian systems of dating
in older times.
The well known later dating system according to which each year was
named from some important event of a political or religious nature presup-
poses in itself the existence of lists in which the "year names" were arranged
24
Sidney Smith {EHA, p. 29) suggested official records of omens kept in the temples as a
possible source, but the "historical" omens which we possess never give information con-
cerning length of reign of the king in question or his relation to other kings. Many of the
"historical" data found in the omen literature can, moreover, be shown to be secondary and
to be derived from chronicles (see Guterbock in ZA n.F. VIII 16 f.). Such a source for the
King List is therefore not probable. Meyer (Die dltere Chronologie . . . . , p. 37) assumes
that historical inscriptions formed part of the sources for the King List—a view shared by
other scholars (e.g. Guterbock, op. cit. pp. 6 f.); but these also could never give information
concerning the order of the rulers and the lengths of their reigns. See also p. 156, n. 46.
25
We are considering the figures as a whole. That certain types of figures, e.g. the exces-
sive reigns given to some of the early rulers, reigns which always come to round numbers,
are fictitious is obvious.
oi.uchicago.edu
148 T H E SUMEKIAN KING LIST
chronologically, for by themselves the individual "names" give little indica-

tion of where the years in question belong. Date lists based on this system
must thus have existed as long as the system itself. Since the name of the
ruler belongs to the full form of every "year name," and since in any case the
accession of a new ruler was always important enough to provide the "name"
of the corresponding year, lists based on this system contain all the material
necessary to find out order, names, and lengths of reign of the various rulers
in the period covered by them and can thus have supplied the author of the
King List with the necessary material. Unfortunately, however, this system
of dating cannot be traced very far back in time. It does not seem to have
come into general use until the time of the Agade dynasty, and the earliest
examples of "year names" are from the time of En-shakush-Anna(k), 26 who
seems to have lived a generation or two before Sargon. It thus covers only a
small part of the period dealt with in the King List.
The system here mentioned, however, is not the only one known to the
Sumerians. Along with it in the period around the beginning of the Agade
dynasty runs a different system which does not use notable events as its base
but dates by reference to the various e n s i ( k ) ' s or other similar high
dignitaries who followed each other in office. And this other system can be
shown to be of very great age indeed. As far back in time as in the texts
from Fara there appear—as Thureau-Dangin was the first to point out—dat-
ings by reference to the names of some high officials, probably local rulers.
These datings take the form b a l a N., "reign of N . , " and examples of
them have been collected by Thureau-Dangin 27 and by Deimel.28
From approximately the same time as the Fara texts, if not earlier, come
26
R. Pohl, Vorsargonische und Sargo?dsche Wirtschaftstexte (TMH V [Leipzig, 1935]) No.
158, has a date of this type from the reign of En-shakush-Anna(k). Cf. Thureau-Dangin in
RA XXXIII (1936) 62.
27
SAKf p. 224.
28
Die Inschriften von Fara I (WVDOG XL [Leipzig, 1922]) 3. Deimel's objections to
considering the b a l a N. phrases as datings are of no value. That these phrases appear
only on contracts of sale is very natural, for there it was clearly more important to have a
date than on ordinary administrative documents which were meant to serve for only a short
time. A parallel observation can be made in the documents from the Isin-Larsa period,
where contracts are generally carefully dated while administrative documents are quite
often left undated. Deimel's own explanation of b a l a N., that it represents "die Quit-
tung, welche man doch auf einer derartigen Urkunde auch erwartet," cannot be right.
The name which appears in the b a l a N. formula is different from that of any party to
the transaction, and the similarity between b a l a N. and X Y - r a / d a d u b - b i
e - d a - b a l g i i - n a e - n i - g a r in the Uru-KA-gina texts upon which that explana-
tion is based is purely superficial. The latter must mean "X turned its (the transaction's)
tablet over to Y and put it on his account" and has nothing to do with b a l a N.
oi.uchicago.edu
COMPOSITION 149
two dedication inscriptions of Me-silim, king of Kish, which we may quote

in full:
me-silim l u g a l k i § i 6-du dnin-gir-su dnin-gfr-su mu-tum
l u g a l - § a g - e n g u r e n s i 1 a g a § u29
"Me-silim, king of Kish, builder of the temple of Ningirsu(k), brought this to Nin-
girsu(k). Lugal-shag-engur (was) e n s i ( k ) of Lagash."
m e - s i l i m l u g a l k i § i 6 - § a r g a - m u - g i 4 6 nf'e n s i a d a b a 3 0
"(I) Me-silim, king of Kish, verily restored E-shar. E (was) e n s i (k) of Adab."
Here too we find dating by reference to the reign of a local official, for the brief
mentions of the men who then held office as e n s i (k) 's in Lagash and
Adab, unconnected with everything else in the inscriptions, could hardly have
another purpose. Exactly similar to these datings in the Me-silim inscriptions,
but later and a little more explicit, are datings found in inscriptions of En-
temena and Lugal-anda:
iu-ba e n - t e - m e - n a ensi laga§uki-kam e n - e - t a r - z i s a n g u
d
n i n - g i r - s u - k a - k a m31
"At that time En-temena was e n s i (k) of Lagash and En-e(n)tar-zi(d) was
s a n g u -priest of Ningirsu(k)."
U4-ba d u - d u s a n g u n i n - g l r - s u - k a - k a m 3 2
"At that time Dudu was s a n g u -priest of Ningirsu(k)."
U4-ba u r - 6 - M U S - k e 4 d i - b i i - k u s l u g a l - a n - d a e n s i - k a m 3 3
"At that time did Ur-E-MU§(k) judge their case, and Lugal-anda was e n s i (k) ."
From these examples it is evident that in pre-Sargonic times, from Me-silim
and the Fara texts to Lugal-anda, documents could be dated simply by ref-
erence to the e n s i (k) or to some other high dignitary who at that time
held office.34 There are, however, traces of a more developed system also,
29
SAK, p. 160.
30
D. D. Luckenbill, Inscriptions from Adab (01P XIV [1930]) No. 5. Our translation of
g a - m u - g i 4 reckons with influence from Akkadian, where the corresponding lu utir can
be either a precative or an emphatic form. Cf. however Poebel, GSG §§ 640 and 640a.
3i RTC, No. 16; cf. SAK, p. 224.
32
Dec., p. xlvii, No. 5:21-22 (cf. SAK, p. 34 h), and p. xlviii, No. 7 viii 8-9 (cf. SAK,
p. 36 k).
33
RTC, No. 28; cf. SAK, p. 224, No. 2 6.
34
This rough dating method persisted until after the Agade period. Cf. e.g. H. F. Lutz,
Sumerian Temple Records of the Late Ur Dynasty ("University of California Publications
in Semitic Philology" IX 2 [Berkeley, Calif., 1928]) pp. 204-5, No. 83 (a text of the Agade
period) iii 39-46: u4 - b a um-mi-mi fq a - § u - d u8 l-nu-mi in(EN)-6tVt<?-d&a-m-t3 isak-ki
iS-nun*1, "At that time Ummimi was cupbearer, in the days of Inpiq-Hanish, ishakku of
Eshnunna(k)" (the name Inpiq-Hanish seems to mean "he heeded Hani," inpiq being Ii of
oi.uchicago.edu
namely dating by the years of the e n s i (k) 's. The evidence for this more
developed form of the system is furnished chiefly by the archives of En-
temena and his successors in Lagash, in which a number of tablets have at
the end a figure of special type which indicates the year in the e n s i (k) 's
reign from which the tablet in question dates.35 Still more exact are datings
on other pre-Sargonic tablets which at the end refer to a year and a month,
for example 4 m u i t u 4, "fourth year, fourth month," or to year, month,
and day: 5 mu 10-LAL-1 i t u 23 u4.36 These dates also must have ref-
erence to the reign of the e n s i (k) or to a similar fixed period.
We see thus that one could date on this system rather roughly by referring
merely to the name of the man who was then e n s i (k) but also, at least
in later times, quite accurately by quoting the year or even year, month, and
day of a given reign. Now it is obvious that even the rougher of these datings,
that by the name of the e n s i (k) only, presupposes lists in which one
could look up the time when a given e n s i (k) held office. Otherwise a dat-
nuppuqu; on the value in of EN see Gelb in AJSL LIII [1936/37] 41); Clay, YOS I, No.
13:1-2: i u - b a i a - a r - l a - g a - a n l u g a l g u - t i - u r a - k a m , "At that time Iar-
lagan was king of Gutium"; Scheil in CR, 1911, p. 319, 11. 14-15: u4 - b a s i - u - u m
l u g a l g u - t i - u m - k a m , "At that time Si3um was king of Gutium"; Hilprecht An-
niversary Volume (Leipzig, 1909) p. 152: i-nu-mi za-H-ku-um isak, "When Zarikum was
ishakku."
36
On this method of dating see Allotte de la Fuye in RA VI (1904-7) 106 f., who was
the first to realize the true nature of the marks in question. His explanation has been ac-
cepted by practically all scholars. Landsberger's objection (OLZ XXXIV [1931] 118, n. 1)
that the figures we find are so low that it seems unlikely that they are dates does not carry
much weight, for from synchronisms contained in the tablets themselves we know that the
reigns of the rulers concerned must have been very short (note, e.g., that En-e(n)tar-zi(d)
occurs as priest already under En-temena [SAK> p. 224, No. 2 a] and that Uru-KA-gina ap-
pears as an official under Lugal-anda [Smith, EHA, p. 39]. Note also that a Dudu, presu-
mably the same man, appears as s a n g u under En-temena, Bar-namtarra [wife of
Lugal-anda], and Uru-KA-gina [see Christian in AOF VIII (1932/33) 207; cf. C. F. Jean,
La religion sumirienne (Paris, 1931) p. 201]). It should also be mentioned that for En-
temena, who is likely to have had a long reign, we have a date as high as the nineteenth year.
Compare, finally, DeimeFs statement in his Sumerische Tempelwirtschaft zur Zeit Uruka-
ginas und seiner Vorgdnger (An. Or. II [1931]) p. 72: "Das sorgfaltige Studium des gesamten
Tempelpersonals, wie es uns in den nach obiger Regel chronologisch nach Jahr und Monat
geordneten Listen in seinem Bestande und in seiner zeitlichen Entwicklung vorliegt, hat es
mir bis zur Evidenz erwiesen, dass die von A. de la Fuije[!] zuerst aufgestellte Erklarung der
am Schlusse der Tafel angebrachten senkrechten Keilchen, die auf ein wagerechtes aufgereiht
werden, unumstosslich sicher ist."
86
Dates of this type were first noted by Thureau-Dangin in RA VIII (1911) 154. The
tablets concerned seem to come from Umma. Other tablets with such dates were published
by T. Fish, Catalogue of Sumerian Tablets in the John Rylands Library (Manchester, 1932)
p. xi, Nos. 2-17. It is not always easy to decide whether the tablets in question belong to
the first half of the Agade period or to the time immediately before that period.
oi.uchicago.edu
COMPOSITION 151
ing by his name would of course be altogether without value. There must
therefore have been lists which gave the names of the e n s i (k) *s in chrono-
logical order and the span of time each of them had reigned. These lists,
furthermore, must have been different for each city, for the local character of
the system is apparent in the datings which we have. The names of the officials
in the Fara dates plainly show that we are dealing with people from Shurup-
pak itself.37 The Me-silim inscription from Lagash is dated by reference to the
e n s i (k) of Lagash; that from Adab is dated by reference to the e n s i (k)
of Adab. Finally, the Lagash tablets from the archives of En-temena and his
successors are all dated from the reigns of e n s i (k) 's of Lagash itself, ir-
respective of what overlords Lagash acknowledged at the time.
We can thus answer our original question and answer it in the affirmative:
It is probable that date lists existed in pre-Sargonic times even as far back
as Me-silim and the Fara texts. These lists must have been local lists, differ-
ent for each city, and they must have enumerated the local e n s i (k) 's—
in some cases perhaps other officials also—with the number of years each
e n s i (k) reigned. In other words, they contained exactly what is required
in the sources of the King List.
We have thus, as with the "notes" of the King List, been able to isolate
a single literary genre as the only one which contains the right kind of ma-
terial and have seen that suitable texts of this genre must have existed and
can have been accessible to the author of the King List when he wrote his
work. With the "notes" we could get no further and had to leave the argu-
ment at that. Here, however, where our material is more extensive, it is pos-
sible to approach the problem from a different angle and thus obtain a check
on the result.
If, as our evidence has thus far seemed to suggest, the King List is a com-
pilation from a number of separate local date lists, the chances are that traces
of such original entities would still remain within the present whole. If we
are mistaken, however, and the source is not a collection of separate lists, such
traces should not appear. But they do. Looking through the list with this
problem in mind, we are struck first by the fact that the 1st dynasty of Uruk,
which now appears well within the list, begins with a ruler who is said to be
son of the sun-god Utu. This beginning is of absolute character. The Uruk
kings traced their lineage back to the sun-god, and originally there cannot
have been anything beyond that. We must therefore conclude that this tradi-
tion was originally separate and that its present place in the list after the 1st
37
E.g. m a s - ds \1 d and d
s t i d - K A - z i - d a , composed with the name of the city
god of Shuruppak, ds u d .
oi.uchicago.edu
dynasty of Kish is secondary and is due to the fact that the editor has com-
piled originally separate traditions.
A similar observation can be made in the 1st dynasty of Kish. Here Etana
occurs in the middle of the dynasty. From the Etana epic, however, we know
a tradition according to which Etana was not only the first king of Kish but
the first king of all.38 Now we find in the King List clear traces that two sepa-
rate units have been joined at just this point. Thus it should be noted that
the kings preceding Etana all have Akkadian names and are further unified
by the fact that many of their names are those of animals, while Etana begins
a series of rulers with names of different type. Further, Etana is introduced
by the formula E t a n a . . . . l i i g a l - & m , the formula characteristic
of the first king in a dynasty. It is therefore clear that the author knew that
Etana introduced something new; the "animal-kings" and the "Etana-kings"
form two separate units. It is obvious that the author of the Etana epic can
have known only the second of these, the one beginning with Etana, for ac-
cording to him there were no kings before that ruler. The second unit must
therefore have existed as a separate entity, and we may accordingly conclude
that the other unit, the "animal-kings," was added by the author of the
King List in front of an originally independent tradition which began with
Etana.
We have thus found evidence of at least two originally independent tradi-
tions which have been embodied in the text of the King List. The most con-
vincing proof, however, that the list represents a conglomerate of several sepa-
rate traditions is furnished by the figures given for the reigns. It has often
been pointed out that the reigns of the kings form a valuable criterion by
which to judge the degree of historicity to be ascribed to various sections of
the King List.39 Where the lengths of the reigns do not exceed what is humanly
possible we may assume that the author still had reliable historical tradition
to go by; where the kings have legendary reigns of several hundred years each,
however, we must be beyond those periods of which the Sumerians preserved
precise historical memories. Since the King List reaches back from historical
to legendary periods and since the criterion just mentioned gives us a means
38
See especially the Old Babylonian version (A. T. Clay, Babylonian Records in the Li-
brary of J. Pierpont Morgan IV [New Haven, Conn., 1923] PL 2 i 1-14), which clearly
states at the beginning of the epic that the gods had not yet appointed a king (1.6: sar-ra-am
la is-ku-nu ka-lu ni-si e-M-a-tim, "A king they had not appointed for all the . . . . people)
and that the insignia of kingship still rested in heaven before An (11. 11-12: fya-at-fii-um
me-a-nu-um ku-ub-hum it U-H-ir-ru ku-ud-mi-is a-ni-4/m ina ha-ma-i Sa-ak-nu).
39
Cf. e.g. Smith, EH A, pp. 35 f., and, still clearer, Ravn in an article in the Copenhagen
newspaper Kjbenhavn, Dec. 31, 1924.
oi.uchicago.edu
COMPOSITION 153
to distinguish between historical and legendary reigns, we should, if the King

List actually did constitute a single homogeneous tradition, be able to ascer-
tain fairly exactly at what point or zone the transition from historical to legend-
ary takes place.40 Looking through the list, however, we find that such a point
or zone does not seem to exist at all. Historical and legendary dynasties are
mixed. A dynasty will begin with legendary reigns, grow historical, and be
followed by another historical dynasty. After that, however, will come a third
dynasty which is again purely legendary. Such obvious inconsistency can mean
only one thing: the King List is not a single tradition but has been compiled
from several independent ones. And if we dissolve the present list, singling
out the material in the way which seems most natural, namely according to
locality, we shall indeed see that all inconsistencies disappear. Instead of one
blurred picture we get several perfectly sharp ones:
KISH
1st dynasty 23 kings legendary reigns
2d dynasty 8 kings legendary reigns
3d dynasty 1 king legendary reign
Transition from legendary to historical
4th dynasty 7 kings historical reigns
URUK
1st dynasty first 5 kings legendary reigns
1st dynasty last 7 kings historical reigns
2d dynasty 1+x kings probably historical reigns
3d dynasty 1 king historical reign
4th dynasty 5 kings historical reigns
UR
1st dynasty 4 kings historical reigns
2d dynasty 4 kings historical reigns
As will be noted, each of these traditions shows a clear and definite point at
which the transition from legendary to historical reigns takes place, just as we
must expect from homogeneous traditions. The inconsistencies of the King
List taken as a whole thus become clear. The author has interpolated these
original and homogeneous traditions one into another; and, since the Kish
tradition becomes historical very late, the combined list shows that inconsist-
40
It should perhaps be expressly stated that the division historical-legendary applies to
the reigns only. A name can obviously be historical, even if later on a "legendary" reign
has been associated with it. See also p. 166.
oi.uchicago.edu
ent and uneven distribution of historical and mythical dynasties on which

we have commented.
We have thus found perfect agreement between our two lines of inquiry.
That the King List is based on independent local lists of rulers—the conclusion
which we reached by comparing its material with that of other genres and by
a consideration of what is known of the oldest Sumerian methods of dating—
is wholly confirmed by the examination of this work itself, which shows traces
of the beginnings of at least two separate traditions within the present list
and which exhibits inconsistencies in the distribution of legendary and histori-
cal reigns which are explicable only if it is a compilation of separate local
traditions.
Our conclusion is therefore imperative: The author of the King List used
as sources a collection of local date lists from a number of Babylonian cities.
INDIVIDUAL SOURCES
We have seen above (pp. 142-47) that the general character of the sources for
the ''notes" of the King List could be determined. Their material derives from
documents of the epic-historical genre. Within this group, however, derivation
from one or more chronicles, from individual epics, or from legends seemed
equally possible; and it is accordingly clear that efforts to reconstruct the in-
dividual documents from which the "notes" were taken could have no chance
of success. We must therefore turn to the other part of our material, that
which constitutes the main body of the list.
In discussing the illogical distribution of historical and legendary reigns in
the King List we have already shown in what way the original sources now
united in the King List can be singled out. Rearranging the dynasties accord-
ing to location, that is, the Kish dynasties together, the Uruk dynasties to-
gether, etc., we found that the inconsistencies exhibited by the list as a whole
disappeared and we got groups which conform to what should be expected
from homogeneous traditions. That this feature, the location of the dynasties,
must indeed be a reliable guide could also have been concluded from our gen-
eral knowledge of the character of thp sources. As we have seen, the sources
must have been pre-Sargonic date lists, and these were local lists, each dealing
with the rulers of a single city only.
By rearranging the dynasties along these lines we obtain a number of groups
each of which contains all the material on a given city which is found in the
King List. In these groups we may—if no other evidence points to still finer
divisions—see the original sources used by the author. One reservation should,
however, be made. The groups which we have singled out can naturally repre-
oi.uchicago.edu
COMPOSITION 155
sent the corresponding sources only to the extent to which the material of the
sources was actually embodied in the King List. Until we know the extent to
which the author used his sources we must reckon with the possibility that our
groups are incomplete and that there are lacunas of unknown length between
their dynasties.
The groups singled out are of two kinds, larger groups made up of several
dynasties and smaller groups which have only one dynasty. We may consider
the larger ones first. In this category there are three: (a) the Kish group, con-
sisting of thef 1st, 2d, 3d, and 4th dynasties of Kish, (6) the Uruk group, con-
sisting of the 1st, 2d, 3d, and 4th dynasties of Uruk, and (c) the Ur group,
which has only two dynasties, the 1st and 2d dynasties of Ur. Only one of these
calls for special comment, namely the Kish group. Here, as we have already
mentioned (p. 152), the 1st dynasty does not seem to be of a piece; it looks as
though a series of Akkadian rulers many of whom have animal names has
been added in front of an originally separate tradition which began with Etana.
Since we cannot assume that the same city had two widely different dynastic
traditions, we must consider only one of the two the genuine Kish tradition.
The choice between them is not difficult. The close connection of the Etana
tradition with Kish is evident not only from the separate testimony of the
Etana epic, where Etana occurs as king of Kish, but also from the fact that
one of the kings of his dynasty has a name composed with that of the city
itself, Melam-Kishi(k).
If, however, the Etana tradition is the genuine Kish tradition, we must ask
how the author could come to assign the other list to Kish and to make it
precede the genuine tradition. A possible explanation suggests itself when we
consider the importance which attached to the title "king of Kish" in ancient
Sumer. That title was so highly prized that a ruler of another city would
proudly adopt it if he was mighty enough to have the actual ruler of Kish
acknowledge him as overlord. Thus for example Mes-Anne-pada called him-
self "king of Kish" although his capital was Ur in the South. 41 Another ruler
from the South, E-Anna(k)-tum of Lagash, also took that title; 42 and the same
is the case with rulers of Agade such as Sargon, Rimush, and Man-ishtushu. 43
The connection of the title with Kish itself, therefore, need not always be
very close, and it is possible to imagine that the author of the King List can
41
See U 13607, published by Burrows in UE II 312 f.
42
This can be concluded from the passage SAK, p. 22 v 20~vi 5: £ - a n - n a - t u m
. . . . - r a dinanna-ke< nam-ensi-laga§uki-ta nam-lugal-kisiki
m u - n a - t a - s l , "to Eannadu Inanna gave the kingship of Ki§ from (i.e., growing out
from) the Isakkuship of Laga§" (Poebel's translation; see PBS IV 1, p. 129).
48
See the inscriptions of these rulers listed in RISA, pp. 100-136.
oi.uchicago.edu
have had a source subscribed, for example, x l u g a l k i § i , "x kings of

Kish," 44 although these rulers actually belonged to a dynasty of another city
and were merely acknowledged as suzerains of Kish. For such a source there
would be only one suitable place, namely before Etana; for we can be fairly
certain that the Etana tradition, the genuine Kish tradition, carried unbroken
down to periods in which the author of the King List was at home and where
he could find no room for such a dynasty. The connection of the kings before
Etana with Kish is thus open to doubt, and it seems better to leave them aside
and to count as genuine Kish tradition only the kings following that ruler,
that is to say, the last half of the 1st dynasty and all of the following dynasties.
To the second category, the smaller groups, belong (a) the dynasty of Ak-
shak, (o) the dynasty of Maeri, (c) the dynasty of Agade, (d) the dynasty of
Awan, (e) the dynasty of Hamazi, and (/) the dynasty of Adab. The first two
of these contain six rulers each, the third contains eleven rulers, and the fourth
three rulers. For each of them we can therefore assume a separate list from
the city in question as the source which the author of the King List used. The
last two, however, the Hamazi group and the Adab group, contain no more
than one ruler each. Since we cannot assume that the author had "lists" which
contained only one ruler, we must suppose that he got his information con-
cerning these two rulers from other channels. It seems possible, for example,
that they appeared as foreign conquerors in one of the other sources.45
Summing up, 46 we may thus present the following conspectus of sources:
I. A collection of date lists from various cities. Source of the material in the body of
the King List. This collection comprised some eight texts:
A. Larger traditions:
1. List from Kish containing 1st dynasty of Kish from Etana down, 2d, 3d,
and 4th dynasties of Kish
44
Such subscriptions telling what the lists dealt with must be presupposed in the in-
dividual lists. As a parallel compare the later Elamite lists of rulers published by Scheil
in RA XXVIII (1931) 2, which are subscribed 12 harrvP^ sd, a-wa-ankW and 12 sarrupl
si-mdk-su-ti,
45
Even more likely is it that these dynasties originally were more extensive than now,
i.e., that each of them counted a reasonable number of kings in the original of the King List.
They would then be in line with the normal lists of rulers. Evidence to support this assump-
tion is discussed on pp. 99 and 102, nn. 174 and 186.
46
The analysis presented here had reached its present form when Guterboek's study of
the problem in ZA n.F. VIII 4-6 appeared. It seemed impractical to work discussions of the
points on which we differ from his results into our main argument, so we have preferred to
discuss his solution separately and as a whole.
Gtiterbock assumes two main sources for the King List, a local tradition from Kish and
a local tradition from Uruk, each of which would consist of mythological, legendary, and
perhaps also historical elements. He differentiates between mythological and legendary as
oi.uchicago.edu
COMPOSITION 157
2. List from Uruk containing 1st, 2d, 3d, and 4th dynasties of Uruk
3. List from Ur containing 1st and 2d dynasties of Ur
4. List of unknown provenience containing the kings who now appear before
Etana in the 1st dynasty of Kish
follows: "Das Wort Sage verwenden wir dabei in der herkommlichen Bedeutung; die Grenze
gegen den Mythus ist dadurch gegeben, dass im My thus weder der Schauplatz noch die han-
delnden Personen der irdischen Realitat angehoren, wahrend die Sage von realen, in die ge-
schichtliche Sphare gehorenden Personen und Ereignissen ausgeht." He counts as mytho-
logical the 1st dynasty of Kish down to Etana and the 1st dynasty of Uruk down to Gil-
games, which we consider legendary, and as legendary the last part of the 1st dynasty at
Kish and at Uruk, Kish II-IV, and Uruk II, which we consider as mainly historical. The
material for this last part of the two traditions he assumes to have come from lists of names
kept in the two cities in question. These two local traditions supplied the names, reigns,
and the material in the "notes" for the kings of Kish and Uruk which we find in the list.
The remaining information, that which concerns the dynasties of Ur, Akshak, Maeri, Awan,
etc., was, according to Giiterbock, derived not from local traditions but from royal inscrip-
tions from the Enlil temple in Nippur: "Wir diirfen annehmen, dass die Gelehrten von Nip-
pur hier allerlei Nachrichten vereinigt haben, die sie wenigstens zum Teil den Bau- und
Weihinschriften im Enliltempel entnahmen."
As will be seen, Giiterbock has on two points reached conclusions similar to ours. He as-
sumes local traditions as sources, even though he does so in only two cases, Kish and Uruk,
and even though his picture of these sources differs from ours. He also considers the pos-
sible existence of "Namenlisten . . . . die in den beiden Stadten gefiihrt wurden," which is
not so far away from our conclusion that the sources were date lists giving names and reigns
of local rulers.
In general, however, we cannot follow his analysis. We must uphold our main division
into epic-historical sources for the "notes" and individual date lists for the names and reigns.
Giiterbock assumes that his two main traditions, those of Kish and of Uruk, combined both
kinds of material; but the material which is now found in the "notes" has been chosen ac-
cording to such definite plan (unusual parentage, native city, profession, etc.) and has been
put in shape and worked into the framework so uniformly that it must have been added en
bloc by a single hand, that of the author. If it had constituted original elements of different
sources this regularity would be unthinkable.
Nor can we accept Giiterbock's suggestion that the first part of the 1st dynasty of Kish
and of the 1st dynasty of Uruk should derive from mythological tales. Mythological tales
are generally concerned with a single hero only. It is assuming a great deal to presuppose a
tale for every one of the five Uruk kings and thirteen Kish kings whom Giiterbock's analysis
assigns to mythological sources. A still greater difficulty is the fact that while mythological
tales can have supplied the names of the list they would contain no information concerning
the mutual order of the rulers, which is an important feature of the list. Our own assumption
that these sections derive from lists of rulers just like the later parts of the dynasties in ques-
tion seems therefore definitely preferable.
Lastly, we must reject the suggestion that royal inscriptions from Nippur formed the
source for that part of the King List which was not covered by the two local traditions from
Kish and Uruk. Royal inscriptions of this period rarely give information concerning the
mutual order of two rulers—the Lagash inscriptions form an exception in this respect—and
they never give any information concerning length of reign. Since figures for reigns could
only have been found in date lists, which would also give the names of the kings in order,
the author of the list could have gained little from royal inscriptions. We must therefore
assume that these sections also are based on date lists.
oi.uchicago.edu
B. Smaller traditions:
1. List from Agade with dynasty of Agade
2. List from Maeri with dynasty of Maeri
3. List from Akshak with dynasty of Akshak
4. List from Awan with dynasty of Awan
II. One or more texts of epic-historical type. Source of the material in the "notes''
METHODS
Having thus formed a general idea of the sources which the author of the
King List had at his disposal, we may next consider in what wray he has treated
this material to produce his final work.
The dynasties of the King List are presented consecutively. At the end of
each dynasty appears the sentence, "The city A was smitten with weapons;
its kingship was carried to the city B , " and the new dynasty is then intro-
duced: "In the city B N. became king." It is thus clear that the author oper-
ates with the idea of a single "kingship" owned at different times by different
cities and that his work is intended as a record of the vagaries of this "king-
ship" in which it is possible to see where it was at any given time. Since our
author wrote his work from date lists, separate lists which enumerated the
rulers of a given city as completely as possible and as far back as possible, it
seems unavoidable that large sections in each of his sources would have been
irrelevant because they dealt with rulers reigning at periods when their city
was not in possession of the "kingship." To record the vagaries of the "king-
ship" as the author intended, he would therefore have had to arrange his local
lists side by side in a sort of synchronistic table and then compile his own list
by working down from the top, deciding for each successive period which city
at that time held the kingship and including only the rulers from there. 47
When we test this assumption with the means at our disposal we find, how-
ever, that it disagrees with all other facts and would lead to untenable results.
An example will illustrate this. From Mes-Anne-pada of the 1st dynasty of
Ur down to Sargon of Agade the King List enumerates some forty-five rulers.
If the author had worked in the manner described, these forty-five rulers must
have followed one another in time. Now we know from archeological and epi-
graphical evidence that Mes-Anne-pada was roughly contemporaneous with
Ur-Nanshe(k) of Lagash, and we know also how many rulers Lagash had from
47
A supposition that the King List was composed in some such fashion has been gen-
erally held; it underlies such statements as "Alle diese, durch Originalinschriften als Ober-
herren erwiesenen Fursten aber sind in den Konigslisten iibergangen, vielleicht, weil sie
Sumerer, Feinde der Semiten, waren, vielleicht auch, weil sie teilweise den religiosen Patesi-
Titel dem weltlicheren Lugal-Titel vorzogen" (Unger in OLZ XXXVII [1934] 364) and
"vermutlich sind diese Herrscher von der Uberlieferung nicht als wirklich berechtigte Voll-
konige anerkannt worden" (Meyer, Die dltere Chronologie . . . . , p. 37).
oi.uchicago.edu
COMPOSITION 159
Ur-Nanshe(k) down to the time of Sargon; they number eleven. It is clearly

impossible to imagine that forty-five rulers, many of whom were father and
son, followed one another on the throne of Babylonia in the same period which
in Lagash was covered by only eleven. The author of the King List therefore
cannot have dealt with his material in the way he would have us believe.48 To
see what actually happened we must look for passages in his work which will
throw light on his methods and where possible retrace his steps.49
A good example is the treatment of the dynasties Kish III (Ku(g)-Baba),
Akshak, and Kish IV. To make clearer what the author actually did we may
first see what he should have done. The sources for the section must have
been pre-Sargonic date lists, and as two cities are involved wre can assume
that the author had two such lists, an Akshak source from which he took the
dynasty of Akshak and a Kish source for the two Kish dynasties. If he had
worked by the selective method described above, these sources should have
looked as follows:
Akshak Source Kish Source
iod A N. Ku(g)-Baba
Unzi N.
Undalulu N.
Ur-ur N.
•iod B <
Puzur-Nirah N.
Ishu-il N.
. Shu-Sin N.
'N. Puzur-Sin
N. Ur-Zababa(k)
N. Simu-dar
-iod C « N. tFsi-watar
N. Eshtar-muti
N. Ishme-Shamash
. N. Nannia
For Period A the author would have decided on Kish as seat of the kingship
and would have entered Ku(g)-Baba in his list, disregarding the corresponding
ruler of Akshak. In the next period, B, he would decide that Akshak was capi-
tal and enter the rulers from there, rejecting the corresponding rulers in the
Kish source. Lastly, in the third period, C, he would come to the conclusion
48
The example here quoted is due to Weidner, who already in 1923, in Archiv fur Keil-
schriftforschung I 95, had pointed out that many of the dynasties which appear as successive
in the King List were in reality contemporaneous. He uses the example as proof of this
thesis in AOF III (1926) 198. Weidner's thesis is now accepted by all scholars.
49
The examples mentioned in the following pages are for the major part well known. They
were pointed out by Weidner in the articles quoted above and have been discussed by other
scholars also.
oi.uchicago.edu
that the kingship had returned to Kish once more and would again use the
Kish source in preference to the Akshak source.
In reality, however, something quite different must have happened. The
first ruler of Kish IV, Puzur-Sin, is stated to be the son of Ku(g)-Baba, who
represents Kish III. These two dynasties must therefore have followed directly
upon each other in the Kish source, and there accordingly cannot have been a
series of rulers between them to correspond to the Akshak dynasty. The au-
thor has not rejected any material. Instead of preferring for a certain period
the rulers of the Akshak source to contemporaneous rulers in the Kish source,
he has forced a continuous list from Kish apart and inserted a list of rulers from
Akshak. In other words, he is not working on a selective method; on the con-
trary, he uses all the material he has and interpolates one source into the other.
Graphically we may express his procedure as follows:
Akshak Source Kish Source

Unzi \ Ku(g)-Baba
Undalulu ^ Puzur-Sin
Ur-ur l^^"^*"^ Ur-Zababa(k)
Puzur-Nirah I Simu-dar
Ishfi-il 1}§i-watar
Shu-Sin J Eshtar-muti
Ishme-Shamash
Nannia
Another passage where we can observe the working method of the author is
the section from Ur-Zababa(k) of the 4th dynasty of Kish to Sargon of Agade,
and here again we get similar results. The author has placed the 4th dynasty
of Kish in front of the 3d dynasty of Uruk, Lugal-zage-si, and this in turn in
front of Sargon of Agade. In a "note" to Sargon he states, however, that this
king was originally cupbearer of Ur-Zababa(k), who is the second ruler of the
4th dynasty of Kish. I t is therefore clear that Sargon must have reigned fairly
close to this ruler in time; and, since we know that Sargon gained his suprem-
acy over Babylonia by defeating Lugal-zage-si, we must assume that Lugal-
zage-si gained his by defeating Ur-Zababa(k). The actual succession should
therefore be
Ur-Zababa(k)
Lugal-zage-si
Sargon
and the author should have listed Lugal-zage-si just after Ur-Zababa(k). In-
stead we find that these two are separated in the King List by as many as five
oi.uchicago.edu
COMPOSITION 161
rulers of Kish, namely Simu-dar, tr§f-watar, Eshtar-muti, Ishml-Shamash,
and Nannia, who cover a period of 66 years.
These five rulers come of course from the Kish source which the author is
using. Being a date list from that city, it must naturally have listed all rulers
of Kish, including those who reigned after the hegemony of Babylonia had
passed to Uruk and to Agade. These later rulers in the source, who were con-
temporaneous with Lugal-zage-si and Sargon and who had to acknowledge
them as suzerains, should accordingly have been rejected by the author of the
King List as only petty kings. Instead he has pulled Ur-Zababa(k) and Lugal-
zage-si apart to make room for them. We see once more that our author does
not reject any of his material but uses it all, even if he has to force apart rulers
who belong together in order to get it in.
The examples which we have here mentioned are by no means isolated in-
stances but could easily be supplemented with others which tell the same tale.
They allow us to draw our first conclusion concerning the working methods of
the author: He used his material to the full, was indeed so far from rejecting
anything that he would even separate kings who belong together to get every-
thing in.
This conclusion is of considerable interest. It means first of all that many
of the dynasties listed as consecutive in the King List were in reality contempo-
raneous, and it explains effortlessly how the King List can give forty-five rulers
for a period which was covered in Lagash by only eleven. Secondly, and just
as important, it means that the groups representing the various sources which
we have singled out earlier must give us fairly complete pictures of the docu-
ments for which they stand; for, if the author was using his material to the full,
we must possess that material unabridged in his work.
Among the groups which we singled out there were three larger ones, the
Kish, Uruk, and Ur groups, made up from material found in various places in
the King List. If these groups, as is likely, contain the full amount of material
given in the corresponding sources, the author can have done no more to the
sources than cut them up and distribute them over his work. And it is obvi-
ously of interest to know how he did this cutting and what reasons can have
prompted him to do such cutting at all. An example of how the editor cut has
already been mentioned—the cut which separated Ku(g)-Baba from her son
Puzur-Sin. This example is, however, exceptional, for nowhere else do we find
an indication that the author has severed palpably organic connections be-
tween two rulers, and since a ruler is often stated to be the son of his predeces-
sor there would be sufficient opportunity. It seems likely that the author gen-
erally placed his cuts more carefully, and there are indications of an original
oi.uchicago.edu
division of the rulers into dynasties which seems to go back to the sources
themselves and which probably influenced him.
The system of division which now dominates the King List is based on the
various movings of the "kingship" from one city to another: "The city A was
smitten with weapons; its kingship was carried to the city B." Since each of
the author's sources dealt with a single city, it is clear that these divisions
must represent the places where he has joined pieces of two sources. Along
with these divisions we find, however, another system in which the beginning
of each new sequence is indicated by explicit use of the words 1 u g a 1 -
k m , "was king" or "became king," which were of course implied after
every name in the list. Comparing the occurrences of the two systems, we
see that the divisions made by the moving of the "kingship," divisions which
represent the places where the author joined two of his units, seem dependent
on the l u g a l - & m divisions, for they always follow them closely, whereas
the latter are self-sufficient and can appear alone. In the beginning of the list
both Etana and En-me(n)-barage-si are marked in this fashion as first rulers
of dynasties although the "kingship" remains stable in Kish all the time.
Since the system which results from the author's cutting and rejoining the
sources, the "kingship" system, thus follows in the steps of the division into
dynasties expressed by the l u g a l - ^ m formula, the latter must be pri-
mary and can be assigned to the original sources. Our author has found such
divisions into dynasties there and has made his cuts according to them.
We have thus seen that the author of the King List has cut up the larger of
his sources into smaller units and that in general he seems to have made his cuts
according to an existing division into dynasties, so that the units he obtained
were natural units, the old dynasties of the sources. It is natural to ask what
purpose he can have had for treating his sources in this fashion.
As we have often had occasion to mention, the author of the King List oper-
ated on the theory of a single kingship of all Babylonia which could belong to
only one city at a time. That this theory disagrees with actual facts—at least
for the older periods of Babylonian history—need not be demonstrated. We
know that different "kingships" existed side by side in Babylonia as late as
the Agade period, probably even later. Since the author was thus working on
a wrong theory, it seems inevitable that he must have run up against difficul-
ties when he began to apply it to his material. The bulk of this material con-
sisted of separate lists, each dealing with rulers o*f a different city. To make it
agree with the theory of the single "kingship" it would be necessary to arrange
these lists successively so that the "kingship" appeared to pass from the city
of one source to the city of the next. Such an arrangement could, however, be
oi.uchicago.edu
COMPOSITION 163
made in good faith only if the author knew of no synchronisms between his
rulers. The lists which formed his sources must in reality have covered largely
the same periods; so if he arranged them successively, rulers of different cities
whom the synchronisms showed to be contemporaneous would of course ap-
pear far apart in his final list and thus expose the fallacy of the arrangement
to him.
Now it can hardly be doubted that our author did know such synchronisms.
One of the "notes" which he inserted in his work states that Sargon began his
career as cupbearer to King Ur-Zababa(k) of Kish, so he must have known
that these rulers were contemporaneous; and he cannot have avoided finding
similar synchronisms in the material from which his other "notes" were de-
rived. As we have seen earlier, the sources he used for this feature of the list
must have been texts of the epic-historical genre: epics, legends, historical
anecdotes, etc.; and such texts are extremely rich in synchronisms. It there-
fore seems certain that our author faced a serious contradiction between his
theories and his evidence. In the light of this dilemma his treatment of the
sources, which at first glance may appear curious, becomes very clear and un-
derstandable.
It is evident that if the author arranged the sources successively as just
suggested and compared the result with what he possessed of synchronisms his
longest sources would be the worst offenders; for, if two long lists which should
actually be placed side by side are arranged consecutively, the distance be-
tween any two contemporaneous rulers in them will naturally be much greater
than if the lists were short ones. If the long sources are cut up into smaller
units, however, and the units are so arranged that a unit of one source is suc-
ceeded by the corresponding unit of the other source, the discrepancies can be
considerably reduced. And this is exactly what has been done with the sources
in the King List. As an example we may mention the treatment of the rulers
of Uruk and Kish. There is here a synchronism between Gilgames of Uruk and
Aka of Kish which the author may well have known.50 If the complete Uruk
source were placed after the complete Kish source, the two contemporaneous
rulers, Gilgames and Aka, would be separated by sixteen other kings. If the
sources are cut up into smaller dynasties and joined as we find them in the
King List, however, only four rulers will intervene—an obvious improvement,
even though it too is not correct.
The arrangement of the sources which we meet in the King List thus proves
to be a compromise between two conflicting forces: on the one side the author's
60
It appears in the text relating how Gilgames liberated Uruk by defeating Aka of Kish
(see p. 145, n. 20).
oi.uchicago.edu
theory of a single "kingship" and the successive arrangement for which it calls,
on the other his knowledge of synchronisms showing up the errors to which
that arrangement leads. The theory has prevented him from radically rectify-
ing the errors, but he has tried to keep the worst of them down somewhat by
cutting his largest sources and joining the resulting units separately.
Summing up, we can therefore characterize the author's working method as
follows: He used all his material. In arranging it he was guided first of all by
a dogma: that Babylonia had always been united in a single kingdom with a
single capital, so that two different cities could never have held the "kingship"
simultaneously. This dogmatic notion presupposed a successive arrangement
of his sources, which were mainly separate lists of rulers of various cities, an
arrangement which would necessarily conflict with the synchronisms which he
knew from other parts of his material. To reconcile in some measure the syn-
chronisms with the successive arrangement, the author therefore cut up the
larger of his sources into smaller units, the dynasties of which they consisted,
and joined the units separately; for by this means he was able to reduce some
of the worst discrepancies, even though he naturally could never get anywhere
near a true chronology. His treatment and arrangement of the sources become
thus in reality a bed of Procrustes. They constitute an effort to force correct
evidence into a mold shaped by an erroneous theory of Babylonian history.
oi.uchicago.edu
V
HISTORICAL VALUE
RELATIVE VALUES OF FEATURES IN THE KING LIST
The results at which we have arrived in the foregoing pages have given us a
general idea of the sources which the author of the King List had at his dis-
posal and have shown us the manner in which he utilized them. We should
accordingly be in a position to judge the historical value of his finished work,
the King List.
Considering the King List and its information from this point of view, we
can discard first of all those features which must be ascribed to the author him-
self, more exactly the arrangement and succession of the individual dynasties.
As we have seen, the successive arrangement of the sources which the King
List exhibits resulted from the author's erroneous theory of a single kingship.
It is true that the author seems also to have had a number of synchronisms and
that by cutting up his sources he tried to reconcile the arrangement in some
measure with their data, but these efforts could hardly do more than reduce
a few of the very worst errors and could not prevent the violent distortion of
his chronology which the successive arrangement implied.
While the arrangement of the King List must thus be considered of negli-
gible value, the actual material from which it has been built up inspires more
confidence.1 The material comes, as we have seen, mainly from local lists of
1
This result agrees in several points with earlier estimates of the historical value of the
King List. The worthlessness of the arrangement of the dynasties seems to have been sus-
pected by Legrain already in 1922. He writes: "The problem of parallel dynasties is one
of the most troublesome for Babylonian chronologists" (PBS XIII 17). It was brought
up energetically by Weidner in 1923 in Archiv fur Keilschriftforschung I 95 and later in
AOF III (1926) 198. From then on parallelisms were admitted as probable or certain by
most scholars, e.g. Meyer, Die altere Chronologie . . . . , p. 38; Langdon, Excavations at Kish
I (Paris, 1924) 5 f. and 100; Meissner in WZKM XXXII (1925) 296 and 303 and in his Die
babyhnisch-assyrische Literatur (Wildpark-Potsdam, 1927) p. 88; Gadd in UE I (1927)
137 f.; Smith, EHA, p. 39; Landsberger in OLZ XXXIV (1931) 119.
As for the material of the King List, the obvious unreliability of the excessively long reigns
found in the earlier part of the list was naturally recognized from their first appear-
ance. But criticism has here gone farther than we have gone above and has raised the ob-
jection that kings who might be expected in the King List do not appear there. This point
was made by Thureau-Dangin as early as 1918 (La chronologie des dynasties de Sumer et
d'Accad, p. 66) and was often repeated later, e.g. by Langdon, OECT II 6 f.; Meyer, op. cit.
pp. 36 f.; Smith, EHA, pp. 37-39; Landsberger, loc. cit. Considered together with the other
165
oi.uchicago.edu
166 THE STJMERIAN KING LIST
rulers, date lists kept in various cities for practical purposes; and the author
has done little to them beyond cutting up a few of the longer ones and dis-
tributing through his compilation the smaller units thus obtained. Such ma-
terials must undoubtedly be considered very reliable sources of information,
for most of these lists stop at the beginning of the Agade period and were
therefore in all probability copies from that period; and the practical purpose
which they served would guarantee their accuracy. Looking through these
sources we find also that only a single point invites criticism. In the longest
of them the earliest rulers appear with abnormally long reigns, often as much
as several hundred years. At a certain point the reigns then change and become
quite normal, keeping within what seems humanly possible. That the immense
reigns are unhistorical is obvious. Their occurrence in our material must be
ascribed to a tendency known also among other peoples of antiquity to form
very exaggerated ideas of the length of human life in the earliest times of which
they were conscious.2 These reigns must therefore be rejected. Since the tend-
ency which we have mentioned could of course have influenced only existing,
more correctfiguresor, if the list gave no reigns for these rulers, have prompted
somebody to add figures of this type, whereas it cannot have influenced or
changed the names and their mutual order, there is obviously no reason to
reject more than the exaggerated figures themselves.3
objections, this point has generally led to rejection of the King List. We may quote Thu-
reau-Dangin (he. cit.): " ... les listes royales sont, pour la periode ant^rieure a Lugalzaggisi,
des temoins tout au moins suspects (historiquement parlant)"; Langdon (OECT II 6): "It
is altogether obvious that the dynastic lists for the early period are totally inadequate";
and Landsberger (he. cit.): "Daraus ergibt sich, dass wir uns von der Konigsliste vollstandig
emanzipieren mussen." It is seldom we find a verdict as lenient as Gadd's (UE I 138):
". . . . it may perhaps be regarded as a principle in dealing with the king-list to consider
that the dynasties, the individual kings, and the general order of the accession of both, are
for the most part authentic, but that the figures, both items and totals, need severe criti-
cism." In the estimate of the value of the King List as a historical source which we give in
this section we do not touch on the missing rulers at all. The importance of this point has,
we believe, been greatly overrated, mainly because of the false impression of the scope of
the King List which its author gives. We shall see later (pp. 180-83) that most of these
"missing" kings have in reality no valid claim to be expected in the sources of the King
List. For this reason our estimate of the value of the King List is higher than the usual one.
2
Cf. Meyer, op. cit. p. 36: "Wie in der aegyptischen Darstellung der Vorgeschichte ist
auch hier der tlbergang von den mythischen Gestalten der Urzeit vor und nach der Flut zu
den altesten wirklich geschichtlichen Erinnerungen durch ein schematisch durchgefiihrtes
Herabsinken der Lebensdauer und der Regierungsjahre von vielen Jahrhunderten . . . .
auf das gegenwartige kurze Mass des menschlichen Lebens vermittelt." Cf. also the similar-
ly high figures attributed to the Hebrew patriarchs in the Old Testament.
8
Cf. also the verdict of Smith concerning the still higher figures of the antediluvian rulers
(EHA, p. 22): "An impartial consideration of all the evidence must allow the possibility
oi.uchicago.edu
HISTORICAL VALUE 167
Our conclusion concerning the historical value of the King List must thus
be that while the arrangement, the succession of the various dynasties, can be
considered a later construction of no significance, we possess in the actual ma-
terial of that document a historical source of high value, from which only some
exaggerated reigns occurring with the earliest rulers should be segregated.
ISOLATION AND RECONSTRUCTION OF THE SOURCES

The conclusions at which we have here arrived are obviously of the greatest
interest for Babylonian chronology. We have seen that the sources which the
author of the King List used can be considered reliable, and we have seen also
that the material of these sources must be embodied unabridged in the present
King List (pp. 159-61). It should therefore be possible to undo the author's
work, reconstruct and rearrange the original sources, and thus obtain a reliable
scheme covering the older periods of Babylonian history. Such far-reaching
consequences naturally make it desirable to test as far as possible the conclu-
sions on which they are based; and the most convincing test is to undertake
the reconstruction suggested, compare its result with what we know from other
sources about the chronology of the period concerned, and see whether they
agree.
Above we have seen how the individual sources used by the author can be
singled out (pp. 153-58). To make our test these sources should be correlated
and then compared with our other chronological material. For such correla-
tion, however, synchronisms are necessary; and, as the synchronisms which
we possess will allow us to correlate only five of the sources, those of Kish,
Uruk, Ur, Agade, and Akshak, we must content ourselves with them. For the
same reason material following the Agade dynasty in the King List has been
left out of consideration for the time being.
THE KISH SOURCE

r
The source from w hich the author of the King List got his rulers of Kish
may be reconstructed as shown on page 168.
The way in which the material can be singled out from that of the other
sources has been shown earlier, where we have also seen that the source in all
probability began with Etana. Instead of the "legendary" reigns which ap-
peared in the source we have given the rulers in question average reigns of
that the names of these pre-diluvian kings and the city names associated with them may
be based on evidence of their existence available to the Babylonians about 2300-2100 B.C.
and unknown to us. In that case their names were connected with the figures arbitrarily, for
some reason not yet apparent."
oi.uchicago.edu
about 20 years, heightening them to 30 years, the average of a generation, in

the eases where a ruler was followed by his son.4 Where two sons succeeded a
Etana ca. 30 years
I
Balih ca. 20 years
En-me (n)-nunna ca. 20 years
Melam-Kishi(k) ca. 20 years

Bar-sal-nunna ca. 20 years
Samug ca. 30 years
Tizkar ca. 20 years
IlkiP ca. 20 years
Ilta-sadum ca. 20 years
En-me (n)-barage-si ca. 30 years

1 ca. 20 years
Aka
Su ca. 20 years
Dadasig ca. 20 years
Magalgalla ca. 30 years
1 ca. 20 years
Kalbum
Si-e ca. 20 years
GA+suB-nun-na ca. 20 years
Enbi-Eshtar ca. 20 years
Lugalmu ca. 20 years
Ku(g)-Baba ca. 30 years

1 25 years
Puzur-Sin
1
Ur-Zababa(k) ca. 20 years
Simu-dar 30 years
t)§i-watar 7 years
Eshtar-muti 11 years
Ishme'-Shamash 11 years
Nannia 7 years
ruler on the throne and where we have thus only two generations, we have as-
signed 20 years to each of the three rulers.
The division into dynasties is that indicated by the 1 u g a 1 - k m formu-
4
Compare with thesefiguresthe average of the reigns found in the 1st dynasty of Baby-
lon, in which son followed father almost from the beginning of the dynasty to the end: a
little more than 27 years.
oi.uchicago.edu
la. However, we have not separated Ku(g)-Baba from her son, because this
separation seems to be secondary and due to the author of the King List, who
needed room for the Akshak kings there. Only a few of the names need com-
ment. The reading of the name Samug has been discussed above (p. 82, n. 85).
Our reasons for preferring the form Magalgalla were given on page 96, note 154.
This leaves only one name still to be discussed, the one which we read Enbi-
Eshtar. The section in which it occurs is preserved only in version WB. The
line in question reads, according to collation (the part outlined in dotted line
represents erasure):
That this line is corrupt is clear. The ancient copyist apparently did not un-
derstand what he was copying. I t seems possible, however, to trace the original
form. The clue is given by the second sign, which can only stand for an original
P&-, that is, REC, No. 381, g u r u n : eribu. This sign may be considered
in connection with the damaged sign at the end of the line which Langdon reads
i § (?) but which more likely represents an original t a r . Since we know
a king of Kish by the name Enbi-Eshtar, 5 it seems highly probable that our
scribe was trying to render as faithfully as he could a damaged line
i
-enbini"ib-e§4-tdr
"Enbi(gloss: Inib)-Eshtar."
His smoothing of the surface or "erasure" would represent the shading of a

modern copyist; and a few accidental scratches of his original were taken over
as wedges, especially between 1*=? and < , so that the latter became **.
THE URUK SOURCE
The Uruk source can be reconstructed as shown on page 170.

As in the Kish source, we have replaced the "legendary" reigns with aver-
age reigns of 20 and 30 years. In the 2d dynasty also we have used average
reigns, since the actual figures are unknown, but we have here made an excep-
tion from our usual procedure, inasmuch as we have assigned the maximum,
30 years, to each of the three rulers of the dynasty. Only one point, namely
the 2d dynasty, which has come down to us in badly corrupted form, causes
6
PBS IV 1, p. 151.
oi.uchicago.edu
Mes-kiag-gasher ca,, 30 years

En-me(r)-kar ca,, 20 years
Lugal-banda ca., 20 years
Dumu-zi(d) ca,, 20 years
Gilgames ca,, 30 years
Ur-Nungal(ak) 30 years
1
Utul-kalamma (k) 15 years
Laba. . . .m 9 years
En-nun-dara-Anna (k) 8 years
MES(?)-HE 36 years
Melam-Anna(k) 6 years
Lugal-ki-tun(?) 36 years
En-shakush-Anna(k) ca. 30 years
[Lugal-kinishe-dudu] ca. 30 years
[Lugal-kisal-si] ca. 30 years
Lugal-zage-si 25 years
difficulties in the reconstruction of this source. WB, where it is best preserved,

reads:
iv45 u n u k i - g a e n - s a k £ n - § a 4 - a n - n a
46 l u g a l - a m m u 1 § u - s i i-a 5
47 n a m - l u g a l - b i m u 2 s u - § i 1 - a5
48 m u 7 l-a 5
As will be seen, the text is in order down to line 46 only. The following line
begins with the words n a m - l u g a l - b i , "its kingship," which in reality
belong to the last part of the formula for change of dynasty, n a m -
l u g a l - b i B k i - s e b a - 1 ti m , "its kingship was carried to the city B , "
and continues with m u 2 § u - s i 1 - a&, "he reigned 120 years/' which
is a formula for the reign of a single ruler. The last line, m u 7 i - a s ,
"he reigned 7 years," also contains such a formula. What we have here can
be no more than gleanings, a few disconnected phrases still readable in an
otherwise badly broken section of the scribe's original. To judge from what
is left, that section must originally have listed two individual rulers with their
reigns and—presumably after a dynasty total which has been completely lost
—closed with the formula for change of dynasty. We can therefore conclude
that our dynasty had three rulers in all, En-shakush-Anna(k) and the two of
whom we have only traces in the lines giving individual reigns.
oi.uchicago.edu
This conclusion is supported by the evidence of another of our versions, P 2 .

This version states that five dynasties of Uruk, numbering 22 rulers, were
enumerated in the King List; and from our other versions we know that the
1st, 3d, 4th, and 5th dynasties of this city had 19 kings in all. There are thus
three left for the 2d dynasty, as the text of WB suggests. The text of Su3+4 also,
which preserves traces of this section, is explicable on the assumption of three
rulers in the dynasty. It has after the reign of En-shakush-Anna(k) the entry
[ l i - g u b a] - a n - d 6 - a [. . m u 1] - a5, "Somebody who has disappeared
reigned . . years," and then gives the dynasty total. This probably reflects
a manuscript in which the second ruler and all but the reign of the third had
disappeared in a lacuna.
Since all our sources thus seem to agree, we can confidently assume that the
2d dynasty numbered three rulers. Less fortunate are we when it comes to
determining how long these rulers reigned. As we have seen, some figures have
been rescued by WB from the damaged section in its original. These figures,
however—60, 120, and 7—are not of a nature to inspire much confidence. We
know, further, that they come from a fragmentary text which must have been
difficult to read, and we have no means to judge how much of the figures the
copyist could actually read, how much he himself restored. It is therefore im-
possible to use this material with any confidence, and we have accordingly
assigned average reigns to the rulers concerned.
It remains to consider the identity of the rulers in this dynasty. The list
has preserved the name of the first of them. It appears completely in WB only
and is there written e n - £££* -DU - a n - n a (in photograph). That this
writing must cover the name of the king En-shakush-Anna(k), who reigned
in Uruk 6 and who on epigraphic evidence can be placed a few generations be-
fore Lugal-zage-si, seems obvious.7 We have therefore assumed that the doubt-
ful second sign «£^ (in Langdon's copy) is to be identified with ^ ^ , earlier
4
S^f which has the value § a k d n , and read the name e n - S a k d n -
§ a4 - a n - n a .7a
After e n - § a k d n - 5 a 4 - a n - n a our source must have listed two
other rulers, who would come immediately before Lugal-zage-si. From his-
torical inscriptions we know of two rulers of Uruk who must be assigned to
6
As suggested by his title l u g a l k a l a m - m a ; see Poebel, PBS IV 1, p. 153.
7
The identification has been accepted as probable by Langdon, OECT I I 6 and 14, n. 9; cf.
also Christian's remarks in AOF VIII (1932/33) 209.
7a
E n s a k a § a n n a ( k ) constitutes a normal later form of e n § a ( g ) k u § a n -
n a (k), in which the unstressed u of k u h has been assimilated to the following a in
a n n a ( k ) . Cf. e.g. d u r a f c , "bouquetin," > d a r a ( f c ) ( d a r a ) , * s u & a r (loan
word from Akkadian vufyarum), "page," > s a f c a r .
oi.uchicago.edu
just that time: l u g a l - k i - GUB-NI - d u7 - d 117, whose name, as recent

texts have shown, is to be read Lugal-kinishe-dudu, 8 and his son(?) Lugal-
kisal-si.9 It therefore seems evident that these are the two names missing in
our source, and we have accordingly restored them in the gap. 10
T H E U E SOURCE
T h e Ur source can be reconstructed as follows:

Mes-Anne-pada ca. 40 years 1 „
<A-Anne-pada> An r 80 years
Mes-kiag-nunna (k) 36 years J J
ca. 40
Elulu 25 years
Balulu 36 years
[Lu]gal-[kinishe-dud]u ca. 16 years

[Lugal-kisa]l-si ca. 30 years
. . . .gi ca. 35 years
[Ka]-ku(g) ca. 35 years t
As in the previous reconstructions, the division of the rulers into dynasties

follows the l u g a l - a m indications. Version WB gives, it is true,
8
L u g a l - k i - GUB-NI - d u7 - d u7 was, according to his own inscriptions, king of
both Uruk and Ur (SAK} p. 156, No, 3 b). He reigned together with his son(?) Lugal-
kisal-si, and the two kings appear together in an inscription (SAK, p. 156, No. 3 c). To the
same time as L u g a l - k i - GUB-NI - d u7 - d U7 belongs—for epigraphic reasons—a cer-
tain L u g a l - k i - n i - S e — d u 7 - d u 7 , who similarly was rider of Uruk (Gadd in RA
XXVII [1930] 125 f.; Barton in JAOS LI [1931] 262-65; Kruckmann in An. Or. X I I [1935]
200 f.) and who in an inscription from Ur also appears together with Lugal-kisal-si (UET I,
No. 3). In this inscription a certain a - n u - z u who is "merchant of the king Lugal-
kinishe-dudu ( d a m - k a r - r a [lu g a l ] - rk i1 [-ni] - s & - d U7 - du 7 l u g a l - k a m ) "
dedicates an alabaster vase to Nanna for the life of Nin-tur and Lugal-kisal-si (n a m - 1 i
l u g a l - k i s a l - [ s i - § & ] ) . Since l u g a l - k i - GUB-NI - d u7 - d U7 with transposi-
tion of the signs GUB and NI can be read I u g a l - k i - n i - § e i 3 - d u 7 - d u 7 , there can
be no doubt that the two rulers are identical. That transposition of signs, which on the
whole belongs to the period before E-Anna(k)-tum, can still occur at the time of our ruler
is shown by the writing l u g a l - s i - k i s a l (BE I 2, PL 37, 1. 7 from end) as against
l u g a l - k i s a l - s i (AJSL XXI [1904/5] 63). Cf. also BE I 2, PL 37, 1. 4 from end:
n a m - t i - § e - l a for n a m - t i - l a - § & . The identity of the rulers Lugal-kinishe-
dudu and l u g a l - k i - GUB-NI - d u7 - d u7 was first pointed out by Langdon in JRAS,
1931, pp. 421-24. He overlooked, however, the important evidence of identity given by
A-nu-zu's inscription, because he read [1 u g a 1] - k i [- GUB - n i] - d U7 - [d u7] instead
of [1 u g a 1] - rk i1 [- n i] - § & - d u7 - d u 7 , which it actually has.
9
SAK, p. 156, Nos. 3 c~d.
10
Thus also Poebel, PBS IV 1, p. 107, Langdon, OECTII14, n. 9, and others. Since these
names can with great probability be restored where the list now has a lacuna, no conclusion
as to insufficient knowledge on the part of the author of the list can be drawn from their
alleged absence, as is done by Giiterbock in ZA n.F. VIII 7.
oi.uchicago.edu
1 u g a 1 - It m after the name of Mes-kiag-nunna(k) also, but this variant does

not appear in our other source for the section, P 2 , and is obviously a mere
scribal error. A copyist strayed from the line m e s - k i - a g - d n a n n a
d u m u m e s - a n - n 6 - p & - d a back into the earlier m e s - a n - n 6 - p & -
da lugal-ltm.
After the name of Mes-Anne-pada we have restored that of his son and suc-
cessor, A-Anne-pada, which in all probability originally appeared in the King
List at this place, although it has been lost by our present versions. As pointed
out by Gadd,11 the much too high reign now assigned to Mes-Anne-pada, 80
years, contrasts with the otherwise reasonable and trustworthy figures of the
dynasty and suggests a conflation of two reigns, Mes-Anne-pada's own and
that of his son. The name A-Anne-pada was probably lost by some early
copyist, while the regnal years survived and were eventually added to those
of Mes-Anne-pada.12
The following name, Mes-kiag-nunna(k), appears in this form in P 2 , whereas
WB gives m e s - k i - a g - d n a n n a . That P 2 has preserved the correct
form is shown by a contemporaneous inscription from Ur which reads [ m e s -
k]i - a [g] - n u n 1 u g a 1 u r i k i , "Mes-kiag-nun(ak), king of Ur." 13 The
form Mes-kiag-Nanna(k) must be due to mishearing; this is especially likely
since it is more natural to expect the element d n a n n a , the name of the
chief god of Ur, than the element n u n n a in a royal name from that city.
The figure for Mes-kiag-nunna(k)'s reign likewise appears in two different
forms. P 2 gives 30 years, and another source, Li, which has preserved a total
for this dynasty agreeing with that of P 2 , probably had the same figure. WB,
however, gives 36 years. There can be little doubt that WB is here the original;
for while it is easy to see how the figure 36 if slightly damaged (^^ffft])
could become 30 ( ^ ) in a later copy, the opposite development would be
difficult to account for.
Passing to the 2d dynasty we come—as in the Uruk source—to the most
troublesome part of the tradition. We have here, however, a better basis to
11
UE I 128. Gadd's conclusion has been accepted by most scholars, e.g. Sidney Smith,
EHA, p. 36, and O. E. Ravn, Babylonske og assyriske Kongers historiske Indskrifter, p. 28.
12
As the high reign of Mes-Anne-pada must be considered an indication that A-Anne-pada
originally appeared in the list, the latter's absence in our present versions cannot be used as
an argument to show that "the authors'' of the list had incomplete knowledge of their sub-
ject, as is done by Guterbock in ZA n.F. VIII 7: "Gerade diese Auslassung des inschriftlich
belegten Konigs zeigt iibrigens, wie unvollstandige Kenntnisse die Verfasser der Konigsliste
hat ten."
» Burrows in UE II 321, n. 10 (U 11675).
oi.uchicago.edu
174 T H E STJMERIAN K I N G LIST
work from. We may consider first the question of how many rulers and how
many years should be assigned to that dynasty.
Version P 2 gives 3 dynasties, 13 kings, and 396 years as grand total for Ur.
As shown by the total after the 1st dynasty, it assigned four rulers to that
dynasty; and, since all our other versions agree that the 3d dynasty had five
rulers, we can assume that P 2 likewise counted five rulers there. Subtracting
the four rulers of the 1st dynasty and the five of the 3d dynasty from the
thirteen given in the grand total leaves four rulers for the 2d dynasty, with
which we are concerned.
As for the reigns, P 2 gives 171 years for the 1st dynasty, but the figures for
the 3d dynasty are missing and must be supplied from other sources. It is
clear that a reconstruction must be based on P 5 , which belongs to the same
main branch of the tradition (the B branch) as does P 2 and comes from the
same place, Nippur. 14 The figures given by P5 in this section
Ur-Nammu(k) 18
Shulgi 58
Bur-Sin 9
Shu-Sin 7
I(b)bi-Sin 25
are by no means correct. Shulgi reigned, as we have seen above (p. 122, n. 321),
only 48 years, not 58, and Shti-Sin reigned 9 years, not 7.15 We are therefore
faced with the problem of whether these errors developed before the tradition
which P 5 represents had separated from that of P 2 , in which case they would
have appeared in P 2 also, or whether they developed later, in which case they
would be peculiar to P 5 . It is very fortunate that this question can be decided,
at least as far as the worst error, that in the reign of Shulgi, is concerned. We
have seen above that the development from the original $W (48) to P 5 's
^ffi* (58) is naturally explained by the fact that P5 places the sign for year,
m u , in front of the figures, so that in >%&W a careless copyist could
easily come to count one of the final < -wedges of m u as part of the figure.
This arrangement must be considered an innovation introduced in Ps's line of
14
Sus+4, which also belongs to the same branch as P2, cannot come into consideration.
This text, as shown above (pp. 25, 51, and 108-9, nn. 219, 221, 224, and 228), has passed
through a badly damaged ancestor which later copyists tried to restore as best they could.
Since that ancestor (Su) must be placed at a point after the tradition represented by Su3+4
and Sui had swerved off from the tradition represented by P2, it is obvious that Su3+4 con-
tains a number of errors—due to restorations made in this damaged text—which P2 would
not have had.
18
The figure for I(b)bt -Sin, 25 years, differs by one year from that given in WB, 24 years,
but is the more likely to be correct (see above, p. 123, n. 331).
oi.uchicago.edu
descent after it had passed Y. P 2 , however, which we are trying to reconstruct,,

does not have this innovation but retains the arrangement characteristic of
the B group as a whole, placing m u after the figure. Here such a mistake
is not likely to happen, and we must accordingly conclude that P 2 had the orig-
inal 48 years for Shulgi. It is somewhat more difficult to decide whether P 2
also retained the original figure for Shu-Sin, 9 years, or whether it already had
the erroneous 7 years which we find in P 5 . However, since P 2 , written in the
reign of Enlil-banl (see p. 6), is considerably older than P 5 , which dates from
the second half of the 1st dynasty of Babylon (see p. 7), it seems probable
that it would be less corrupt, and we may therefore restore it with 9 years
for Shu-Sin. In any case this question is not of great importance, since it is a
matter of only two years. We thus get
Ur-Nammu(k) 18
Shulgi 48
Bur-Sin 9
Shu-Sin 9
I(b)bi-Sm 25
Total 109
as the most likely text for P 2 in the section with which we are concerned. Sub-
tracting P 2 's total for the 1st dynasty of Ur, 171 years, and this reconstructed
total for the 3d dynasty, 109 years, from the 396 years which P 2 has as grand
total for all three Ur dynasties, leaves 116 years to represent the duration of
the 2d dynasty of Ur.
The next question must then concern the identity of the four rulers. Besides
the kings appearing in the 1st and 3d dynasties of Ur, we know from inscrip-
tions three other kings of that city. Lugal-kinishe-dudu and Lugal-kisal-si,
whom we have mentioned above, were kings of Ur as well as rulers of Uruk.
A third king of Ur, Ka-ku(g), is mentioned in inscriptions of Bimush of Agade,
who defeated and captured him.16 Utilizing this knowledge, we may now turn
to the sparse remnants of the section concerning the 2d dynasty of Ur which
have been preserved in version Sui and try to restore its text (parallel to WB
v 2-13):
iv 1 [u r ikl - s e] r
b a - t li m1 to Ur was carried,
kL(7
[u r] f > In Ur
[1 u g a 1 - k i - n i] - rs e - d U7 - d U71 Lugal-kinishe-dudu
[1 u g a 1] - % m1 became king
18
PBS IV 1, pp. 189 f., Text N xvii 3-9; p. 192, Text O xix 10-12; p. 193, Text P xix
31-34.
oi.uchicago.edu
176 T H E SUMEBIAN K I N G LIST
[. . m u] I 1 - a5 (and) reigned . . years;

[1 u g a 1] - rk i s a 1 - siU?) Lugal-kisal-si,
[ d u m u l u ] g a l - [ k i - n i - § & - d u7]• -r d U71 son of Lugal-kinishe-dudu,
r
[. . m u] l' - a5 reigned . . years;
[....] - g i 4 • • • -gi
r
10 [. .] m u1 i-a5 reigned . . years;
r 1
ka - 'k u1 Ka-ku(g),
[dumu ] - g i4 -ke4 son of . . . .gi,
[. .] fmu 1 i-a5 reigned . . years.
r
4' 1 u ga1 4 kings.
As so little is left of the signs, it is necessary to comment on the restoration

line by line.
In line 1 H H M « t is left; we consider it remnants of M M . In line 2
WMSl^M is left; assuming that the shaded NA of the copy should be KI, we
consider this remnants of W$%0^M. In line 3 WMM%M is left; we consider
it remnants of WWM^M. In line 4 there is 'WMZWfr/-, which we consider
remnants of « » S 1 In line 5 Wmm<M is left; this is obviously Wmmm.
In line 6 WSmm is left; it seems possible that this could represent WMmmmi.
In line 7 Wmmtmm is left; these traces agree with a restoration W&Jmtim.
In line 8 Wm0WM. is left, obviously remnants of M O W . In line 9 the
sign W& (gi 4 ) remains at the end. Line 10 has rather complete traces of
[ . . ] rm u1 i - a 5 (NA). In line 11 we find H S W , which we have restored
as T c a - k i i 1 , assuming it to be the remnants of H H ! . Line 12 has
Hf4' ff ^ | g ^ . Line 13 has the end of m u and quite clear l - a B (NA).
Line 14, finally, gives the head of a single wedge of ft (4) and a clear
iugal.
The text thus restored gives the order and names of the rulers but has not
preserved a single reign. We must therefore content ourselves with the total,
which, as we saw above, was probably 116 years. The figures given in our re-
construction of the source add up to this total but are otherwise chosen arbi-
trarily.
THE AKSHAK SOURCE
The reconstruction of this source presents little difficulty:

Unzi 30 years
Undalulu 12 years
Ur-ur 6 years
Puzur-Nirah 20 years
Ishu-il 24 years
Shu-Sin 7 years
oi.uchicago.edu
HISTOKICAL VALUE 177
The reigns here quoted are those found in version S. WB, which has preserved
the first two kings and their reigns, gives Undalulu 6 years only; and Sui,
which has the last two reigns, gives Shu-S!n 24 years instead of 7. Since both
these variants can have developed out of the figures of S by simple dittog-
raphy, whereas the opposite development would be hard to account for, we
must assume that S preserves the original text. In the case of WB we must
assume that the copyist erroneously looked at the figure 6 for Ur-ur which
occurred in the line following the one he should copy; and in Sui the scribe
has strayed from the line § u - ds i n d u m u i - § u - i I m u 7 1 - as back
into the preceding i- S u - i 1 m u 24 i - as and thus got 24 years for Shu-
Sin also.
That the figures of S are indeed the original ones is shown also by a curious
correspondence pointed out by Langdon17 between their sum, 99 years, and
the 100-year reign given to Ku(g)-Baba. As we have seen earlier, the author
of the King List inserted the dynasty of Akshak between Ku(g)-Baba and her
son Puzur-Sln, apparently because he knew a synchronism between Ku(g)-
Baba and a ruler of that dynasty. This meant, however, that Ku(g)-Baba
became separated from her son by as much as 99 years, the length of the
dynasty of Akshak. To bridge the gap he prolonged her reign so that it became
one year longer than the Akshak dynasty, namely 100 years.
T H E AGADE SOURCE
This source may be reconstructed as follows:

Sargon 56 years
1 9 years
Rimush
Man-ishtushu 15 years
11
Naram-Sln 37(?) years
1
1 181 years
Shar-kali-sham 25 years
Igigi
Nanum
Imi 3 years
Elulu
Dudu 21 years
Shu-Dur ul 15 years
We have based the reconstruction on WB, which, as we saw above (pp.
26 f.), seems to give the betted text for this section. The lacuna covering the
17
OECTII 7.
oi.uchicago.edu
reigns of Nar&m-Sfn, Shar-kali-sharrl, and the four rulers of the interregnum

has to be restored from the other versions. We have accepted the 3 years for
the interregnum given by P3 and S. The reign of 25 years for Shar-kali-sharrl
follows Li and Su3+4, although it is not quite certain that thisfigureis preferable
to the 24 years of P3. The very high figure for Naram-Sin, 56 years, which
we find in Li and, as it seems, in P3 would make our individual reigns come to
a sum higher than the total which WB indicates, 181 years; we have therefore
given Nar&m-Sin a reign of only 37 years to make the reconstruction agree
with WB's total.18
CORRELATION OF THE SOURCES
The next step after the reconstruction of the sources must obviously be to
place them in correct relation to one another in time. To do this we must seek
the help of synchronisms.
U R AND AGADE
We have mentioned above that the traces left of the name of the last king
of the 2d dynasty of Ur could be restored as Ka-ku(g) and also that Rimush
of Agade in his inscriptions says that he defeated and captured "Ka-ku(g),
king of Ur." This event is mentioned by Rimush immediately after he has
stated that Enlil gave him kingship,19 and we may therefore assume that it
took place early in his reign. Allowing a year after Rimush's accession for
the young king to get settled and to prepare for his first great military under-
taking, we may tentatively place the capture of Ka-ku(g) and the end of
the 2d dynasty of Ur around Rimush's second year.20
AGADE, K I S H , AND URUK
Rimush's father, Sargon, began as cupbearer of Ur-Zababa(k) of Kish.

Later on, however, he must have become independent for he founded a city
of his own, Agade, and finally, having defeated and captured Lugal-zage-si
of Uruk, became suzerain of all Babylonia. From Sargon's inscriptions we can
conclude that the reign of his master Ur-Zababa(k) ended in a catastrophe
which left Kish ruined and partly deserted; for, after the victory over Lugal-
zage-si, Sargon states that he "restored" Kish, which must thus have been in
18
Cf. the discussion of this figure on p. 112, n. 251.
19
RISA, p. 118 xvi 52-xvii 6.
20
Note also that Rimush's campaign against Elam, which could hardly have been under-
taken before he had Babylonia itself firmly in hand, falls as early as the third year after he
had received the kingship (ibid. p. 124, No. 10 xxiii 50-55: in sa-an-tim sa-li-is-tim §a-ti
d
en4il sar-rut-tdm i-li-nic-sum)—a fact which confirms our dating of the war against Ka-
ku(g) to Rfmush's second year.
oi.uchicago.edu
ruins at that time.21 Now the only one who could thus have destroyed Kish
is obviously Lugal-zage-si, who in hisfightfor the hegemony of Babylonia must
have defeated Ur-Zababa(k) and destroyed his city in the same ruthless way
in which he had earlier destroyed other opponents. The defeat of Ur-Zababa(k)
and the destruction of his city would of course make Ur-Zababa(k),s cup-
bearer, Sargon, independent, so that he could found his own city and set up
as ruler there. We can therefore count Sargon's 56 years, which must repre-
sent his reign in Agade,22 from the fall of Ur-Zababa(k); and we must place
that event in the reign of Lugal-zage-si. Since we do not know exactly when
in that reign it occurred, we may place it at the middle, where we shall get the
smallest margin of error.
AKSHAK AND K I S H
The Chronicle Weidner contains an anecdote relating how Ku(g)-Baba,

who lived under King Puzur-Nirah of Akshak, was given "the kingship of all
lands" because of a pious deed.23 The historical kernel in this quite legendary
story would seem to be that Ku(g)-Baba of Kish vanquished Akshak and de-
throned its king, Puzur-Nirah. We can therefore assume that Ku(g)-Baba and
Puzur-Nirah were contemporaries; and since we do not know exactly when in
Ku(g)-Baba,s reign she gained her victory over Akshak, we may as above
place it at the middle to get as small a margin of error as possible.
URUK AND LAGASH
By means of the synchronisms just mentioned we can correlate the five

sources which we have reconstructed, but to facilitate comparisons with other
chronological evidence it will be practical also to fix their position in relation
to the list of rulers of Lagash which at present forms the backbone of Baby-
lonian chronology for the older periods.
When Lugal-zage-si was as yet only e n s i (k) of Umma he captured and
sacked Lagash severely.24 His opponent in Lagash, Uru-KA-gina, carries from
his second to his eighth year the title "king"; since his dates then stop abruptly
21
Poebel, PBS IV 1, p. 176: "§arru-kin, king of the land, restored Ki§ (i.e., the people
of Kis) in its (old) place. Their city (or the city) he gave to them as a dwelling place."
22
As we have seen, the sources used by the author of the King List were local date lists.
23
Giiterbock in ZA n.F. VIII 51 and 54. Giiterbock objects to Weidner's reading, Puzur-
Saân (AOF III [1926] 198; Altorientalische Studien Bruno Meissner . . . . gewidrnet....
[Altorientalische Gesellschaft, "Mitteilungen" IV (Leipzig, 1928-29)] p. 230), on the grounds
that there is not room for the sign MTJS r S a f t a n , n i r a j j ; but the remnants shown
in the copy and photograph published by him suit indeed a somewhat cramped MUS very
well.
24
SAK, pp. 56 and 58. Lugal-zage-si is called " e n s i (k) of Umma" in rev. iii 11-13.
oi.uchicago.edu
we may place Lugal-zage-si's attack there. In the King List Lugal-zage-si's reign
is given as 25 years, but as this figure comes from an Uruk source we must
assume that it refers to his reign in that city only. The sack of Lagash should
therefore be some years earlier; as we do not know exactly how many, we may
use the minimum and equate the eighth year of Uru-KA-gina with the year
before that in which Lugal-zage-si ascended the throne of Uruk, admitting
that we may be skipping some years.
From tablets dated to their reigns we know that Uru-KA-gina of Lagash
ruled 8 years and that his immediate predecessors, En-e(n)tar-zi(d) and Lugal-
anda, ruled 5 and 7 years respectively. For the time before Lugal-anda we
lack such indications; but, since the eight rulers from Gurshar to En-Anna-tum
II represent seven generations, we may ascribe the average length of a genera-
tion, thirty years, to each of them except A-Kurgal(ak) and his two sons,
E-Anna(k)-tum and En-Anna-tum I. These three together represent only two
generations, and we must therefore count only twenty years for each.
We thus get Table I.
COMPARISONS WITH DATA FROM OTHER DOCUMENTS

The table which we have drawn up cannot, of course, claim absolute exacti-
tude. In correlating the sources we often had to reckon with a small margin
of error, and in several sections we had to use average figures for the reigns.
These possible inexactitudes can, however, have influenced only the details.
The main lines of the scheme are not dependent on them and can be wrong
only if the foundation on which we have built—that the dynasties in the King
List can be pieced together into complete individual lists with reliable material
—is itself erroneous.
The comparison of the table with other chronological evidence will therefore
be a test of fundamentals. If the chronological facts to be found in royal in-
scriptions and other independent sources do not fit the picture which this
table gives, then the analysis and evaluation of the material in the King List
which we have given above must have been false. Correspondingly, essential
agreement will prove that analysis correct.
"MISSING RULERS"
The first question which such a comparison raises must concern the fact
that relatively few rulers appear in both the King List and outside material.
This has sometimes been quoted as an indication that the King List is not re-
liable (see pp. 2-3; nn. 10 and 12 above) but is in reality only to be expected in
view of its scope and nature. As we have seen, the author had as sources date
oi.uchicago.edu
TABLE I
TABULATION OF THE SOURCES AS CORRELATED
AKSHAK KISH AGADE URUK UR LA6ASH
Etana
[ca. 3 0 ]
Balih
[ca. 20]
En-me(n)-nunna
[ca. 2 0 ]
Meiam-Kishi(k)
[ca. 2 0 ]
Bar-sal-nunna
[ca.20\
Samug
[ca, 3 0 ]
. Mes-kiaq-gasher
j [ca. 3 0 ]
Tizkar
[ca. 20]
llku3
[ca. 20] En-me(r)-kar
[ca.20]
llta-sadum
[ca. 20] Lugal-banda
[ca.20]
Errme(n)-barage-si
[ca. 30] Dumirzi(d)
[ca.20]
4 Gilgames
[ca. 3 0 ]
Aka
[ca. 20]
Su....
[ca.20] Ur-Nungal(ak) -: Gurshar
30 j [ca. 3 0 ]
Mes-Anne-pada
ca. 4 0
Dadasig
-: [ca. 2 0 ]
Utul-kalamma(k) 4 Gunidu
15 [ca. 3 0 ]
Magalgalla
[(». 3 0 ]
Laba....m(?)
A'Anne-pada
9
ca. 4 0
En-nun-dara-Anna(k)i
Ur-Nanshe(k)
MES(?)-HE [CA 3 0 ]
Kalbum 36
[ca. 20]
SE-e Mes-kiag-nunna(k)
-: [ca. 20] 36 A-Kurgal (ak)
[ca.20]
Melam-Anna(k)
Lugal-ki-tun(?)
GA+SUB-nun-na 36
•: [ca. 2 0 ] E~Anna(k)-tum
Unzi [ca.20]
30
Elulu
Enbi-Eshtar 25
[ca.20] ErrAnna-tum I
[ca.20]
En-shakush-AnnaO^-:
Undalulu [ca. 3 0 ]
12 "" Lugalmu Balulu
[CA 20] En-temena
36
[ca. 3 0 ]
-fllr-ur
- 6
Puzur-Nirah
20 Ku(g)-Baba
-i [c». 30] Luqa!-kinishe-dudu -j
-*i ErrAnna-tum I I
IshO-iT Lugal-kinishe-dudu J [ca.30]
24
[ca. 16]
Puzur-STn
25 Lugal-kisai-si Lugal-kisal-si
[ca. 3 0 ] j [ca. 3 0 ]
Shu-Sin
. 7 Lugal-anda
•_ 7
En-e(n)tar-zi(d) 5
Uru-KA-gina
v-8
Ur-Zababa(k)
[ca. 2 0 ]
' Lugal-zage-si ....gi
25 [ca.35]
• * - - ^ -
* Simtrdar Sargon
30 -i 56
- i Ka-ku(g)
" [ca. 3 5 ]
Ushwatar
7
Eshtar-muti
II
lshm§-Shamash
If
Rimush ** ">'-
Nannia 9
7
Man-ishtushu
-: 15
oi.uchicago.edu
oi.uchicago.edu
lists from a limited number of cities. A few of these were fairly long, but most
of them were only short lists. His material was thus limited in both time and
place. There is, to quote an example, nothing extraordinary in the fact that
none of the many rulers of Lagash whom we know from their inscriptions re-
appears in the King List, for the author had no list from Lagash among his
sources. The same reason explains the absence of rulers from Umma, the neigh-
bor of Lagash, and from many other cities. The limitations in time must also
be taken into consideration. Inscriptions of E-Anna(k)-tum of Lagash state
that he defeated a ruler of Akshak by the name of Zuzu.25 This ruler is not
mentioned in the King List. But, as our table shows, the author's source for
Akshak began sometime within the reign of E-Anna(k)-tum, and Zuzu can
very well have preceded the first ruler mentioned in that source, Unzi. It may
even have been the very defeat of Zuzu at the hands of E-Anna(k)-tum which
brought a new dynasty, that of Unzi, to the throne in Akshak.
Just as in the case of Akshak, we know rulers of Ur who do not appear in
the Ur source used by the author of the King List, namely Mes-kalam-du(g)
and A-kalam-du(g).26 But the Ur source begins with the 1st dynasty of Ur,
and stratigraphic evidence has shown that Mes-kalam-du(g) and A-kalam-
du(g) belong to the time before that dynasty;27 it is therefore natural that we
do not find their names.28
Lastly we must consider the case of Kish. We know a certain number of
rulers who in their inscriptions designate themselves as kings of Kish but who
do not appear in the Kish source, namely Me-silim,29 Lugal. . . .,30 Mes-
Anne-pada,31 Ur-zaged,32 Lugal-tarsi,33 Sargon, Rfmush, and Man-ishtushu.34
™SAK, p. 2 0 i v 2 7 - v 8 .
28
UE II316 and PL 191 (U 11751, U 11825, U 10001, U 10002, U 10004, U 10081).
27
Woolley, UE II 218-22. The evidence of stratigraphy there presented seems to us de-
cisive and unshakable. We have seen no argument for a lower date which could be con-
sidered convincing.
28
The special case of A-Anne-pada, who probably appeared in the King List in its original
form, has been discussed in detail on p. 173, n. 12.
M
SAK, p. 160; Luckenbill, Inscriptions from Adab (01P XIV [1930]) Nos. 1 and 5.
Mentioned as king of Kish by En-temena (SAK, p. 36 n i 8-9).
80
Dec., PL 5 ter, No. 1. See below, n. 34.
KUE II 312 f. (U 13607): [me] s - a n - n 6 - p a - d a l i f g a l k i § i k i .
™SAK, p. 160.
«Ibid, and YOS I, No. 6.
34
See the inscriptions of these rulers listed in RISA, pp. 100-136. The famous spearhead
from Telloh (Die, PL 5 ter, No. 1) is inscribed l u g a l . . . . l u g a l k i § i . The traces
following the initial l u g a l are at present undecipherable. In the first reproduction
oi.uchicago.edu
The reason why these rulers are missing is easy to see. It is well known that
the title "king of Kish" was so highly prized that rulers of other cities preferred
to style themselves "king of Kish" rather than king of their own city if they
could claim that Kish acknowledged their suzerainty. The sources of the King
List, on the other hand, were local date lists concerned with local rulers only.
In such a source we should therefore find the local rulers of Kish and not the
suzerains. Now we can show from other material that four of the kings men-
tioned above actually were only such suzerains and belong elsewhere, namely
Mes-Anne-pada, whose city was Ur, and Sargon, Rimush, and Man-ishtushu,
who belong in Agade. And since the others, Me-silim, Lugal. . . . , Ur-zaged,
and Lugal-tarsi, have no more claim to connection with Kish than those four,
it is natural to assume that they are missing in the Kish source for the same
reason; they are suzerains and not genuine rulers.35
In reality there is only one case where we have reason to assume that one
of the sources has omitted a ruler whom we could have expected to find. We
possess the inscription of a certain Utuk who styles himself "e n s i (k) of
Kish" and dedicates a vase to the city god Zababa.36 In this case, therefore,
we have a genuinely local ruler; and we must accordingly assume that the
Kish source as we have it has lost his name in the course of tradition, as could,
naturally, happen.
of the spearhead (RA III [1896] 53, Fig. 1) they were rendered ru r 1 , which would
give the name u r - l u g a l ( l u g a l - u r with transposition of signs) l u g a l k i § i ,
"Ur-lugal (ak), king of Kish." Since this reproduction was based on photographs
alone (Heuzey in RA III 54, n. 2), whereas the reproduction in Dec. (cf. also RA IV [1897]
111, Fig. 18) is made from the monument itself, we can hardly accept its reading, however
tempting an identification of its Ur-lugal(ak), king of Kish, with Ur-lugal(ak), son of Gil-
games, would be. (The date of the lance—it was found discarded!?] at the Ur-Nanshe(k)
level [RA IV 111]; note also absence of determinative after k i hi—is clearly before Ur-
Nanshe(k); the Gilgames text mentioned on p. 145, n. 20, suggests that Kish at Gilgames'
time acknowledged the supremacy of Uruk.)
We have not included any reference to a 1 - [ ] 1 u g a 1 k i h i ki in the Stele of Vul-
tures rev. xii, since, as Poebel has pointed out (PBS IV 1, p. 167, n. 1), it is "entirely un-
certain" that the sign a 1 - begins the name of a king. We are also disregarding the name
l a - . . - . . - § i - i mentioned by Langdon, Excavations at Kish I 5, until full publica-
tion of the inscription in which it appears makes it possible to check the reading.
35
That this explanation holds good can be directly shown in the case of Ur-zaged, who
in his inscription styles himself l u g a l k i § i ki l u g a l [ . . . . ] , "king of Kish, king of
" Here the title "king of Kish" is clearly placed before the original title of the king
because of its greater importance. It should be mentioned that Langdon also (OECT II
6 f.) considers the explanation given above as a probable reason why certain "kings of
Kish" are not mentioned in the King List. Cf. also Landsberger in OLZ XXXIV (1931)
121, n. 3: ". . . . ist die Lokalisierung des Me-silim in Ki§ nicht mehr zwingend."
*SAK9 p. 160.
oi.uchicago.edu
We have thus reviewed the question of the relative scarcity of overlaps be-
tween the material in the King List and the contemporaneous inscriptions.
We see too that there is only a single case in which we could actually have ex-
pected to find a "missing" ruler mentioned in the King List,37 and that one
case is easily explicable as an accidental omission in the King List or in its
source in the course of tradition. There is accordingly no disagreement here
between the evidence of the inscriptions and the sources of the King List.
SYNCHEONISMS
We may now turn to the more positive and definite part of the comparison
and consider the outside evidence which has a bearing upon the time and
relative position of rulers who do appear in the sources which we have corre-
lated in our table.
EN-TEMENA-LUGAL-KINISHE-DUDTJ
An inscription of En-temena states that this ruler concluded "brotherhood"

with the e n s i (k) of Uruk, Lugal-kinishe-dudu.38 These two rulers must
therefore have been contemporaries. Comparing our table, we see that the
reigns of En-temena and Lugal-kinishe-dudu partly overlap. There is thus at
this point full agreement between the table and the outside evidence.
EN-SHAKUSH-ANNA(K)-ENBI-ESHTAR
An inscription of En-shakush-Anna(k)39 relates that he defeated and cap-

tured Enbi-Eshtar, king of Kish. Reconstructing the Kish source used by the
author of the King List, we found that the name of the king before Lugalmu
should be read Enbi-Eshtar (gloss based on form Inib-Eshtar; see p. 169);
now that the sources have been correlated, we see that the reign of this ruler
corresponds in its last part to that of En-shakush-Anna(k).40 There is thus
here also perfect agreement between the table and the outside evidence.
37
Since we are here concerned with the question of whether or not the sources of the King
List were so elliptical that their evidence cannot be used, we have not counted the two
cases in which existing traces show that a ruler originally appeared in the list but was lost
during the course of tradition, namely Mashda (pp. 21-22) and A-Anne-pada (p. 93, n. 145),
nor those cases where rulers known from inscriptions can plausibly be restored in existing
lacunas, such as the rulers of Ur II and certain Gutian kings. Comments on these cases
will be found in the notes to the text of the King List in chap. iii.
38
Gadd in RA XXVII (1930) 125 f.; Barton in JAOS LI (1931) 262-65.
39
PBS IV 1, p. 151.
40
The King List uses a late form of the name, En-shakash-Anna(k); see p. 171.
oi.uchicago.edu
EN-SHAKUSH-ANNA(K)-ELULU
An inscription of En-shakush-Anna(k) 41 tells that this ruler was son of
E -1 i -1 i - n [a]. Elili(n) appears as E -1 i -1 i , king of Ur, in his own
brick inscription42 and as E -1 u -1 u in the 1st dynasty of Ur in the King
List.43 Comparing our table, we see that Elulu's reign immediately precedes
that of En-shakush-Anna(k) in time, as should be expected on the basis of the
synchronism.44 There is thus again agreement between the table and outside
evidence.
A-ANNE-PADA-ABOUT U R - N A N S H E ( K )
c
Excavations in al- Ubaid produced a few short inscriptions of the ruler
A-Anne-pada of Ur. From the orthography of these inscriptions and of a few
others found with them it is possible to date A-Anne-pada fairly exactly in
relation to the rulers of Lagash. Fundamental here is Gadd's thorough study
of the material in Ur Excavations I 128-37, which calls attention to the fol-
lowing significant points:
a) The A-Anne-pada inscriptions use the verbal infix - § & - instead of - § i - .
The change from - § & - to - § i - begins with En-temena and ends with Uru-
KA-gina.
b) The A-Anne-pada inscription TO 160 is written on a marble tablet shaped like
a plano-convex brick, which recalls the inscribed bricks of Ur-Nanshe(k) and a
contract of E-Anna(k)-tum written on a brick.
c) Signs constituting a word are still written in arbitrary order; this usage practically
disappears before E-Anna(k)-tum.44a
d) The signs on stone most resemble those of Ur-Nanshe(k), but these in turn do not
vary much in the course of his dynasty. The signs on clay occupy a position be-
tween those of the Fara tablets and those of the En-temena-Uru-KA-gina tablets,
but seem to resemble the latter more closely.
41
W. K. Schileiko, Votivi Natpisi Sumerinski, p. 11, 1. 5, PL III 3, quoted by Unger in
RLA II 309; cf. ibid. p. 490. We do not have access to Schileiko's book.
« S. Smith in JRAS, 1932, p. 306.
43
The identifications were already made by Unger in RLA II 490.
44
That En-shakush-Anna(k) was king of Uruk, as shown by the King List and supported
by his title (see Poebel, PBS IV 1, p. 153), whereas his father Elulu, as shown by his title
and by the King List, ruled in Ur, suggests that Elulu had one of his younger(?) sons (En-
shakush-Anna(k)) made ruler of Uruk when the dynasty there came to an end (note that
En-shakush-Anna(k) begins a new dynasty, and cf. p. 162). That the same family would
thus have ruled both in Ur and in Uruk throws light on the history of the following times,
explaining satisfactorily why the kingship of Ur on BakuVs death passed to Lugal-kinishe-
dudu, the successor of En-shakush-Anna(k) in Uruk.
44a
Sporadic examples of the old usage in texts which otherwise write the signs in their
true order can be found in later times (examples have been collected above on p. 172, n. 8).
This does not, however, detract from the value of that usage as a dating criterion as long as
we judge by the general usage of the text or texts to be dated, not by isolated instances.
oi.uchicago.edu
The evidence of these texts Gadd summarizes as follows:

a) gives (probably) before En-temena
6) gives Ur-Nanshe(k) and E-Anna(k)-tum
c) gives before E-Anna(k)-tum
d) gives (probably) about Ur-Nanshe(k) but later than the Fara tablets.
He draws the conclusion that "so far, then, as the successive tests have given
tangible results, the agreement is striking, and the date indicated is somewhat
before, but not very long before, the reign of Ur-Nin& [Ur-Nanshe(k)]." To
a very similar date, somewhat before Ur-Nanshe(k), Landsberger comes in
OLZ XXXIV (1931) 117-26.45
46
V. Christian's treatment of the criteria in ZA n.F. IV (1929) 234 f. is not satisfactory
in method, and his conclusion that the data favor a date within the period from E-Anna(k)-
tum to En-temena cannot be accepted. The diagram in which Christian arranges the cri-
teria tends to obscure their true significance. Thus this diagram gives the tablet form as
indication for the period from En-temena to Uru-KA-gina. In reality, however, the tablet
form in question appears already in the Fara tablets, so that this criterion is compatible
with any date from somewhere before Ur-Nanshe(k) down to Uru-KA-gina. A similar case
is the criterion represented by the shape of the signs on clay. We have only two groups of
texts with which we can compare the A-Anne-pada signs: the Fara tablets and the Uru-KA-
gina tablets. Gadd has shown that the A-Anne-pada signs resemble the signs used in the
En-temena-Uru-KA-gina period more than those used in the Fara tablets. Since we do not
know when in the period which separates the Fara period from the En-temena-Uru-KA-gina
period the sign forms which we find in the latter developed, the fact that the A-Anne-pada
signs resemble the En-temena-Uru-KA-gina signs more than the Fara signs only proves
that A-Anne-pada is later than Fara, i.e., belongs somewhere in the time from after the Fara
period down to the end of the En-temena-Uru-KA-gina period. It is therefore not correct,
as Christian does, to use this criterion as indicating the period from En-temena to Uru-KA-
gina only, for En-temena does not represent an upper limit.
The way in which Christian uses his diagram is likewise open to criticism. The diagram
is arranged chronologically, and each of the various criteria is represented by a vertical black
line. Counting how many such black lines cross the horizontal bands allotted to E-Anna(k)-
tum and En-temena, Christian states: "Aus dieser Aufstellung ergibt sich, dass von acht
Kriterien sechs fur die Periode Entemena's, funf fur die Eannatum's sprechen, nur eines
scheint eine friihere Zeit zu verlangen." But this statistical method can never lead to cor-
rect results, for the important thing is to find a period at which all the criteria overlap or
toward which they approximate in time. Thus if we had an inscription which showed three
orthographic or stylistic peculiarities which lasted down to the time of Ur-Nanshe(k) and then
went out of use, and ten such peculiarities which were known only from documents of the
time of E-Anna(k)-tum and later rulers such as En-temena and Uru-KA-gina, we should
obviously have to conclude that the inscription was written in the period between Ur-Nan-
she(k) and E-Anna(k)-tum, i.e., under A-Kurgal(ak), because only this intermediate period
could have produced an inscription which had both old and new features. It would obviously
be wrong to state that because ten of the criteria "spoke for" the period of En-temena, where-
as only three "called for" an earlier date, the inscription should be dated to the time of
En-temena.
Lastly we must mention Christian's reasons for rejecting as a criterion the transposition
oi.uchicago.edu
This result is undoubtedly correct in all essentials and can be modified only
within narrow limits. There is reason, however, to call attention to a criterion
which Gadd has rejected as insignificant. The A-Anne-pada inscriptions use
suffixes; and, as is well known, suffixes, although common in the inscriptions
of E-Anna(k)-tum, do not appear in those of Ur-Nanshe(k). In Gadd's opinion
it is doubtful whether this peculiarity of Ur-Nanshe(kys inscriptions is a true
indication of archaism, since "no great change in the usage of the language can
be supposed to have taken place within the narrow limits of two generations."
He therefore rejects the criterion. In reality, however, the change in question
does not reflect a change in language but is merely a further step in that long
process which brought the system of writing ever nearer to the spoken word.
It is thus merely an orthographic change, which of course would not need so
much time as a change in language. Since Ur and Lagash are situated near
each other and must have had close cultural connections, we can obviously—
even if we admit that Ur may have been the more progressive—hardly assume
that the improvement would take more than a generation at most to penetrate
to Lagash. GacUTs dating "somewhat before, but not very long before, the
reign of Ur-Nin& [Ur-Nanshe(k)]" is therefore better modified to "approxi-
mately the time of Ur-Nanshe(k)." This agrees very well with the places
which the two rulers take in our table, and there is thus once more agreement
between the two groups of evidence.
GILGAMES-AKA
An epic text deals with a war between King Aka of Kish, to whom Uruk
owed allegiance, and Gilgames, who defeated Aka and liberated Uruk (p. 145,
n. 20). We have stated earlier that the material of epics and historical legends
must be fairly old and can hardly postdate by many generations the historical
persons concerned. We can therefore accept the synchronism here presented.
Comparing it with the table, we see that it agrees perfectly. The reign of Aka
falls within that of Gilgames by several years.
of the signs, the criterion which suggests a date to the time of Ur-Nanshe(k) or earlier:
"Die Abweichung von der regelmassigen Zeichenfolge, um die es sich hierbei handelt, findet
sich auf dem Kurlil-Torso, der aus stilkritischen Grunden etwa der Periode En-temena's
zuzuweisen sein diirfte. Das aus der Inschrift geschopfte Argument, das ein hoheres Alter
zu fordern scheint, wird dadurch wohl hinfallig." Since it can safely be stated, however,
that—with our present limited knowledge of Sumerian statuary—the criteria of style
which Christian quotes (op. cit. p. 236) are of exceedingly doubtful value as dating evi-
dence, we should only be rejecting a safe criterion in favor of a very unsafe one if we followed
him.
oi.uchicago.edu
TYPES OF PERSONAL NAMES
TO a time somewhat before Ur-Nanshe(k) of Lagash belongs a group of

tablets of economic and administrative character found in Fara. 46 These tab-
lets mention a large number of persons and give a good impression of the types
of proper names then in vogue. The next large group of tablets which enables
us to study Sumerian proper names comes from the period of Lugal-anda,
En-e(n)tar-zi(d), and Uru-KA-gina of Lagash, and the names of this later group
show that considerable changes have taken place: certain names do not occur
any more, elements which were frequent in the Fara period are now no longer
current, etc. We may mention some of these changes:47
a) In the Fara texts the name A k (or A k a) is common. Deimel, who has col-
lected the names in these texts, lists seventeen occurrences. This name has dis-
appeared completely in the Uru-KA-gina period.
b) The Fara texts know an element b a r , occurring in such names as A b - b a r -
t a b , B a r - m e s ( ? ) - r a (i.e., M e s - b a r - r a ) , B a r - s i - s a r , dS u d -
n u - b a r , £ - u r u - b a r , E d i n - b a r , U r - b a r - d a , K a - b a r , Lti-
b a r - z i - d a , and M e s - b a r - r a . This element has disappeared in the
names of the Uru-KA-gina period, where we find b a r only in the compound
verb i g i — b a r , "to see."
c) Another element which is fairly common in the Fara names is PA4-GIS-BIL or paPA4-
pa
GI&-BIL, read p a4 - g i § - g eio or p a4 - g i § - g do, "sprout of a new tree,"
which occurs in the names P a 4 - g i § - g e i 0 - & - n u - k u § (written g i & - p a p -
BIL - d - n u - k u s " and p a p - g i § - BIL - & - n u - k i i § ) , P a4 - g i § - g eio -
k a 1 a m - d U10 (written g i s - p a p - BIL - k a l a m - d Ui0) or pa P a4 - g i § -
g do - k a 1 a m - d U10 (written p a - g i s - B f L - p a p - k a l a m - d u i o and
p a p - g i § - BIL - p a - k a l a m - d Uio), P a4 - g i § - g eio - d U10 (written p a p -
g i § - BIL - d Uio), P a4 - g i § - g eioga, and P a4 - g i s - g eio - dd u g u dmuien
(written d p a p - g i § - B i L - d u g u d mu5en ). This element also has disappeared
before the Uru-KA-gina period.
d) A very common element of the Fara names is k a 1 a m , which appears in A m a r -
k a l a m , G i g - l a (?) - k a 1 a m , E n - k a l a m ( ? ) , E n - k a 1 a m - d u w ,
E n - k a l a m - x ( L M , No. 503), E n - n u - k a l a m - s e , En-nin-
k a l a m , U r - 6 - k a l a m , K a l a m , K a l a m - d Uio, K a l a m - k u - l i ,
K a l a m - x(LAK, No. 503), L u g a l - a - k a l a m , Nin-kalam-?,
S a l - e n - k a l a m ( ? ) , and P a4 - g i § - g eio - k a 1 a m - d Uio or paP a4 -
g i s - g eio - k a 1 a m - d Uio (both written as above). In the Uru-KA-gina period
46
Deimel, Die Inschriften von Fara I and III (WVDOG XL and XLV [1922 and 1924]).
On the date of the texts see Vols. I 4-5 and II l*-2*; Deimel in Orientalia No. 6 (1923) pp.
51-54; Landsberger in OLZ XXXIV (1931) 123.
47
The changes were determined through a comparison of the proper names from the
Fara tablets listed by Deimel, Die Inschriften von Fara III 18*-48*, with the onomastic
material contained in the texts from the En-temena-Uru-KA-gina period which Deimel has
transliterated in various numbers of Orientalia from No. 14 to No. 44.
oi.uchicago.edu
this element is so rare that we can quote only two examples, E n - k a l a m ( ? ) -

m a (?) and M u - n i - k a l a m - m a .
e) Several Fara names contain the element m e n : M e n - i u - g f d , M e n -
m - . . - d u , M e n - p a - e , M e n - m u , and L u g a l - m e n . In the Uru-
KA-gina period this element is limited to the stereotyped connection m e n -
z i - d i m 4 and appears outside it only once: M e n - e n - t u .
/) The element m e s is extremely frequent in the Fara texts. Deimel lists twenty-
two different names beginning with this element. In the Uru-KA-gina period it
has practically disappeared and is known to the writer from a single name only,
Mes-zi.
g) As frequent as m e s is the element n u n . Deimel lists about fifty names in
which it occurs. In the Uru-KA-gina period it has disappeared and is found only
in names composed with old names of deities or buildings, such as d f i § - n u n ,
fi-ka-nun-di, G i - n u n , G f r - n u n , and I r - n u n .
Comparing the rulers1 names which appear in the King List, we observe
that several of the earlier names in the traditions of Kish, Uruk, and Ur are
composed with the elements which we have here mentioned, elements which
were common in the Fara period but had disappeared already in the time of
Uru-KA-gina. This observation is of considerable interest. It shows first that
we actually have genuine and old tradition in the King List, for such names
naturally could not have been invented in late periods. Secondly it gives us a
means of testing the validity of our table. We have seen that the elements
mentioned flourished in the Fara period, that is, shortly before Ur-Nanshe(k),
and died out in the time of Uru-KA-gina. If our arrangement of the various
traditions is correct, we should expect to find that names with such elements
would occur in our table rather frequently in the period just before Ur-Nan-
she(k), would then become scarce, and would finally disappear before we get
down to Uru-KA-gina.
This is indeed exactly what we do find in the table. In the period before
Ur-Nanshe(k) the table has E n - m e n - n u n - n a , Bar-sal-nun-
na, E n - m e n - b a r a - g e - s i , A5-ka, M e s - k i - a g - g a - § e - e r ,
Gis-bil-ga-mes,48 Utul-kalam-ma, En-nun-dara-
48
GiU-BfL-GA, the first element of the name Gilgames (written dGis-BfL-GA-MES, dGis-
BfL-GfN-MES, etc.; see p. 89, n. 128), cannot, of course, be separated from the element *>ap a4 -
GIS-BIL-GA, pa p a4 -Gi§-BfL, or p a4 -GIS-B!L of the Fara names but must represent an ab-
breviated form of it. The meaning of this element is reasonably clear. P a* ( A r , older
^ and perhaps «$ [LAK, Nos. 229 and 42]) is the picture of a sprouting seed of grain <*>,
as shown by the value b u 1 u g or d i nu , "malt," i.e., sprouting grain, which the sign
has when it is repeated: A*T /*r • This idea, "sprouting seed," obviously underlies the value
p a p , "father" (§L, No. 60.2); the father is called "sprouting seed" as the germ from
which a family springs up. The idea is still more clearly expressed in p a -ai§-Bh>GA,
"forefather" (En-temena Cone [SAK, p. 38] i 35; §L, No. 295.131, has incorrectly BIL),
also abbreviated as pa-Bn>GA:a& abi: "grandfather," "forefather" (§Lf No. 295.109),
oi.uchicago.edu
a n - n a , and M e s - a n - n 6 - p & - d a . Then, however, such names grow

scarce; in the generation which follows Ur-Nanshe(k) we find only two,
G A + § U B - n u n - n a and M e s - k i - a g - n u n - n a . After that time
there is not a single name which is composed with any of these elements.
Once again we have found our table to be in agreement with other chronologi-
cal evidence.49
The test has thus on all points confirmed our previous results. We have
compared the material of the King List with that of historical inscriptions
and found that there is only one case in which the King List seems to have
missed a ruler whom we should expect to find in it, and this one case may be a
simple omission in the course of tradition. We have singled out and correlated
p a4 -Biii~Qi:a-bi a-bi (SL, No. 60.54), etc., which must represent p a - g i s - g eio - (a) k -
a (m), "who was (- a m) the germ (p a) of (- a k) a new (g eio) tree (g i §) ." The
ancestor is considered as the germ, the acorn, from which the "family tree" has grown.
That this explanation is correct is also shown by the early names themselves, for we have
the parallels p a 4 - g i s - g e i o - £ - n u - k u s : a - n u - k t i § ( < a - d - n u - k u s ) and
p a4 - g i s - g eio - k a 1 a m - d UIQ : a - k a l a m - d u i o , in which p &* - g i s - g eio,
"germ," "(fore)father," varies with a , which also means "germ" (lit., "semen virile")
and "father." The name Gilgames should thus in its full form have been p a« - g i § -
g eioga - m e s , "man (m e s) who is germ of a new tree," i.e., "a man who is to become
originator of a family."
49
One further point concerning the Fara names should be mentioned here, although it
means anticipating the results of our next chapter, which shows how the above synchronistic
list is related to absolute chronology. In Zeitschrift fur dgyptische Sprache und Alterturns-
kunde LXXI (1935) 89-106 Scharff has discussed a series of finds from Egypt which give
evidence of early relations between that country and Mesopotamia. The series in question
can be dated on the Egyptian side to the end of the predynastic period; it stops around the
time of Menes (ibid. p. 93, n. 1). According to Egyptian chronology that would be in the
centuries just before 3000 B.C. (Meyer, Die altere Chronologie . . . . , pp. 68 f., puts the ac-
cession of Menes at 3197 with margin of 100 to 200 years; Scharff, op. cit. p. 90, n. 3, favors
a lower date, shortly before 3000: "Nach meiner Uberzeugung liegt nach wie vor keinerlei
Anlass dazu vor, mit Menes wesentlich liber 3000 hinaufzugehen").
On the Mesopotamian side the affinities of the series extend from the final phase of the
Uruk period through Jamdat Nasr down to Early Dynastic II, the latest parallels being
with the seal impressions from Fara which date from that period (Scharff, op. cit. pp. 102 f.;
Frankfort, Progress of the Work of the Oriental Institute in Iraq, 1984/35 [01C No. 20 (1936)]
p. 42 and Chronological Table). Now the approximate time when Early Dynastic II ended
can be fixed by means of our King List; for the tablets with the characteristic Fara names
discussed above belong to the immediately following period, Early Dynastic III (ibid.
Chronological Table), and in the King List these names begin to appear with En-me(n)-
nunna. Early Dynastic II should thus have ended slightly before En-me(n)-nunna; and
this ruler reigned, as shown by the King List correlated with absolute chronology (see Table
II), around 3050 B.C. There is thus a very noticeable agreement between Egyptian chronol-
ogy, which places the end of the interrelations in the centuries immediately before 3000 B.C.,
and our Mesopotamian chronology, which places the end of those interrelations in the period
which ended ca. 3050 B.C.
oi.uchicago.edu
190 T H E STJMERIAN K I N G L I S T
the individual sources in the King List on the basis of a single set of synchro-
nisms. Comparing the resulting table with other synchronisms
Gilgames-Aka
A-Anne-pada-about Ur-Nanshe(k)
En-shakush-Anna(k)-Elulu
En-shakush-Anna(k)-Enbi-Eshtar
En-temena-Lugal-kinishe-dudu
we saw that in the table the rulers in question were actually placed correctly.
Lastly, we have shown that a number of royal names appearing in the table
are composed with elements which were current in the Fara period but died
out before the time of Uru-KA-gina. Noting the distribution of these names
in the table, we saw that, as should be expected if the table is correct, they
center around the time before Ur-Nanshe(k), then grow scarce, and disappear
completely before we get down to the period around Uru-KA-gina.
oi.uchicago.edu
VI
RESULTING CHRONOLOGY
It remains to tie in the reconstruction and correlation of the King List with
the data of absolute chronology. To that end we must examine briefly extant
possibilities of finding reliable fixed points of absolute chronology in older
Mesopotamian history, and we must consider the sections of the King List
and of additions to it which link the main body of its evidence with such
fixed points.
RELEVANT FIXED POINTS OF ABSOLUTE CHRONOLOGY1

RECKONING FROM LATE CHRONOLOGICAL LISTS
The general trend of recent years has been to choose as basis the thoroughly
established dates of late Assyrian times and to reckon back from them toward
the period of Hammurabi by means of Assyrian and Babylonian king lists,
supplementing their data with such occasional chronological evidence as can
be found in other historical texts. The material falls into two large groups,
Assyrian and Babylonian, which we may discuss separately. 2
The Assyrian material consists in the main of king lists of various types,
Zmmu-lists, and chronological statement^ in historical inscriptions. Synchro-
1
From the extensive literature bearing on these problems we may quote F. X. Kugler,
Sternkunde und Sterndienst in Babel II 2 (Munster in Westf., 1912); Weidner, Studien zur
assyrisch-babylonischen Chronologie und Geschichte auf Grund neuer Funde (MVAG XX 4
[for 1915; 1917]) and Die Konige von Assyrien; neue chronologische Dokumente aus Assur
(MVAG XXVI 2 [1921]); Kugler, Von Moses bis Paulus (Munster in Westf., 1922); S. A.
Cook in Cambridge Ancient History I (1923; 2d ed., 1924; reprinted, 1928) 145-56; C. Sehoch,
Ammizaduga (Berlin-Steglitz, 1925); P. Schnabel, "Zur astronomisehen Fixierung der altba-
bylonisehen Chronologie mittels der Venustafeln der Ammizaduga-Zeit," ZA n.F. II (1925)
109-22; Meyer, Die dltere Chronologie Babyloniensf Assyriens und Agyptens (1925; 2d ed.,
1931); Weidner, "Die grosse Konigsliste aus Assur," AOF III (1926) 66-77; Thureau-Dan-
gin, "La chronologie des trois premieres dynasties babyloniennes," RA XXIV (1927) 181-98;
Sehoch, Die Ur-Finsternis (eine Hypothese!) (Berlin, 1927); S. Langdon, J. K. Fothering-
ham, and C. Sehoch, The Venus Tablets of Ammizaduga (London, 1928); P. V. Neuge-
bauer, "Zur Frage der astronomisehen Fixierung der babylonischen Chronologie," OLZ
XXXII (1929) 913-21.
2
The following short discussion naturally does not pretend to treat exhaustively the
numerous and knotty problems in which the late chronological material abounds. It merely
intends to blaze a single trail through the jungle: to determine by applying the principles
of source criticism which of our sources must be considered most reliable and to build on them
only.
191
oi.uchicago.edu
nisms found in historical inscriptions, chronicles, synchronistic lists, etc. offer

possibilities of connecting these data with data from the lists of Babylonian
rulers. In judging the value of this material, we must consider first the basis
of Assyrian chronology, the limmu system. Assyrian chronology, as is well
known, was based on eponym lists in which each year was represented by the
name of the person officiating as eponym for that year.3 It is obvious from
this that records going back for several centuries, containing many hundreds
of names, must have been fairly bulky and must have occupied a considerable
number of tablets. For that reason we can be sure that on the whole relatively
few copies of the complete eponym list existed and that damage to the records
—especially prone to happen in view of their bulk—would be difficult to repair
for want of duplicates. With fairly bulky records, likely to be damaged, and
with few duplicates to fill in the lacunas which arose, Assyrian records should
show a gradual shrinking of figures as in the course of time the few standard
sources became more and more defective and more and more eponym names
disappeared in lacunas. This is indeed what we actually find. Our oldest chron-
ological source is a statement in an inscription of Shalmaneser I.4 According
to this, Shamshl-Adad I had rebuilt the Ashur temple 580 years before Shal-
maneser himself, and a still older ruler, Erishum, had rebuilt it 159 years be-
fore Shamshl-Adad. Already under Shalmaneser's son and successor, however,
we find somewhat lower figures: Tukultl-Ninurta I reckons for the larger
span of time from Ilu-shuma, father of Erishum, down to his own time only
720 years.5 Still later, under Esarhaddon, the distance from Erishum to
Shamshl-Adad has shrunk from the 159 years given by Shalmaneser to only
126 years, and similarly the 580 years between Shamshl-Adad and Shalmaneser
have become only 434, a total loss of 179 years.6 The source which the Assyrian
kings consulted for such information, the standard copy or copies of the
eponym list, had clearly become steadily more damaged as time went on, so
that the scribes of Esarhaddon found far fewer itmrau-names preserved than
did the earlier scribes of Shalmaneser I. Their additions therefore gave very
much lower figures. Since we cannot assume that it WBS only with Shalmaneser
I that the gradual deterioration of the older parts of the eponym list here
reflected began, it is probable that the figures which he gives are already on
the short side.
If we must thus suspect our oldest evidence, we should obviously approach
8
See e.g. Meyer, Geschichte des Altertums I 2 (3. AufL; Stuttgart und Berlin, 1913) § 324;
Cook, op, a t pp. 148 f.
« KAH I, No. 13 iii 32 ff.
8 6
KAH II, Nos. 48 and 59. KAH I, No. 51, and II, No. 126.
oi.uchicago.edu
RESULTING CHRONOLOGY 193
that of the king lists, which would draw from the same source at still later
stages, with even more caution. For our purpose this means that we cannot
hope to fix the dates of old dynasties such as that of Hammurabi by reckoning
back on the evidence of the Assyrian king lists to the Assyrian side of syn-
chronisms for the older period such as Ilu-shuma-Sumu-abum, Erishum-Su-
mu-la-El, or Shamshl-Adad-Hammurabi. We must realize that we can rely
on the figures of these Assyrian lists in their later part only, approximately
from Ashur-iiballit downward, where the shrinking should be less pronounced
and where synchronisms with Egyptian chronology give us some means of
checking the figures.6*
Turning to the Babylonian material, wefindit to consist—much as does the
Assyrian—of king lists, to wit the famous Babylonian King List A,7 and of oc-
casional chronological information given by historical or similar inscriptions.
King List A is a relatively late source; it lists Kandalanu and possibly had
still later rulers. Of importance in judging its value as a source is the fact that
Babylonian chronology was based on the regnal years of the kings. Records—
even those going back for centuries—would thus be compact, and it would not
be too great a task to copy them. We can therefore assume that more copies
of the standard Babylonian chronology were about than was the case with the
Assyrian, so that lacunas could comparatively easily be filled in from dupli-
cates. The evidence of the Babylonian king list will therefore inspire some-
what more confidence than that of the Assyrian. We should not forget, how-
ever, that the list is a late document and that we have no other copies to check
its text. In the long period of tradition from Hammurabi to Kandaltou or
later many scribal errors in figures and totals may have crept in, and we have
few means of detecting them.
Among the material from inscriptions we should mention the information
given on a boundary stone from the fourth year of Enlil-nadin-apli8 that 694
years had passed from Gulkishar of the Sea Land to Nebuchadnezzar I. As
our earliest Babylonian source this passage must rank comparatively high. A
••For this reason we have little confidence in the recent tendency to explain the syn-
chronism Shamshl-Adad I-Hammurabi by pulling Hammurabi down to the date assigned
to Shamshl-Adad on the basis of the Assyrian king lists (cf. e.g. Albright in American Schools
of Oriental Research, Bulletin No. 69 [1938] pp. 18 ff.). Reasons for doubting the com-
pleteness of the older parts of the Assyrian king lists have also been expressed by Lewy in
ZA n.F. IV (1929) 95-105. We do not agree with all of his conclusions, but see esp. pp.
104 f. and his correct observations on the variant figures of Esarhaddon's scribes as inten-
tional corrections of Shalmaneser's figures (ibid. pp. 102 f.).
7
CT XXXVI (1921) Pis. 24-25.
8
BE I 1 (1893) Pis. 30-31, No. 83.
oi.uchicago.edu
number of references from late times in inscriptions of Nabonidus need not be

specially discussed, since their unreliable character is generally recognized.
Our survey has thus shown that, while the Assyrian material is of rather
doubtful value and should be used only for the later periods, the Babylonian
material inspires somewhat more confidence, although here also we must
reckon with possible errors in figures and totals. Foremost among the Baby-
lonian material ranks the information given by the boundary stone, next the
evidence of King List A. To get the most reliable basis for our computation
we should therefore use the material in the following order of preference:
boundary stone, King List A, Assyrian material.
The information on the boundary stone reaches down to Nebuchadnezzar I
only. To get connection with the established dating of later times our second
best source, King List A, is of no help, since it is too broken to give a con-
tinuous series. Even the name of Nebuchadnezzar himself is lost. We must
accordingly use the Assyrian lists, which for this part, after Ashur-uballit,
must be considered relatively reliable. According to the text VAT 102819 the
Assyrian king Ashur-r&sh-ishi I fought with Ninurta-nadin-shumati of Baby-
lon, and according to the Synchronistic History10 Ashur-r&sh-ishl also fought
with the successor of Ninurta-nadin-shumati, Nebuchadnezzar I. He was, ac-
cording to the synchronistic lists KAV, No. 12, and Assur 14616c,11 even con-
temporaneous with the successor of Nebuchadnezzar, Enlil-nadin-apli. Ashur-
r6sh-ishl reigned 18 years; Nebuchadnezzar, as shown by a boundary stone
dated to his 16th year, 16+x years.12 Since we must allow one or two years
at the beginning of Ashur-r6sh-ishfs reign for his war with Ninurta-nadin-
shumati, the 16 years which represent the minimum for Nebuchadnezzar's
reign will take us down to around the last year of Ashur-r^sh-isht. It therefore
seems likely that the death of Nebuchadnezzar, the accession of Enlil-nadin-
apli, and the death of Ashur-r&sh-ishi all happened in the same year. Assyrian
king lists show that the last year of Ashur-rMi-ishi was 1113.13 Adding to this
9
Weidner in AOF IV (1927) 213-17.
" CT XXXIV (1914) PI. 39:2-13. « AOF IV 70 f.
12
W. J. Hinke, Selected Babylonian Kudurru Inscriptions ("Semitic Study Series" XIV
[Leiden, 1911]) pp. 21-27 v 26.
13
This statement rests in part on information from the Khorsabad king list, which Prof.
Poebel is preparing for publication. We are indebted to him for permission to quote this
material as basis for the date of Ashur-r6sh-ishf. We may mention that this date agrees
with the so-called "Bavian date," which has long been considered a cornerstone for Assyrian
chronology. Sennacherib states (III R 14:48-50) that when he had conquered Babylon he
brought back ' 'after 418 years" the statues of Adad and Shala of the city Ekallate which
Marduk-nHdin-ahe had taken to Babylon in the time of Tiglathpileser. It is uncertain to
what point Sennacherib's scribes counted back. Contrary to what is generally taken for
oi.uchicago.edu
RESULTING CHKONOLOGY 195
the 694 years of the boundary stone, we get 1807 as the last year of Gulkishar.
Using then thefiguresfor Gulkishar and his predecessors given in King List A
(55+24+16+56+60 = 211), we arrive at 2017 as the first year of Iluma-ilum.
Now we know that Iluma-ilum was contemporaneous with Samsu-iluna and
Abi-Eshuh of the 1st dynasty of Babylon.14 His rebellion against Samsu-iluna
is in all probability referred to in the formula for the 10th year of that ruler:
"Year: 'He . . . .ed the hosts of Idamaraz, Yamutbal, Uruk, and Ism' "; but
the occasion for it would seem to have been the Cassite invasion referred to in
the formula for the preceding year, the 9th. Since the date formulas presum-
ably are a year behind the events to which they refer, we should place the Cass-
ite invasion and the beginning of Iluma-ilum^ rebellion in Samsu-iluna's 8th
year.15 With Samsu-iluna's 8th year as 2017 his 1st year becomes 2024, and
the 43-year reign of his father, Hammurabi, who preceded him, can thus be
dated to 2067-2025 B.C.
granted, we consider it highly unlikely that they possessed information as to the exact year
in which these statues were taken. More probably they were working from king lists and
counted to a figure given there, e.g. the beginning of Marduk-nadin-ahe's reign. Four hun-
dred and eighteen years before Sennacherib's capture of Babylon gives 1107 B.C. According
to the king lists Tiglathpileser, the successor of Ashur-re'sh-ishf, began to reign in 1112, and
the date thus falls, as it should, within his reign. As we have shown above, Nebuchadnezzar I
probably died around 1113. His successor, Enlil-nadin-apli, reigned 4-He years (BE I 1,
Pis. 30-31, No. 83), i.e., to 1109 or later. Thus 1107 can very well represent the beginning
of the reign of his successor, Marduk-nadin-ahe.
14
King, Chronicles concerning Early Babylonian Kings II 18-21.
15
Of importance for the correlation of the reigns of Samsu-iluna and Iluma-ilum are also
tablets from Nippur dated in Iluma-iium's 2d year (Poebel, BE VI2, No. 68, and in ZA XX
[1907] 241; Chiera, PBS VIII 1, No. 89). The persons mentioned on these tablets recur on
other documents ranging in time from the 33d year of Hammurabi to the 18th year of
Samsu-iluna (see Poebel in ZA XX 241 and Chiera, op. cit. p. 66). The year in which Iluma-
ilum—as shown by these tablets—held Nippur must therefore be sought within this period.
Now the dates on the Nippur tablets published in BE VI 2 and PBS VIII 1-2 show that
Nippur must have been in Samsu-iluna's possession from his 1st to his 8th, from his 10th
to his 15th, and from his 17th to his 23d year; for tablets dated to these years have been
found in Nippur. As years for which tablets with his dates are missing and in which the city
can therefore have been out of his possession there are thus only his 9th and his 16th years.
In his 16th year, however, Samsu-iluna was occupied with restoring fortresses in Emutbal
(formula for 17th year) south of Nippur, so there can be no question that he then held that
city, and the fact that no Nippur tablet with his formula for that year has been found must
thus be accidental. This leaves only the 9th year as a year in which Nippur can have be-
longed to another ruler, and we must accordingly date the Iluma-ilum tablets to that year.
Since these tablets belong to the 2d year of Iluma-ilum, his rebellion would thus have begun
in the 8th year of Samsu-iluna.
This result is strongly supported by the historical evidence contained in Samsu-iluna's
date formulas. In Samsu-iluna's 8th year occurred the Cassite invasion (date formula for
9th year), which furnished an opportunity for a rebellion in the south: Iluma-ilum's 1st
oi.uchicago.edu
ASTRONOMICAL DATING OF THE HAMMURABI DYNASTY

AND OF THE THIRD DYNASTY OF UR
A totally different approach is represented by efforts to fix by astronomical
computation the absolute dates of astronomical observations dated according
to the Old Babylonian calendar. Foremost among these stand the Venus tab-
lets of Amml-§aduqa, the importance of which for ancient Babylonian chronol-
ogy was recognized by Kugler in 1912. The history of the problem which
these tablets present is too well known to call for much comment. We need
only state that repeated tests and improvements of Kugler's results coupled
with new discoveries make it clear that the observations in question furnish a
series of possible datings only. In choosing among these, the best criterion is
dated Babylonian contracts dealing with delivery of dates by date-growers.
T&e days of delivery stipulated in these contracts will fall differently according
to which of the above-mentioned possibilities is chosen; and, since we must
assume that the deliveries would follow soon after the harvest, that possibility
which will make them fall most nearly in the time just after date harvest,
ending in the middle of October, has obviously the best chance of being correct.
As shown by Fotheringham, 16 the possibility which gives the best results is the
one which places the reign of Hammurabi from 2067 to 2025, and that dating
has now been accepted as the most satisfactory by the majority of the scholars
who have worked on the problem.17 It will be noted how well this astronomical
date agrees with the one at which we arrived after a critical sifting of the chron-
ological sources, using for reckoning only those which could be considered the
most reliable.
The Venus tablets of Amml-§aduqa are not the only ancient observations
year. In the next year, Samsu-iluna's 9th, this rebellion must have spread to Nippur, for
in the rebel confederation which met Samsu-iluna that year was Isin, just south of Nippur
(date formula for 10th year). This agrees therefore with the occurrence of Iluma-ilum tab-
lets in Nippur that year. In the encounter just mentioned Samsu-iluna was victorious, and
in his next (10th) year he was beleaguering Uruk and Ur, destroying their walls (formula for
11th year)—a situation which presupposes that Nippur farther north was already securely
in his hands. This fits with the fact that Nippur, as shown by tablets, now again dated by
Samsu-iluna formulas. On the correlation of Iluma-ilum and Samsu-iluna see also Schnabel,
Die babylonische Chronologie in Berossos' Babyloniaka (MVAG XIII 5 [Berlin, 1908]) pp.
19 f. Schnabel wrote before Nippur tablets dated to the 10th year of Samsu-iluna had
been published and therefore considered both the 9th and 10th years Iluma-ilum years, so
that his synchronism comes out one year too low.
"Langdon, Fotheringham, and Schoch, op. cit. pp. 69 ff. Neugebauer's objections to
this criterion (OLZ XXXII 918 f.) seem too-theoretical and general in character to carry
weight.
17
Besides Fotheringham, also Schoch, Langdon, and Schnabel (ZA n.F. II 109-22) ac-
cept this date.
oi.uchicago.edu
which afford a possibility of establishing an absolute date in older Mesopota-

mian history by means of astronomical reckoning. On the basis of observations
of a lunar eclipse foreboding the fall of I(b)bi-Sin and the end of the 3d dynasty
of Ur, Schoch has calculated that this omen occurred on the night of February
17/18 (Gregorian), 2283 B.C.18 Now it is a highly important fact indeed that
the two mutually independent astronomical dates, Fotheringham's for the 8th
year of Amml-§aduqa, which places Hammurabi at 2067-2025 B.C., and
Sehoch's for the omen foreboding the fall of Ur III, 2283 B.C., give just the
span of time between these two events which from our chronological lists can
be seen to have separated them. Langdon has commented briefly on this fact;19
but the revised text of the additions to the Sumerian King List established
above (pp. 122-27) and a reconsideration of the correlation of Ur III, Isin,
Larsa, and Babylon I makes it possible to state the argument wTith greater pre-
cision than he has done.
The correlation of the dynasty of Larsa with the 1st dynasty of Babylon has
been safely established by Thureau-Dangin.20 According to the date formula
for the 31st year of Hammurabi, which refers to his capture of Rim-Sin of
Larsa, that event can be placed in the 30th year of Hammurabi, that is, in
2038 B.C. according to Fotheringham's chronology. A list from Larsa giving
14 years for Hammurabi again shows that Larsa must have come into his pos-
session in his 30th year, for Hammurabi reigned 43 years in all.21 As for the
18
Schoch, Die Ur-Finsternis (eine Hypothese!); cf. Langdon in Langdon, Fotheringham,
and Schoch, op. cit. p. 82; Neugebauer in Astronomische Chronohgie I (1929) 96. Schoch
added "eine Hypothese" because he found difficulties in correlating this date with the one
found by Fotheringham for the Hammurabi dynasty. These difficulties disappear, however,
on closer examination of the chronological lists at our disposal and of the correlation of
the dynasties involved.
19
Langdon, Fotheringham, and Schoch, op. cit. p. 82.
20
La chronohgie des dynasties de Sumer et d'Accad, pp. 40-42.
21
Additional evidence is offered by the fact that the date formula for the 30th year of
Hammurabi actually seems to be the earliest date formula of that ruler found in Larsa and
also by a date list found in Larsa (YOS I, No. 33) which begins with this year (see Thureau-
Dangin, op. cit p. 42). We should also mention a tablet with double dating published by
Langdon (RA XXVII [1930] 23-25). This tablet gives the year in which it was written as
both the 8th and the 10th year after Isin was captured. On the correlation assumed, the
8th year of Rtm-Sln's Isin era corresponds to the 10th year of an Isin era counted from the
Babylonian conquest of the city under Stn-muballit. That our tablet actually dated from
this year and that its figures 8 and 10 refer to Larsan and Babylonian Isin eras respectively
is especially probable since the year in question is just that in which Hammurabi brought
Isin back under Babylonian rule. A revival of the Babylonian Isin era in that year is there-
fore very understandable. (Langdon's solution, which assumes that the figure 8 refers to
the 8th year of Hammurabi, seems to us less probable, especially as we have no other evi-
dence of dating by regnal years from this period.)
oi.uchicago.edu
reign of Rim-Sin, a slight discrepancy exists between the tablet just mentioned,
which gives him 61 years, and a large prism with Larsan date formulas, which
sums up his date formulas as only 60. This discrepancy is naturally explained
on the assumption that the scribe who wrote the tablet counted Rim-Sin's
formula for the year in which Hammurabi took the city, 2038 B.C., which
would have both a Rim-Sin and a Hammurabi formula, whereas the scribe
who wrote the prism did not.22 Rim-Sin's reign of 61 years thus comprises the
years from 2098 to 2038 B.C. inclusive; and, since the dynasty lasted 263 years,
its first year, the one in which Naplanum made himself independent in Larsa,
becomes 2300 B.C.
To correlate the dynasty of Isin with those of Larsa and Babylon we must
determine when in the date formulas of these dynasties the fall of the Isin dy-
nasty is recorded. In the period which can come into consideration Isin is re-
ferred to four times, and the earliest reference recording the capture of the city
by a foreign power presumably gives us the date when the independent dynasty
of Isin came to an end. The four references are:
Rim-Sin year 25 ( = Sin-muballit year 14)
Sin-muballit year 17 ( = Rim-Sin year 28)
Rim-Sin year 30 (=Sin-muballit year 19)
Hammurabi year 7 ( = Rim-Sin year 38).
Now the first of these references, that from the 25th year of Rim-Sin, can
hardly, as is generally assumed, refer to a capture of Isin. The formula23 reads:
[mu & - m] a h a n d e n - l i l d e n - k i - g a - t a u r u k i d a m - q f - T - l f -
§ u d - d a m z a g § u - d i b - b i i - s i - i n k i - k a s i p a zi d[r i - i] m -
d
s 1 n(EN-zu) i n - d i b - b a [ § i r - § ] i r - r a L#[+KAR- a] 1 a r s aki - m a - § e
b f - i n - t u - r e ruV-ul-a-ta u - m a - a - n i m u - u n - g u b - b a
"Year when by the exalted power of An, Enlil, and Enki the city (called) Damqi-
ilishu, (namely) crowds from the conquered border district of Isin,24 which the true
shepherd Rim-Sin had seized, he (i.e., Rim-Sin) brought into Larsa in chains and
captive and established his victory forever."25
22
Thus also Ungnad in ZDMG LXXIV (1920) 424 and in RLA II155.
23
Thureau-Dangin, op. tit. p. 55, AO 7025 obv. iii 47-51.
24
I.e., a d a m z a g § u - d i b b - a I s i n - ( a ) k - a ( k ) . We assume that the neigh-
boring ifa of d i b b and I s i n changed the - a of d i b b - a into - e .
26
Thureau-Dangin translates: "[annee ou, avec la force su]blime d'Anu, Enlil et Enki, le
pasteur legitime [R!]m-Sin prit la 'ville de Damiq-iliSu,' la foule (des habitants)
d'Isin, transfera a Larsa . . . . [ . . . . ] , e*tablit pour toujours sa puissance victorieuse.,,
Ungnad in RLA II 163, No. 227, translates: "Mit der hohen Kraft Anus, Enlils und Enkis
eroberte der wahre Hirt Rim-Sin die Stadt des Damqi-ilisu und die vollig(?) eingeschlos-
sene(?) Einwohnerschaft von Isin, brachte nach Larsa und errichtete fur alle
Zeit seinen Triumph." A serious objection to these translations—apart from the fact that
oi.uchicago.edu
Here, as will be seen, only a minor victory is reported, a capture of the popula-
tion of a city in the border district of Isin, not the capture of the capital itself.26
As the first actual capture of Isin itself we must therefore consider only the
second of the references listed, that of the 17th year of Sin-muballit, which
reads:
mu l - s i - i n - n a k i i n - d f b - b a
"Year when he (i.e., Sin-muballit) took Isin."
Since in general a date formula may be assumed to refer to an event of the
preceding year, this capture of Isin should have taken place in Sln-muballit's
16th year, that is, in 2072 B.C. according to Fotheringham's chronology, and
here we may therefore place the end of the independent dynasty of Isin.
The duration of the dynasty of Isin was 226 years, as we have seen above
(pp. 124-27 and notes). Since 2072 B.C. was the last year of the dynasty, we get
2297 B.C. as its first year, the year in which Ishbt-Irra set himself up as inde-
pendent ruler in Isin according to Fotheringham's dating.
Now according to Sehoch's computation, the eclipse which heralded the fall
of I(b)bi-Sin took place in the night between the 17th and 18th of February,
2283 B.C., and, since the Babylonian New Year fell toward the end of March,
I(b)bf-Sfn,s last year was thus 2284/83 B.C. Inasmuch as I(b)bt-Sin's reign
lasted 25 years, the rebellions of Naplanum and Ishbt-Irra should thus, if both
Fotheringham's and Schoch's datings are correct, have taken place in his 9th
and 12th years respectively; that is, his kingdom began to break up in his
9th year.
This, however, is exactly the year which other evidence indicates as the
fateful year of his rule. From the well known letter of I(b)bi-Sfn27 we know
that during the later part of his reign he ruled only the city-state of Ur and was
contemporaneous with Ishbt-Irra of Isin. The event which caused his authority
over all Babylonia to crumble was apparently connected with a campaign
they do not account satisfactorily for z a g § u - d i b - b i —is the casual way in which
Isin would be mentioned: "the city of Damiq-ilishu." That this is not a natural way of
referring to that important and time-honored capital shows up very distinctly by compari-
son with the formula in which Rfm-Sin actually does relate its capture (30th year): "Year
when with the exalted weapons of An, Enlil, and Enki the true shepherd Rtm-Sln seized
Isin, the city of kingship, and its population, as many as there were, granted pardon to its
widespread people, and had its royal name appear forever" (i.e., allowed it to retain its
status as royal city). To refer to such a city merely as "the city of Damiq-ilishu" would
minimize the importance of the victory.
28
That the city in question is not Isin is also held by George Tyler Molyneux in JSOR
IV (1920) 87-90.
27
PBS XIII, No. 3; Barton, Miscellaneous Babylonian Inscriptions, No. 9.
oi.uchicago.edu
200 THE SUMERIAN.KING LIST
against Anshan, for in the text BE XXXI, No. 3, we hear first about the divine
decision to bring evil times on the land, then, after a lacuna, about an expedi-
tion of I(b)bi-Sin to Anshan, and immediately afterward about the calamity
of Sumer. This evidence is supported by a date list dealing with part of
IO^bi-Sin's reign, for there the date formulas after the first, which tells of the
campaign against Anshan,28 suddenly have no more political references but
deal only with religious events centered around the cult in Ur itself. It is there-
fore obvious that I(b)bi-Sin suffered a severe setback in the year when the
campaign to Anshan took place. Now we can form a fair idea of where in his
reign this event should be placed, for it is obvious that those of his date formu-
las which have been found on tablets from Nippur or other northern sites in
Babylonia must belong to the early part of his reign when he ruled the whole
country, and the same is true of formulas which by their content show that
they are the formulas of an important king, not just a petty king ruling merely
the territory around Ur. On the other hand, formulas which deal only with re-
ligious events, centered around the cult in Ur, and have been found only in
Ur but not in other parts of the country are likely to belong to the later part
of his reign.
Now we find that there are only ten formulas which from the place where
they were found or from internal evidence show that they belong to I^bi-Sin's
early years, namely:29
(1) m u d i - b i - d s i n ( E N - z u ) l u g a l - a m
"Year: 'I(b)bi-S!n became king.' "
(2) rau e n d i n a n n a u n u k i - g a m & § - e i - p a
"Year: 'The enu of Inanna(k) of Uruk was envisaged on the (entrails of the
omen-Jkid.'"
(3) m u d i - b f - d s i n(EN-zu) lugal u r fki - m a - k e4 s i - m u - r u - u m ki
mu-l)ul
"Year: *I(b)bi-Sfn, king of Ur, sacked Simurum.' "
( 4 ) m u e n - a m - g a l - a n - n a e n d i n a n n a b a - fc ti n
"Year: 'En-am-gal-Anna(k), the enu of Inanna(k), was invested/ "
28
UET I, No. 292; cf. No. 290.
29
The known formulas of I(b)bt~Stn's reign are listed by Schneider in An. Or. XIII
(1936) 36-39. The first six of the formulas here quoted have been found in Nippur. The
remaining four prove by their content that when they were written I(b)bt~Stn was a king
of considerable power and influence. In our list parentheses are used with the year num-
bers of formulas the chronological position of which is fixed by date lists or similar evidence.
Brackets are used for those the order of which is hypothetical. The place of No. [10] as the
last of the early formulas is assured by the change to formulas referring only to local
events in UET I, No. 292.
oi.uchicago.edu
(5) m u t u - k i - i n - f r a t t a - r a i - i g - r i " " ^ DUMU-SAL l u g a l e n s i z a -

a b - § a -1 iki - k e4 b a - a n - t u k
"Year: The e n s i ( k ) of Zabshali married Tukin-hatti-niigrisha, the
daughter of the king.' "
[6]mu b a d - g a l n i p r u k i u r i k i - m a b a - d u - a
"Year when the great walls of Nippur and Ur were built."
[7] m u di - b f - ds i n(EN-zu) l u g a l u r f k i - m a - k a de n -111 -1 e [me-
ldm-a-ni] k u r - k u r - r a bf-in-§udun
"Year: 'Enlil put the terrible splendor of I(b)bi-Sin, king of Ur, as a yoke on
all lands.' "
[8] mu di - b f - ds i n(EN-zu) l u g a l u r i k i - m a - r a MAR-TU ID-IM-QAL
u l - t a u r u k i n u - z u gii i m - m a - n a - a n - g d - a r
"Year: To I(b)bi-Sin, king of Ur, did Amurru, a . . . . which from of old knew
not cities, make obeisance.' "
[9] mu di - b f - ds i n(EN-zu) l u g a l u r f k i - m a - k e 4 § u § a n k i a - d a m -
d u n k i m a - d a a - w a - a n k i - k a u 4 - g i m s i d b f - i n - g i4 u4 1 - a
m u - u n - g a m u e n - b i LU-HKAR1 - a m i - n i - i n - d f b - b a - a
"Year when I(b)bt-Sln, king of Ur, roared like a storm in Susa, Adamdun, and
Awan, subdued them in one day, and took their lords captive."29*
[10] m u di - b i - ds 1 n(EN-zu) l u g a l u r fki - m a - k e4 b u - t i b - n u - r i k i
s a k ( ! ) - k u l (!) m a - d a a n - § a - a n k i - § e (var. e l a m k i - [§ e]) . . . .
dugu-d-bi b a - § i - g e n . . - b i s a - b i - i n - g a r
"Year: 'I(b)bi-Sin, king of Ur, went in . . . . fashion to Huh(u)nuri, the key
(lit., "lock")29b of the land of Anshan (var., Elam) and '"
The last, that which records the campaign against Anshan, would thus be the
formula for I(b)bt-Sin's 10th year, and the event which it records should have
taken place in his 9th year. This is the very year in which, according to
Schoch's and Fotheringham's dates, Naplanum made himself independent in
Larsa.
There is thus full agreement between (1) the absolute date for the 1st dy-
nasty of Babylon as derived from the best of our chronological texts, (2)
Fotheringham's astronomical date for that same dynasty, and (3) Schoch's
astronomical date for the fall of the 3d dynasty of Ur. This fact speaks strong-
ly for their correctness, and we have accepted them with little hesitation.290
29a
Cf. Poebel's translation of this formula in AOF IX (1933/34) 248.
29b
In the reading and interpretation of this word we are following Poebel, quoted in
George G. Cameron, The History of Early Iran (Chicago, 1936) p. 58, n. 43.
290
The author has chosen this phrase deliberately and as distinct from "without hesi-
tation." The date here accepted is in our considered opinion head and shoulders above
other possibilities in probability. The convergence on it of the three independent lines of
oi.uchicago.edu
202 T H E SUMEBIAN K I N G L I S T
CONNECTION OF DATABLE DYNASTIES WITH THOSE

IN THE KING LIST PROPER
THE REIGN OP UTU-HE§AL
As shown by our detailed discussion of the text of the additions to the King
List (pp. 122-25 and notes), the duration of the 3d dynasty of Ur can be set at
109 years. Its first year would thus be 2392 B.C. The first king of the dynasty,
TJr-Nammu(k), held office as governor (shakkanakku) of Ur30 under Utu-hegal,
the last ruler listed in the King List proper, so it is clear that he must have
followed fairly closely upon the latter. An exact correlation of the two rulers
meets, however, with certain difficulties, of which the most important is that
presented by the reign of Utu-hegal.
As we have seen, the King List can be dated to the reign of Utu-hegal
(p. 141). If it was written under him, it seems highly probable, however, that
the figure which it gives for his reign, 7 years, 6 months, and 15 days, represents
not his full reign but only his reign up to date, that is, to the day on which the
scribe finished his work. The very exactness of the figure supports such an as-
sumption; for, since our scribe usually gives the reigns in full years, it is likely
that he would have rounded off Utu-hegaPs reign also if he had had it com-
plete. The dating down to month and day in this last figure of the list is, on
the other hand, natural if he here reckons to the day on which he finished his
work. We must therefore allow for the possibility that Utu-hegal may have
reigned longer than the time given in the King List.31
investigation mentioned places it almost as high in probability as that conception allows.

But we do not think it exceeds those borders; it still hovers slightly below the line separat-
ing highest probability from proven fact.
so JJET I, No. 30, a building inscription concerning E-kishirgal in Ur. The restoration
of u r - [ . . . . ] in i 10 as u r - [dn a m m u] can be considered certain since the writ-
ing points to his time and since we know from the Ur-Nammu(k) hymn in TRS I, No.
12:109: 6 - k i § - § i r 6 - g d l b u r - s a g - g a ( ! ) - g i m k i - g a l - l a b i - g u b , "E-
kishirgal like a mountain I set upon . . . . , " that Ur-Nammu(k) really did build E-kishirgal.
Landsberger's skepticism (OLZ XXXIV [1931] 118, n. 1) is therefore unfounded.
81
Another difficulty may at least be mentioned briefly. Analogy from Lagash indicates
that the Sumerian cities dated according to local date formulas down into Ur-Nammu(k)'s
reign. The local date lists from Ur may therefore have contained date formulas of Ur-Nam-
mu(k) back into the time when he ruled that city as governor of Utu-hegal. If the additions
to the King List were made in Ur on the basis of such a list, the 18 years assigned to Ur-
Nammu(k) could thus include all or part of the time in which he was merely governor. Since
we have no special reason to believe that the additions were made in Ur and since—as
Landsberger has pointed out (be, tit.)—the extensive building activities of Ur-Nammu(k)
outside Ur suggest a long reign, it is more probable that his 18 years actually stand for the
time in which he ruled all Babylonia. The Chronicle Weidner, which has been brought
into the discussion (Christian and Weidner in AOFV [1928-29] 140), states that Utu-hegal
oi.uchicago.edu
The difficulty here met with cannot be solved on the evidence of the King List
and its additions. It is therefore fortunate that we have other material which
indicates in which direction the solution lies. Schoch's article in which he cal-
culated the eclipse of the moon foreboding the fall of Ur deals also with obser-
vations concerning an eclipse which took place on the 14th of Hu§ u (- n u m u n)
and which foreboded: "A decision will be given the king of the Gutians; down-
fall of the Gutians by weapons will take place. The land will be lying desti-
tute."32 That this omen has reference to Utu-hegaPs famous victory over the
Gutians, which freed a country bare and destitute after years of Gutian mis-
rule, seems obvious. That, more important, the late text in which the observa-
tions are contained actually rests on a sound historical basis is indicated by the
inscription of Utu-hegal himself.33 According to Utu-hegars account of the
campaign, he camped the night before the battle near Muru. As he had prayed
to other city gods on his route, so he prayed to Ishkur also. But the account
does not, as usual, stop there; a passage, unfortunately broken, goes on: "In
the midst of the night " Something of importance must thus have oc-
curred in the night just before the decisive battle. What it was can hardly be
doubtful; it must have been the lunar eclipse foreboding victory over the king
of Gutium. The moon-god Sin, the special god of the Gutians,34 became dark-
ened in token of their imminent defeat. Now Schoch has shown that in the
period which can come into consideration there actually was a very noticeable
eclipse of the moon which took place, as stated in the text, on the 14th of
itu
§ u - n u m u n and, also as stated in the text, lasted from the first watch
to the middle watch. This was the eclipse of the 20th of July, 2403 B.C.,
committed some evil act and that his body was carried away by the river. It then goes on
(ZA n.F. VIII 49,1. 30): "[To] Shulgi, son of divine Ur-Nammu(k), he (i.e., Marduk) gave
the kingship of all countries." The reason why the author of the chronicle has Marduk give
the "kingship of all countries'' only to Shulgi, not to his father and predecessor, Ur-Nam-
mu(k), although he knows him, is probably that Ur-Nammu(k)'s kingdom never came up
to the standards implied in the term sarrut kissat malate, "kingship of all countries." Note
that, in contrast to Utu-hegal and Shulgi, Ur-Nammu(k) does not use the title "king of
the four regions." (An exception is perhaps Barton, RISA, p. 274, No. 13. We do not feel
certain, however, that Barton's rendering is correct.) The chronicle apparently implies that
Marduk held back the "kingship of all countries" after the death of Utu-hegal until the
reign of Shulgi. This agrees with historical conditions. The statement cannot therefore be
used—as it is by Christian and Weidner in AOFV 140 f.—to indicate that Shulgi followed
almost directly upon Utu-hegal. Nor does this statement—as maintained by Landsberger
(loc. cit.)—disprove the value of the chronicle as a chronological document.
32
C. Virolleaud, L'astrologie chaldemne (Paris, 1905-12) Sin XXXIII 18-36.
™RA IX (1912) 112 f.
84
Note the inscription of Laerab {ZA IV [1889] 406; SAK, p. 170), 11. 18-21: ila gu-ti-4m
Hnanna ii d8in, "the gods of Gutium, Inanna(k) and Stn."
oi.uchicago.edu
lasting from 9h9m to 12h51m in the night.38 There is therefore little reason to
doubt that our evidence is reliable, and we can date Utu-hegaFs victory over
the Gutians at Muru, which took place the day after this portent, to July 21,
2403 B.C.
This dating corroborates our suspicion that the figure for Utu-hegal may
not represent his complete reign. As we have seen above, the 3d dynasty of
Ur began in 2392 B.C., which would represent the first year of Ur-Nammu(k).
Utu-hegaFs reign must thus have lasted more than 7 years, 6 months, and
15 days to reach from his victory over the Gutians in 2403 down to that date.
The dating of the victory over the Gutians still leaves a small amount of
uncertainty concerning Utu-hegal, inasmuch as we do not know exactly where
within the 7 years etc. given by the King List that event should be placed.
That it represents the beginning of Utu-hegars reign is not very likely, for
from his inscription it is clear that when the summons from Enlil came he had
so complete control of Uruk that he could gather the army and lead it out to a
campaign without even stating what the object was to be. We must therefore
imagine that he was already well established as ruler. Under these circum-
stances, and until further evidence turns up, it therefore seems best to place
this victory where we will get the smallest margin of error, namely in his 4th
year, the middle of the 7 given in the King List. This leaves a margin of error
of only 3 years.
GUTIUM AND AGADE
With the period from Utu-hegal back to Agade we are within the limits of
the King List proper, and it is therefore natural to ask whether the peculiarities
of its author and the methods which he used in dealing with his material may
not have influenced this, the last part of his work, as they did the earlier parts
studied above. Our first concern must therefore be to get as clear a picture as
possible of the sources at his disposal.
The evidence which we have concerning conditions during the Gutian period
tends to show that the larger cities of southerh Babylonia enjoyed a certain
measure of independence. They were ruled by their own e n s i (k) 's and
—to us the most important point—they apparently dated from local calen-
85
A smaller eclipse on July 20, 2384, can hardly come into consideration, since placing
the defeat of Gutium there would unduly shorten the reign of Ur-Nammu(k) as king of all
Babylonia. It would also presuppose that Ur-Nammu(k) began to rule in Ur before Utu-
hegal in Uruk. This is, however, very unlikely; for the fact that Ur-Nammu(k)'s personal
deity is Ninsun (UET I, No. 47; TRS I, No. 12:112-13), a goddess closely connected with
Uruk, suggests that he hails from that city. I t is therefore natural to assume that he was
one of Utu-hegaFs men who was made governor of Ur after Utu-hegaFs victory over
Gutium had brought Sumer into his hands.
oi.uchicago.edu

dars featuring local events in the year dates.36 We may therefore assume that
Uruk also had its local calendar. Since the Agade kings, who ruled Babylonia
before Gutium, had succeeded in establishing a central calendar used all over
the country, we may assume furthermore that the local Uruk date lists con-
tinued that calendar (i.e., they had Agade date formulas) until the fall of that
city, then continued with local formulas featuring local events and rulers, of
whom Utu-hegal would be the last.36a
To such lists as these, the date lists of his day, the author of the King List
naturally had access, and from what we have learned about him and about the
methods which he employed in dealing with his sources it must be considered
highly unlikely that he discarded any of the material which they contained.
He worked, as we have seen, almost exclusively by a method of interpolation.
Therefore when wefindthat the King List gives the Gutian dynasty wedged in
between rulers of Uruk, Ur-Utu(k) and his predecessors above, Utu-hegal be-
low, it begins to look suspiciously as though that dynasty had been interpo-
lated in a succession of Uruk rulers taken from the local Uruk date lists. If this
suspicion is justified, the Gutian dynasty instead of separating the 4th and 5th
dynasties of Uruk actually ran parallel with them. Placing—as suggested
above—the defeat of Tirigan in the 4th year of Utu-hegal, which limits the
margin of error to some 3 years, and assuming that Ur-nigin(ak) of Uruk in-
stituted his own calendar with local date formulas when the last Agade king,
Shu-Durul, died, we get the revised arrangement shown on the following page.
To decide whether this or the successive arrangement of the dynasties pre-
sented by the King List comes nearer to the truth we may compare first the
evidence of Sumerian historical tradition. We possess a number of texts which
state that the reign of Naram-Sin ended in disaster. The Sumerian text BE
XXXI, No. 1, relates a tale of unsuccessful omens and divine decision to over-
throw his rule. The same tradition is found in the long Nar&m-Sin text edited
by Guterboek,37 where the gods likewise determine to overthrow his reign
and give his kingship to someone else. In its most precise form the tradition
appears in Chronicle Weidner,38 according to which Naram-Sin incurred divine
wrath so that his rule was overthrown and his kingship given to Gutium. That
this widespread tradition rests on a solid historical basis is shown, further, by
36
See the local year dates of this period from Lagash in RLA II 133 f.
36a
Cf. the parallels from, e. g., Eshnunna(k), where the calendar similarly changed from
"centralized dates" to local dates when the city became independent after the 3d dynasty
of Ur.
" ZA n.F. VIII 28,11. 31-33.
38
Ibid. p. 49, 11. 20-22; cf. Christian and Weidner in AOF V (1928-29) 140.
oi.uchicago.edu
AGADE URUK 6UTIUM

Naram-STn
37
-< .->•-• [Erridupizir]

Shar-kali-sharrl •pmta 5 3
25 -i Inkishush 7
Sarlagab 6
figigi, Nanum,) Shulme' 6

/llmi.Elulu^Ll Elulumesh 7
Dudu ^ ^
21 Inimabakesh 5
Igeshaush 6
larlagab 15
Shu-Durul rlbate 3
15 •larlatngab] 3
-i-i/Kurum r
HfrHabi!-kTn 3
^<- ->--, "p^Laerabum 2
Ur-nigin(ak)
^-Xlrarum 2
j Ur-gigir(ak) IJvlbranum I
-p«-HabIum 2
-fkudda 6 Puzur-STn 7
•j Puzur-ili 5 larlaganda 7
-fur-Utu(k) 5 Si5u(m)
Utu-hegal<- • > - .
Tirigan 40 days
oi.uchicago.edu
a contemporary document, a letter written in the time just after Nar&m-Sin,

which shows that the Gutian hordes were then already the actual masters of
the country. 39
Now it will be seen that the revised arrangement given above actually has
the beginning of the Gutian dynasty at the end of Nar&m-Sin's reign, in full
agreement with the Sumerian tradition that Nar&m-Stn was overthrown and
that the Gutians took over his kingship. It is therefore clearly preferable to the
successive arrangement of the dynasties now presented by the King List; for
there the Gutians are separated from Nar&m-Sin by twelve later rulers, cover-
ing nearly a century. This result is not surprising, for the successive arrange-
ment rests solely on the authority of the order of dynasties given in the King
List, and that feature of the list is, as we have seen above, quite unreliable.
The evidence from Sumerian tradition is not the only argument which shows
that the parallel arrangement of the dynasties is correct. Among the Gutian
rulers is one Elulumesh, whose name is evidently Akkadian Elulum slightly
"Gutianized" by the Gutian case(?) ending -es.40 This Gutian ruler Elulum
is obviously the same man whom we find participating in the scramble for
power after the death of Shar-kali-sharri; his name appears there in Sumerian
form without mimation as Elulu. The correctness of the revised arrangement
comes out plainly when we find that there Elulum's reign as Gutian ruler and
his participation in the struggle after Shar-kali-sharri belong to the same time.
By the successive arrangement they would be pulled more than eighty years
apart.
Lastly we may mention a date formula of Shar-kali-sharri, according to
which this ruler was a contemporary of the Gutian king Sarlag (sar-la-ag),
whom he defeated.41 Now the King List has among the Gutian rulers one
whose name is written s a r - KIL - 1 a - g a b (var. z ^ - a r - l a - g a - b a ) .
Since KIL may itself be read 1 a g a b , it seems reasonable to assume that
the repetitious 1 a - g a b following it represents a gloss which has got down
into the line, so that we should read s a r - 1 a g a b l a _ g a b , an assumption
supported by the variant also.42 It is not difficult to recognize in the name Sar-
lag an abbreviated form of this Sarlagab; and the correctness of the revised
39
JRAS, 1932, p. 296, BM 121205.
40
That (e)s is a grammatical ending is suggested by the frequency with which it occurs
at the end of Gutian proper names. Cf. Inimabak-es, Inki§u-(e)s, IgeSau-(e)s, and Iaria-
ga-(e)s.
41
RTC, No. 118; SAK, p. 225 c.
42
See the discussion of the name above (p. 118, n. 291).
oi.uchicago.edu
arrangement is once more evidenced when we find that in it Sarlagab and

Shar-kali-sharrI appear as contemporaries.
The parallel arrangement, suggested by consideration of sources and of the
methods used by the author of the King List, is thus vouched for by Sumerian
historical tradition, which makes Gutian domination begin after Nar&m-Sin,
by the fact that it places the apparently identical Elulu and Elulumesh in
the same period, and by the fact that its correlation of Shar-kali-sharri with
Sarlagab corresponds with the synchronism Shar-kali-sharri-Sarlag. We have
therefore adopted it in preference to the successive arrangement, for which
there is no other authority than the highly suspect present order of dynasties
in the King List.
oi.uchicago.edu
TABLE E
CHRONOLOGY OF MESOPOTAMIAN RULERS DOWN TO THE FIRST DYNASTY OF BABYLON
AKSHAK KISH AGADE URUK UR LAGASH
Etana
-. [ca. 50] >
JO
3100
Balih >
CO
[ca 20]
H
O
En-me(n)-nunna
[ca. 20]
3050
Mebm-Kishi(k)
[at. 20]
Bar-sal-nunna
[ca.20]
5000 Samug
[ca. 50]
Mes-kiag-gasher
[ca. 30 J
Tizkar
[ca. 20]
2950 llku 3
[ca. 20] Errme(r)-kar
[ca.20]
llta-sadum
[ca. 20] Lugahbanda
to. 20] >
2900
Errme(n)-barage-si
-. [ca. 30] Dumu-zi(d)
[ca.20]
r
Gilgames
Aka
to-30] d
[ca. 20]
K:
>
Su....
[ca. 20] Ur-Nungal(ak) Gurshar
50 to. 30] i—i
2850 Mes-Anne-pada
ca. 40 o
Dadasig
-: to. 20]
Utui'kalammafk) Gunrdu
15 [ca. 30 ]
Magalgalla
-j [ca. 30]
Laba...jR(?)
A-Anne-pada
9
ca. 40
En-n un-dara-Anna (kj^
2800 8
Ur-Nanshe(k)
MES(?)-HE
[ca. 50]
Kalbum 36
[ca. 20]
to. 20] 36 A-Kurga! (ak)
[ca.20]
Melam-Anna(k)
- 6
2750 Lugal-ki-tun(?)
GA+suB-nun-na 36
[ca. 20] E~Anna(k)-tum
Unzi [ca.20]
30
Elulu
Enbi- Eshtar 25
[ca.20] En-Anna-turn I
to-20]
En-shakush-AnnaO^-j
J Undalulu 1 » ^^W • M^BBM* V M M « « H ^ • MBMK • to-30]
12 ~\ Lugalmu Balulu
-i [ca.20] 36 En-temena
[ca. 30]
2 7 0 0 * Ur-ur
J- ©
Puzur-Nirah
20 Ku(g)-Baba
[ca. 30] Liflal-hinishe-dudu -: ___ i__ §
O
En-Anna-tum E
lshu-il [ca.50]
24 Lugal-kinrshe-dudu
[ca. 16]
3
Puzur-Sln
25 Lugai-kisal-si -i Lugal-kisai-si
:
jo
to-30] [ca. 30 ]
2650- Shu-Sm
i_ 7 -: Lugal-anda
L 7
En-e(n)tar-zi(d) 5
Ur-Zababa(k) Uru-KA-gina
X)
[ca. 20]
' Lugal-zage-si •>**'- w
25
....gi
[ « . 35]
o
-*--v<- -**-i
" Simtrdar Sargon
30 -i 56
2600
J Ka-ku(g)
* [ca. 35 ]
7
i—i
Usrwatar
-j_ 7
Eshtar-muti en
li
H
Ishme-Shamash i—i
ri
Rimush < ->'-
X
Nannra >
2550
7
9
Q
m
Man-ishtushu
15 >
d
Naram-STn
Pi
37
Q XJ
w
* i—i
G O
GUTIUM
2500
Shar-kalrsharrl
25
• ^ - f [Erridupizir]
-Dmta s
-i Inkishush
I
"i
>
2
4
2
Sarlagab 6
H
W
u^ [ j g r g i , Nanum,]
L / l l m i . Elulu +L
' Dudu
3
Shulme*
Elulumesh
6
7 l—I
pa
O •—*
i
21 Inimabakesh a
Igeshaush
W
larlagab
2450
Shu-Durul I bate >
f5 larla[ngab] t-
Kurum
. Habil-kin
—*<- ->-, Ur-nigin(ak) i=>Laerabum -1—d
J^/rlrarum
Ur-gigir(ak) ^-V'branum pa
-—<-HabIum i—i
-i Kudda
Puzur-Stn
Puzur-ili larfaganda
o
Ur-Utu(k) Si3u(m) a
Utu-hegal^-
->'-. Tirigan 40 days ^ 1
2400
Ur-Nammu(k)
18
0)
o
o
Shulgi
48 a
2
2350 C rn
?o pd
Bur-5?n
9
o ni
ShQ-S?n o
9
a
ISIN LARSA BABYLON l(b)bi-Sin
25 V v
2300_ Naplanum
Ishbi- Jrra 21
33
Emizum
28
Shirilishu
10
l(d)din-Dagan
2250- 21 Samu^jm
35
XI
oi.uchicago.edu
Gi [games
[ca. 30] <
Aka
[ca. 20] Z
>
Su....
[ca. 20 ] Ur-Nuncjal(ak)
30
Gurshar
[ca. 30 j
Hi—«
2850
-i
Mes-Anne-pada
Cd,4Q o
Dadasig
[ca. 2 0 ]
0
U t u I kalafrifna(k) -i Gunidu
15 [ca. 3 0 ]
Magalgalla
[ca.30]
Laba....iR(?)
A-Anne-pada
9
c<* 40
2800 Errnun-dara-Anna(k)^
Ur-Nanshe(k)
MES(?)-HE
[ca. 30 ]
Kalbum 36
[ca. 20]
-»*
[CA 20] 36 A-Kurgai (ak)
[ca.20]
Melam-Anna(k)
-- 6
2750 Lugal-krtun(?)
GA+suB-nun-na 36
[ca. 2 0 ] E - Anna(k)~tum
Unzi [ca. 20 ]
30
Elulu
Enbi-Eshtar 25
[ca.20] ErrAnna-tum I
[ca.20]
En-shakush-Anna(I^-j
J Undalulu [ca.$Q]
\2 Lugalmu Balulu
[ca.20] 36 En-temena
[ca.30]
270CH Ur-ur
J- 6
Puzur-Nirah
-* 20
Ku(g)-Baba
[ca. 30] Ujgal- kinishe-dudu
-*j
[ca. 3 0 T "
En-Anna-tum IT
3
ishu-il
•i 24 Lugal-kinishe-dudu J [ca. 30]
[ca 16]
2
Puzur-STn
25 Lugal-kisal-si Lugal- kisal-si
[ca\30] [c<?. 30 ]
2650-i Shu-Sfn
i_ 7 Lugal-anda
L 7
En-e(n)tar-zi(d) 5
Ur-Zababa(k) Uru-ftA-gina
[ca. 20]
->*•- X)
' Lugaf-zage-sr I—I
25 [CA35]
o
>v-*
Srmu-dar
-*--v*
Sargon
30 56
2600
Ka-ku(g)
[c*35 ]
Ust-watar
7
Eshtar-muti en
li H
ishme-Shamash
II
RTmush < -*>- 2
Nannia 9 >
2550 7 Q rn
Man-ishtushu
15 >
d !>
r
Naram-Sm
PI
37
m
G O
OUTIUM
2500
Shar-kali-sharri
25
• •^-*[Erriduprzir]
-pruta1
J Inkishush
>
z 4
2
Sarlagab 6 13 H
Shulme' 6 W W
• rigigi.Nanum,]
W i
L n l m i , Elulu ^ l i J Elufumesh 7 i—i
" Dudu ^ ^ o
21 i Inimabakesh 5 a §
Igeshaush 6 "a
lariagab 15 w
2450 Id
ShO-Durul (bate 3
>
15
i-««- ->--,Ur-nigin(ak)
J larla[ngab]
Kurum
Habil-kin
"sff/'Laerabum
JryArarum
3
I
3
2
2
r
m
Ur-gigir(ak) g > I bran urn I
-Z*-Hablum 2 Xi
J Kudda i Puzur-STn 7 o
Puzur-iii larlaganda 7
a
-TUr-Utu(k)
Si a u(m)
Utu-hegal ^ - ->>--
2400 Tirigan 4 0 days «..
J Ur-Nammu(k)
18
Shulgi
48
T 03
171
O
2350 cm
jo ?d
»—i
Bur-STn
9
o tn
i—i
Shu-STn
o
9
a
ISIN LARSA BABYLON l(b)bt-Sm
25
2300_ Napianum A A
IshbHrra 21
33
Emizum
28
Shu-ilishu
10
l(d)din-Daqan
2250-i 2! Samu'um
35
Tl
ishme-Dagan
O
20
0)
D
Zaba a
Lipit-Eshtar 9
H
2 2 0 0 ^ Ur-Ninurta(k)
Gungunum
27
'2
:
28
> Pi
AbT-sare > >

Bur-STn
21
II
Sumu-EI Sumu-abum
r
29 14 •o
jo tD
Sumu-fa-Ef
J Lipit-Enlif 5 J 36 O
2150-
'" lrra-imi(t)ti
8
t—I
Enlil-bam
24
Nur-Adad
16 o
D
S?n-i(d)drnam
— ?5?n-inbam Zabum
-IZambia 3 ^ 5?n-iqfsham L4
jjter-pisha 4 J=vsn|j-Adad
j_Ur-Du(l)kuga(k) 4 j W a r a d - S ? n
STrrmagir 12 Apil-STn
2\0O> I' 18
RinrrSTn
Damiq-ilishu 61
23
STn-mubalfit
20
Hammurabi
43
2050
§2
O '
2000 2
oi.uchicago.edu
INDEX
Italic numbers refer to the most important pages, especially those in the text of the King
List (chap, iii) where the various rulers and dynasties appear.
A, 18-23, 50, 54-55, 67, 82-85, 102, 104-5, antediluvian rulers, 166-67; — section of
108-9, 115-18, 143 King List, 5, 11, 29, 55-68, 70-77, 136
A-Anne-pada, 92-94, 172-73, 181, 183, 184- Anu-banlni, 140
86, 190, and Tables I-II A-nu-zu, 172
Aba, 15-16, 18, 19, 22, 79 Apil-Sin, Table II
Abl-Eshuh, 195; time of -—, 102 Aqqi, 145
Abi-sare, 138 and Table II Arad-gin, 76
Abraham, 89 Arad-Nanna(k), 98
D
Adab, 102,149,151; dynasty of —, 100-103, Arbum, see Arwi um
156 arrangement of dynasties, see dynasties, ar-
Adad, 194 rangement of
Adamdun, 201 ArwiDum (Arbum, Arwi, Arwiu), 15-16, 17-
18, 19-20, 21-22, 23, 56, 80-81
additions to the King List, 21, 32, 56-57, Ashur temple, 192
61-77, 86, 122-27, 135-36 Ashurbanipal, 57
Aelian, 87, 89-91 Ashur-r&h-ishl I, 194-95
Agade, 13, 111, 143, 155, 167, 179, 182, 205;
— dynasty, 6, 9-11, 14, 23-28, 45, 47, 50- Ashur-uballit, 193-94
51, 52-53, 110-15, 156, 158, 160-61, 177- assimilation of vowels, see vowels, assimila-
79, 204-8, and Tables I-II; — period, 97, tion of
119, 144, 148-50, 162, 166; — variants, 3, Assyrian chronology, 191-94; — period, 11,
14, 23-28, 42-46 60, 90, 145, 191
Aka, 84-85, 163, 168, 186, 187-88, and astronomical dating, 196-97, 199, 201,203-4
Tables I-II Atab, 16, 18,19, 20, 21-23, 56, 78-81
A-kalam-du(g), 181 Atabba, 15-16, 18-20, 21-23, 56, 79-80;
Akkadian influence on Sumerian, 149; — — variants, 15-23, 28, 42, 44-46
loan words in Sumerian, 17-18, 171; — author of the King List, 18, 61-63, 136, 139,
143, 146-47, 152-56, 158-64, 204-5
passage in King List, 52-53, 112-13
Akshak, 106, 158-60, 167, 181; dynasty of Awan, 95, 158, 201; dynasty of —-, 94-95,
156-58
—, 3, 6, 8-10, 12, 29, 102, 104-5, 106-7, Azupiranu, 145
156-60, 169,176-77,179, 181, and Tables
I-II B, 18-23, 27-28, 42-48, 50, 55, 64, 67, 79,
A-Kurgal(ak), 180, 185, and Tables I-II 82-85, 102, 104-5, 108-9, 112, 115-18,
Al , 182 143, 174-75
Alalgar, 65, 70-71 Baba, 104-5
A-lulim(ak), 65, 68, 70-71 Babylon, 7-8, 60, 70, 146, 194; 1st dynasty
of —, 168, 195, 197-98, 201; period of 1st
amissible final consonants, 69, 87-88 dynasty of —, 5-7, 9-13, 90, 144, 175
Ammi-§aduqa, time of, 102; Venus tablets Babylonian chronological material, 191,
of —, 196-97 193-97
Amorite names, 124 Bad-tibira(k), 58-59; antediluvian dynasty
Amurru, 201 of —, 65, 70-73, 74
An, 58, 152, 198-99 Bahina, 78-79
Anam, 89 Balih (Walih), 80-81, 168, and Tables I-II
animal names, 17-18, 78-80, 152, 155 Balulu, 94-95, 172, 184, and Tables I-II
Anshan, 137, 200-201 209
Ansud, 102-8
oi.uchicago.edu
Bar-Ilitabba(k), 139 dialectal forms, see e m e - SAL forms

Bar-nam tarra, 150 dictation, errors due to, 14, 17, 76, 81, 91,
Bar-sal-nunna, 82-881 168, 188, and Tables 94, 173
I-II Dudu, 130
"Bavian date," 194 Dudu, king of Agade, 24-26, 114-15, 177,
Berossus, 60, 70-76, 87-88 206, and Table II
Bu-AN-?-um, 78-79 Dudu, s a n g u -priest of Ningirsu(k), 149-
Bur-Sin of Isin, 124-25, 130, 135, and Table 50
II Duggagib, see Zuqaqlp
Bur-Sin of Ur, 122-23, 130-31, 133, 135-36, Dumu-zi(d) (Tammuz) of Bad-tibira(k), 72-
174-75, and Table II 73
calendars, see date lists and dating systems Dumu-zi(d) (Tammuz) of Uruk, 88-89,
ease endings, Akkadian, 69; Gutian —, 207 142-43, 145, 170, and Tables I-II
Cassite invasion, 195 Dungi, see Shulgi
"causative" - n - , 131-35, 139 Dur-an-ki(k), 75
chronology, absolute, 189, 191, 194-99, 201- dynasties, arrangement of, 2, 13, 58-60, 66,
4; Assyrian —, see Assyrian chronology; 70-72, 74-75, 102, 104-5, 159-67, 204-8;
Babylonian —, see Babylonian chrono- division of -—, 161-62, 168-69, 172-73;
logical material parallel —, 2-3, 165, 205-8
city king lists, see date lists, local dynasties of one ruler each, formulas for
collation of WB, 29, 69 totals in, 30, 33, 61-63, 99, 102
collective form of nouns, 39; — of verbs, see E, e n s i (k) of Adab, 149
verbs, collective form of
E-Anna(k), 37, 84-85, 87, 91, 143
consonants, amissible, 69, 87-88; develop-
ment of —, 18, 71-72; see also e m e - SAL E-Anna(k)-tum (Lumma), 39, 99, 155, 180-
forms and nasalization 81; inscriptions of —, 137, 184; time of —,
conventions observed in transliteration, 16, 71, 137, 172, 184-85
69 Early Dynastic II-III, 189 and Table II
Egypt, relations of, with Mesopotamia, 189
Dadasig, 96-97, 168, and Tables I-II
Egyptian chronology, 189, 193
Dagan, 113
E-Ishkur(ak), 139
Damiq-ilishu, 5-7, 127, 130, 135, 198-99,
and Table II Ekallate, 194
Damqi-ilishu, 198-99 E-kishirgal, 202
date formulas, 122-23, 137-38, 144, 147-48, E-kur, 98-99
195-201; local —, 202, 204-5 Elam, 82, 84-85, 88, 95, 137, 141, 178, 201
date lists, 123, 144, 147-48, 150-51, 200; Elamite documents, 10, 90, 156; — lan-
local —, 151, 154, 156-59, 161, 165-66, guage, 40-41; — versions of King List, 1,
179-82, 197, 202, 204-5 10-11 25-26, 34, 40-41, 46, 51, 54r-55
date of original King List, 128-41 {see also Sui, »U2, SU3+4)
dates of extant manuscripts of King List, Elili(n), 184
5-7, 9-13, 23, 124 Elulu (Ilulu) of Agade, 24-26, 52, 114-15,
177, 206, 207-8, and Table II
Dati-Enlil, 113 Elulu of Ur, 94-95,172,184,190, and Tables
dating, criteria for: ideology, 138-40; lan- I-II
guage, 67-68, 129-40, 186; names, types
of, 187-89; orthography, 2, 7,124,131-35, Elulumesh, 12, 118-19, 120-21, 206, 207-8,
139, 172, 182, 184-86; sign forms, 5-7, 9- and Table II
13, 184-85, 202; stratigraphy, 181; style e m e - SAL forms, 18, 71, 74
of statuary, 186; tablet forms, 184-85 Emizum, Table II
dating systems, Assyrian, 191-92; Baby- Emutbal, 195
lonian —, 193; Sumerian —, 147-51
En-am-gal-Anna(k), 200
deified kings, 68, 98-99
En-Anna-tum I, 180 and Tables I-II
Deluge, 56, 58-60, 65-66, 76-77
En-Anna-tum II, 180 and Tables I-II
determinatives, use of, as dating evidence,
7, 68, 124, 182 Enbi-Eshtar, 96-97, 118, 168-69, 183, 190,
and Tables I-II
oi.uchicago.edu
INDEX 211
En-e(n)tar-zi(d), 114, 149-50, 180, and 137-40,170; for dynasty total, 14, 28, 30-
Tables I—II; inscription of —, 137; time 44, 46-49, 56, 61-63, 129; for introducing
of —, 187 new dynasties, 29, 64, 67-68, 107, 122,
135-36, 143, 152, 162, 168-69, 172-73;
Enki, 60, 198-99 for introducing single rulers, 14, 28-36,
Enkidu, 89 38, 42, 49, 54-55, 61-62, 92, 100, 131,
Enlil, 5&-60, 89, 93, 178, 198-99, 201, 204; 170; for king's father, 18, 21, 28-29
— temple in Nippur, 99, 157
G, 12, 36, 38, 42, 106, 118-19
Enlil-bant, 6-7, 23, 126-27, 130, 145, 175,
Ganhar, 98
and Table II
Gashananna(k), 62
Enlilla(k)-zi(d), 98-99
GA+suB-nun-na, 96-97, 168, 189, and
Enlil-nadin-apli, 193-94 Tables I-II
En-me(n)-barage-si. 56, 82-85, 142-43, 162, genitive case, see grammar
1681 188, and Tables I-II Gilgames, 18, 87, 88-91, 142-44, 146, 157,
En-men-dur-Anna(k), 74-7$ 163, 170, 182, 186, 188-89, 190, and
En-men-gal-Anna(k), 72-78 Tables I-II
En-men-lu-Anna(k), 70-73 Gimil- , . . . , see Shu- . . . .
En-me(n)-nunna, 73, 80-88, 168, 188-89, glosses, 82, 96, 99, 118, 169, 183, 207
and Tables I-II grammar: case endings, 69, 207.; genitive
En-me(r)-kar, 56, 77, 85, 86-87, 89, 142-43, case, 21-22, 30, 71, 80; infix - r a - , 62;
170, and Tables I-II locative case, 30; nouns, plurals of, 39;
preposition "after/ 1 66; relative sen-
En-nun-dara-Anna(k), 90-91, 170, 188-89, tences, 29, 105, 143; subject element - e ,
and Tables I-II 17, 29-36, 135; suffix - b i , 37, 40, 84;
Enridapizir, see Erridupizir verb, identifying, and verb, transitive,
En-shakush-Anna(k), 100-101, 170-71, 183, see verb; verbs, collective, passive, and
singular, see verbs
190, and Tables I-II; time of —-, 148 Greek transliterations of Sumero-Akkadian
En-sipa(d)-zi(d)-Anna(k), 60, 74-75 names, 70-77, 87-88, 90, 103
En-temena, 139, 149-50, 183, 190, and Gudea, 138, 140; inscriptions of —, 131,
Tables I-II; inscriptions of —, 137, 139, 137-38, 146; statue of —, 99
149-51, 181, 184-85; statue of —, 99 Gulkishar, 193, 195
epithets, 29-32 Gungunum, 137 and Table II
eponym lists, 192 Gunidu, Tables I-II
Era, see lira Gurshar, 180 and Tables I-II
Eridu(g), 58-61, 64, 89; antediluvian Gutian language, 119, 207; — period, 138
dynasty of —, 65-68, 70-71 Gutium, dynasty of, 6, 11-13, 37, 116-21,
Erishum, 192-93 138-40, 183, 208-8, and Table II
Erridupizir (var. Enridapizir), 117, 206, and Ga. .ur, 76-77
Table II
error, source of, see dictation Habil-ktn, 118-19, 206, and Table II
E-sag-ila, 146 Hablum, 120-21, 206, and Table II
Esarhaddon, 192-93 Hamazi, dynasty of, 96-99, 156
E-shar, 149 Hammurabi. 113, 193, 195-98; — dynasty,
Esh-barra, 98 193 {see also Babylon, 1st dynasty of); —
Eshnunna(k), 140, 149, 205; see also Tell period 5-6, 10, 129, 191; post period,
38
Asmar Hani, 149
Eshtar-muti, 108-9, 110, 159-61, 168, and Hatanish, 98-99
Tables I-II
Hebrew name, 78
Etana, 56, 80-81, 142-43, 145-46, 152, 165, Huh(u)nuri, 201
156-57, 162, 168, and Tables I-II
Fara names, 150, 187-90; — texts, 84, 147, Iarlagab, 118-19, 206, and Table II
150-51, 184-^85, 187-89; see also Shurup-
pak Iarlagan, 120, 150
Mood, see Deluge Iarlaganda, 120-21, 206, and Table II
formulas: for change of dynasty, 14, 28-37,
39, 41, 43-44, 4 6 ^ 9 , 55, 61-64, 100, 131,
oi.uchicago.edu
Iarlagash, 118-19, 207 Isin, 109, 138, 195-99; dynasty of —, 5-10.

Iarla<ngab>, 118-19, 206, and Table II 12, 67-68, 109, 124-27, 136, 197-99, and
lasM-ila, 124 Table II; — period, 99, 105, 128-29,
*Iasbi-Irra, 124 135-36
Ibate, 118-19, 206, and Table II Isin-Larsa period, 8, 38, 61, 89-90, 98, 129,
I(b)bt-S!n, 12, 122-23, 130, 134, 174-75, 137, 145, 148, and Table II
197, 199-201, and Table II Mr-pisha, 126-27 and Table II
Ibranum, 118-19, 206, and Table II J, 11, 12-13, 86, 38, 41-42, 49-50, 54-56,
Idamaraz, 195 120-24, 131, and plate at end
I(d)din-Dagan, 68, 124-25, and Table II Jamdat Nasr period, 189
I(d)din-Eshtar, 126
K, 11, &?-&?, 37-38, 41-42, 48, 50, 55-57, 66,
ideograms, 113; pseudo , 18, 106
80-82, 129, 131
ideology of the King List, 138-41, 158,
162-64 Ka-ku(g), 100-101, 172, 175-76, 178, and
Igeshaush, 12, 118-19, 206-7, and Table II Tables I-II
Igigi (var. Irgigi), 24-26, 52, 112-18, 114, Kalbum, 96-97, 168, and Tables I-II
177, 206, and Table II Kalibum, 15-18, 19-21, 55-56, 78-79
Ikun-pi-Eshtar, 8 Kandalanu, 193
II, 139 Kazallu, 138
Ilbaba, see Zababa "king of Kish," 155-56, 181-82
IlkiP (Ilkum), 82-88, 168, and Tables I-II king lists, Assyrian and Babylonian, 191-95;
Ilta-sadum, 82-83, 168, and Tables I-II local —, see date lists, local
Ilulu, see Elulu of Agade kings, missing, 2-3, 158, 165-66, 172-73,
Iluma-ilum, 195-96 180-83, 189
Ilushu-ilia, 140 kingship, descent of, from heaven, 58-59,
Ilu-shuma, 192 65-66; movements of — on earth, 37, 66,
Imbia, see Imta^ 139-40, 158, 162; theory of single -—, 66,
Imi, 24, 26, 52, 114-15, 177, 206, and Table 139-40, 158, 162-65
Kirkuk, 98
Imta 3 (var. Imbia), 12,118-19,120, 206, and Kish, 3, 5, 12, 69, 77, 104, 106, 111, 141-42,
Table II 149, 154-57, 159-63, 167-69, 178-79,
Inanna(k), 86-87, 139, 155, 200, 203 181-83, 188; 1st dynasty of —, 5-7, 9-11,
14-23, 55-56, 67, 73, 76-85, 152-53, 155-
infix - r a - , 62; s e - , 184 56, 157, 186, and Tables I-II; 2d dynasty
Inimabakesh, 12, 118-19, 120-21, 206-7, of —, 96-97, 153, 155-57, 188, and
and Table II Tables I-II; 3d dynasty of —, 11, 15,
Inkishush, 12, 118-19, 206-7, and Table II 53-54, 102, 104-5, 153, 155-57, 159-60,
179, and Tables I-II; 4th dynasty of —,
Inpiq-Hanish, 149 10-11, 53-54, 102, 104, 106-11, 153, 155-
interregnum in the Agade period, 24, 26, 57, 159-61, 178-79, and Tables I-II
52-53, 112-15, 177 Ku>a(ra), 29, 58, 70, 88-89, 142
Irarum, 118-19, 206, and Table II Kudda, 51, 114-15, 116, 206, and Table II
Irgigi, see Igigi Ku(g)-Baba, 15, 53-54, 104-5, 106-8, 110,
Irra (Era), 7, 124 142-43, 146, 159-61, 168-69, 177, 179,
Irra(?)- ,8 and Tables I-II.
Irra-imi(t)ti, 126-27, 130, and Table II Kul , 94-95
Ishbi-Irra, 67-68, 124-25, 199, and Table II Kullab, 90-91, 142-43
Ishkur, 139, 203 Kurlil torso, 186
Ishkur- ,125 Kurum, 118-19, 206, and Table II
Ishm^-Dagan, 9, 124-25, and Table II Kutir-Nahhunte, 10
Ishm£-Shamash, 108-9, 159-61, 168, and Kuyunjik, 11
Tables I-II Li, 9, 13, 24-27, 28, 88, 35-36, 38, 41-42,
Ishu-il, 12, 106-7, 159-60, 176, and Tables 48-46, 47, 49-50, 52-55, 56, 94-104,
106-8, 110-21, 131, 173, 178
L2, 9, 83, 38, 42, 55-56, 76-78
oi.uchicago.edu
INDEX 213
Lata IB, 90-91, 170, and Tables I-II Mari, see Maeri
Laerabum (Laerab, Lasirab), 118-19, 206, Mashda, 15-16, 17-18,20, 21-22, 79, 80-81
and Table II; inscription of —, 119, 203 Mash-Sud, 151
Lagash, 114, 138, 149-51, 155, 157-59, 161, Melam-Anna(k), 92-93, 170, and Tables
179-81, 184-87, 202, and Tables I-II I-II
Lal-ur-alimma(k), 72 Melam-Kishi(k), 82-83, 155, 168, and
Lamkurru, 75, 86 Tables I-II
language, see dating, criteria for; also Menes, 189
grammar and orthography Mes-Anne-pada, 2, 92-95, 155, 158, 172-73,
Larak, 58-59; antediluvian dynasty of —, 181-82, 189, and Tables I-II
60, 74~75, 76 MES(?)-Q6, 92-93, 142-43, 170, and Tables
Larsa, 5, 72, 75, 137-38, 197-98, 201; ante- I-II
diluvian dynasty of —, 58, 71-72, 74; Me-silim, 3, 149, 181-82; inscriptions of —,
dynasty of —, 197-98 and Table II; — 149, 151
list, see WB; — period, see Isin-Larsa Mes-kalam-du(g), 181
period Mes-kiag-gasher, 21, 84-87, 92, 142-^3,170,
Lasirab, see Laerabum 188, and Tables I-II
UlM-demon, 18, 29, 90-91, 142 Mes-kiag-nunna(k) (var. Mes-kiag~Nan-
limmu lists, 191-92 na(k)), 94-95, 172-73, 189, and Tables
Lipit-Enlil, 124-25, 135, and Table II I-II
Lipit-Eshtar, 124-25 and Table II Mesopotamia, relations of, with Egypt, 189
Lugal , 181-82 missing kings, see kings, missing
Lugal-anda, 149-50, 180; statue of —, 99; Muati, 120
time of —, 89, 149, 187 Muati-qu(b)btsin, 120
Lugal-Anne-mundu, 102-3 Muru, 139-40, 203-4
Lugal-banda, 56, 88-89, 91, 142-43, 146, Nabonidus, 194
170, and Tables I-II Nabu of Borsippa, 120
Lugal-kinishe-dudu, 100-101,170,172,175- names, Akkadian, in Sumerian context, 16
76, 183, 184, 190, and Tables I-II names, meanings of: A-lulim(ak), 71;
Lugal-kisal-si, 3, 100-101, 170, 172, 175-76, Arwi'um, 17-18, 80; Etana, 80-81; Gil-
and Tables I-II games, 188-89; Hatanish, 98; Inpiq-
Hanish, 149-50; Kalibum, 17, 78; Ku(g)-
Lugal-ki-tun, 92-93, 170, and Tables I-II Baba, 104; Lugalmu, 97; Magalgalla, 96;
Lugal-me-lam, 51, 54, 116 Man-ishtushu, 112; Mashda, 17-18, 80;
Lugalmu, 96-97, 168, 183, and Tables I-II Muati~qu(b)b!sin, 120; Nangish-lishma,
Lugal-shag-engur, 149 78; Pala-kmatim, 78; Qalumu(m), 17,
78-79; Tizkar, 83; Zuqaqlp, 79
Lugal-tarsi, 181-82 names, readings of, see readings of names;
Lugal-zage-si, 110-11,160-61,170,171,178- spelling of — in English context, 69, 77,
80, and Tables I-II 82-83
Lumma, 98-99; see also E-Anna(k)-tum Nangish-lishma, 78-79
Maeri (Mari), dynasty of, 10, 102-3, 106, Nahi, see Nanum
156-58 Nanna, 92-93, 172-73
Magalgalla (var. Mamagalla), 96-97, 142- Nannia, 108-9, 142^3, 159-61, 168, and
43, 168, and Tables I-II
Tables I-II
Malgia, 137-38
Nanum (var. Nani), 24, 26, 52,114-15,177,
Mamagalla, see Magalgalla
206, and Table II
Man-ishtushu, 24, 26-27, 69, 112-13, 155, Naplanum, 198-99, 201, and Table II
177, 181-82, and Tables I-II Naram-Sin, 24, 26, 28, 112-13, 117, 177-78,
manuscripts of King List, extant, see G, J, 205-8, and Table II
K, Li, L2, P2, P3, P4, p6, Pe, S, Sui, Su2, nasalization, 119
SU3+4, WB; hypothetical ancestral —, see Nebuchadnezzar I, 194
A, B, Su, X, Y, Z Neo-Babylonian period, 70, 145
Marbiti-apal-usur, 11 Nidaba, 76-77, 88
Marduk, 203
Marduk-nadin-ahe, 194r-95
oi.uchicago.edu
Nigi, 86 phonetic writings, see orthography

Ningirsu(k), 138, 149 preposition "after," 66; — "with," 112
Ninhursag, 60 Puzur-ili, 51,114-15, 116, 206, and Table II
Ninsun, 91, 204 Puzur-Nirah (Puzur-Sahan), 12, 104,106-7,
Nintu, temple of, in Adab, 102 159-60, 176, 179, and Tables I - I I
Nin-tur, 172 Puzur-Stn of Gutium, 120-21, 206, and
Ninurta, 108 Table II
Ninurta-nadin-shumati, 194 Puzur-Sin of Kish, 53-54,104-5,106-9,159-
Nippur, 5-9, 72, 75, 89, 93, 99, 157, 174, 60,168, 177, and Tables I-II
195-96, 200-201; — lists, 57 (see also Qaltimu(m), 15-20, 78-79
L1-2 and P2-6)
notes, copyists', 77, 97, 100-101, 117, 171 readings of names: Akshak (ug ki ), 106;
"notes," historical and genealogical, 13, 18, Ansud (an-Bu), 103; Bad-tibira(k), 72;
29,31,105,141-47,151,154,157,160,163 Eluldan ( L I - 1 U - U 1 - DAN), 114; Enbi-
nouns, plurals of, 39 Eshtar, 96, 169; En-shakush-Anna(k),
171; Irra, 124; EaPa(ra) (A-gAki/gA-Aki),
numbering of lines, 10, 15, 69 88; Ku(g)-Baba (ku-dBA-tx), 104-5;
numerals, positions of, 38-39, 112, 120-22, Laerab, 119; Lugal-kinishe-dudu, 172;
174-75 Maeri, 29, 102; MES(?)-B;E, 92; Puzur-
numerical system, Sumerian, 97, 108 Nirah, 106, 179; Samug, 82-83; Sarlagab
Numkurru, 75, 86 ( s a r - K i L - l a - g a b ) , 118, 207; Shu-
ruppak, 75; Ur-Zababa(k) (ur-dZA-MAL-
Nur-Adad, Table I I MAL), 107; Zuqaqip, 79
Nur-Ninshubur, 126-27 readings of signs: 15, 31, 46, 77-78, 83, 90,
Nuzi texts, 98 119, 201; BU= s u d i , 103; E N = in, 150;
order of dynasties, see dynasties, arrange- GE=k e 4 ,16; iA-exorix, 114; NAinSui =
ment of; — of rulers, 13,19-23, 59-60, 73, mistake for a5, 34; PA in PA m u - n a - e =
79, 104, 107-8, 114, 166; —- of signs, see p a , 132; TUR-DA-TI den4U~liUiH den-lU,
orthography 113
orthography: amissible final consonants, relative sentences, see grammar
69, 87-88; "causative" - n - , writing of, Rlm-Sln, 10, 197-99, and Table II
131-35, 139; genitive, writing of, 30, 71; Rfmush, 24, 26, 27, 28, 42,110-13, 155, 175,
ideographic writing, 113; infix - § e - or 177-78, 181-82, and Tables I-II
- s i - , 184; order of signs, 18, 88, 172, rulers, missing, see kings, missing
184-86; phonetic writings, 75, 77, 86, 88,
104-6; preposition "with," writing of, 112; S, 9, 13, 24-27, 28, 33-34, 36-38, 41-42,
pseudo-ideographic writings, 18, 106; sub- 43-46, 47-48, 50, 51-54, 55-56, 66, 104-
ject element - n - , writing of, 36, 38, 16, 129, 131, 177-78
41, 85, 131-35, 138-39; suffixes, writing
of, 186 gabeans of Harran, 73
Samsu-iluna, 7, 195-96
P 2 , 6-6, 7, 15-18, 21, 23, 24-27, 28, 31, 37- Samug, 56, 82-83, 168-69, and Tables I-II
38, 41-42, 43-47, 48, 50, 53, 55-57, 64, Samuûm, Table II
69, 78-95, 98-101, 104, 111-12, 114-15,
120-21, 124-25, 131, 143, 171, 173-75 Sargon (Sharru(m)-kfn), 13, 18, 24, 26-27,
P 3 , 6, 15-18, 23, 24-27, 28, 31-32, 38, 42, 107, 110-18, 143, 145-46, 148, 155, 158-
61, 163, 177-79, 181-82, and Tables I-II
48-46, 50, 52-53, 55-57, 64, 78-81, 84~85,
111-14, 178 Sargonic times, 89, 137; pre times, 146-
P 4 , 6-7, 32, 36, 38, 41-42, 49, 51, 55-56, 47, 149-51, 154
116-18 Sarlag, 207-8
P5, 7, 15-18, 23, 32, 36-38, 41-42, 43-46, 48, Sarlagab, 12, 118-19, 206, 207-8, and Table
50, 54-57, 66, 78-84, 122-27, 129, 131,
135, 174-75 Sennacherib, 194-95
P6, 7-8, 32, 38, 41-42, 47, 55-56 Shala, 194
Pala-kmatim, 78-79 Shaimaneser I, 192-93
patronymics, formula for, 18, 21, 28; use of Shamshl-Adad I, 192-93
subject element with —, 30-34 Shar-kali-sharrl, 24-26, 28, 52,112-18, 177-
Peleg, 88 78, 206, 207-8, and Table II
oi.uchicago.edu
INDEX 215
Sharru(m)-khi, see Sargon rangement of; — of rulers, see order of
Shiruduh, 10 rulers
Shu-Durul (Gimil-Durul), 24, 26, 114-15, Sud, 151
177, 205-6, and Table II Sud-KA-zida, 151
Shu-ilishu (Gimil-ilishu), 68, 109, 124-25, suffix - b i, 37, 40, 84
and Table II suffixes, use of, as dating evidence, 186
Shulgi (Dungi), 89, 122-28, 124, 130, 133, Sumu-abum, 7-8, 193, and Table II
138, 174-75, 203, and Table II Sumu-El, 138 and Table II
Shulme', 12, 118-19, 206, and Table II Sumu-la-El, 193 and Table II
Shulpaê, 132 sun-god, see Utu
Shuruppak, 58-59, 69, 151; antediluvian Susa, 10-11, 25,95,201; — lists, see Elamite
dynasty of —, 56, 58, 60, 68, 74~77 versions of King List and Sui, SU2, Su3+4
Shuruppak (ki), 76 Syriac names, 78, 87, 89-90
Shu-Sin (Gimil-Sin) of Akshak, 106-7, 159- Si-e, 96-97, 168, and Tables I-II
60, 176-77, and Tables I-II
Shu-Sin (Gimil-Sin) of Ur, 12, 98, 122-23, Tammuz, 91; see also Dumu-zi(d)
129-31, 133-36, 174r-75, and Table II Tell Asmar, 119; see also Eshnunna(k)
sign forms as dating evidence, see dating, Telloh, 181-82; see also Lagash
criteria for texts of King List, see manuscripts
signs, order of, see orthography: order of
signs; readings of —, see readings of signs Theodore bar K6ni, 87, 89-90
§illi-Adad, Table II Tiglathpileser, 194-95
Simashu, 156 Tirigan, 120-21, 138-39, 205-6, and Table
Simu-dar, 108-9, 110, 159-61, 168, and Ta- II
bles I-II titles: "king of Kish," 155-56, 181-82;
Simurum, 200 "king of the four regions," 117, 203; "off-
spring of Dagan," 113; "offspring of En-
Sin, 203 lil," 113; 1 u g a 1 k a 1 a m - m a, 171
Sin-i(d)dinam, Table II Tizkar, 55, 82-83,168, and Tables I-II
Sin-iqisham, Table II translations, explanations of, 36-37, 58-59,
Sin-iribam, Table II 61-62, 77, 84, 86, 88, 90-91, 105, 113,
Sin-magir, 5, 126-27, 135, and Table II 137-38, 148, 198-99, 201; see also names,
Sin-muballit, 197-99 meanings of
Sippar, 58-59; antediluvian dynasty of —, transliteration, conventions observed in, 16,
56, 60, 68, 74-75 69
Sim(m), 120-21, 150, 206, and Table II transposition of signs, see orthography: or-
spelling of Sumerian and Akkadian names in der of signs
English context, 69, 77, 82-83 Tukin-hatti-migrisha, 201
statues of gods, 194; — of kings, 99 Tukultl-Ninurta I, 192
stratigraphy, 181
Su, 10, 28, 44-48, 50, 53-55, 174 al-<Ubaid, 2, 184
Sui, 10, 15-18, 23, 24-27, 28, 84, 36-38, 40, Ubar-Tutu(k), 58, 60, 69, 74-76
41-42, 43-46, 47-50, SI, 53, 54-55, 56-57, Umma, 90, 120, 139, 150, 179, 181
64, 78-82, 86-91, 99-104,106-12, 114-16, Ummimi, 149
122-25, 131, 174-77 Undalulu, 12, 106-7, 159-60, 176-77, and
Su2, 10, 84-35, 37-38, 40, 41-42, 46-47, 48, Tables I-II
50, 54-55, 56, 69, 76-78, 82-85, 92, 131 Unzi, 12, 106-7, 159-60, 176, 181, and Ta-
Su3+4, 10-11, 24-27, 28, 35, 36-38, 40, 4 1 - bles I-II
42, 43-46, 47-50, 51-55, 56, 98-100, 104- Ur, 69,141, 155, 157, 161, 167, 172-76, 181-
18, 120-24, 131, 171, 174, 178 82, 184, 186, 188, 196, 200-202, 204; 1st
Su , 96-97, 168, and Tables I-II dynasty of —, 5, 9, 92-95, 153, 155, 157-
subject element - e, 17, 29-36, 135 58, 173-75, 181, 184-86, and Tables I-II;
2d dynasty of —, 10, 100-101, 153, 155,
subject element - n - , writing of, see orthog- 157, 178-76, 178, 183, and Tables I - I I ;
raphy 3d dynasty of —, 7, 10-13, 67-68,122-25,
succession of dynasties, see dynasties, ar- 136, 174-75, 197, 199-202, 204-5, and
oi.uchicago.edu
Table II; period of 3d dynasty of —, 98, Utu-hegal, 12-13, 89, 120-21, 136, 138-41,
112,129, 131,139-40, 147 146, 202-4, 205-6, and Table II
Ur- ,8 Utuk, 97, 187
Ur-Du(l)kuga(k), 126-27 and Table II Utul-kalamma(k), 90-91, 170, 188, and
Ur-E-MM(k), 149 Tables I-II
Ur-Engur, see Ur-Nammu(k) value of variants, 36, 40-43, 46-49
Ur-gigir(ak), 51, 114-15, 116, 206, and Ta- values of signs, see readings of signs
ble II
Ur-Ishkur(ak), 98 verb, identifying (-am), 30-32, 36, 67-68,
Ur-lugal(ak) (var. Ur-Nungal(ak)), 90-91, 136; transitive —-, 36-37
182 verbs, collective form of, 30, 33, 36, 39, 62-
Ur-Nammu(k) (Ur-Engur), 67-68, 89, 122- 63, 99, 102, 117, 129-30, 135; passive—,
23, 132, 138, 174-75, 202-4, and Table II 36-37, 87; singular —, 30, 34, 37, 40, 62-
Ur-Nanshe(k) (Ur-Nina), 158-59, 184-88, 63
190, and Tables I-II; statue of -—, 99; vowels, assimilation of, 18, 75-76, 83, 171,
time of —, 187-88, 190; — level at Telloh, 198
182
Ur-nigin(ak), 51, 114-15, 116, 205-6, and Walih, see Balih
Table II Warad- . . . . , 8
Ur-Nind, see Ur-Nanshe(k) Warad-Sin, 10, 138, and Table II
Ur-Ninurta(k), 21, 124~25, and Table II WB, 5, 11, 13, 15-16, 18, 21, 23, 24, 26-27,
28-31, 36-38, 41-47, 49-50, 51-58, 59-68,
Ur-Nungal(ak) (var. Ur-lugal(ak)), 90-91, 69, 70-127, 131, 135-36, 143, 169-71, 173-
170, and Tables I-II 74, 177
Ur-Stna(k), 98-99 writings, see orthography
Uruk, 8, 21, 69, 85-87, 89, 91, 139, 141-43,
154-57, 161, 163, 167, 169-72, 184, 188, X, 47, 48, 50, 55
196, 204-5; 1st dynasty of —, 6-7, 10,
55, 84-93, 151-53, 155, 157, 171, 182,186, Y, 43, 44-46, 48, 50, 54-55, 66-67, 175
and Tables I-II; 2d dynasty of —, 11, Yamutbal, 195
98-101, 153, 155, 157, 169-72, 183-84,
and Tables I-II; 3d dynasty of —, 10-11, Z, 47-48, 50, 55
110-11, 153, 155, 157, 160-61, 171, 178- Zabaâ, Table II
80, and Tables I-II; 4th dynasty of —, Zababa (Ilbaba), 107, 182
6, 9-11, 50-51, 54, 114-17, 120, 153, 155,
157, 171, 205-6, and Table II; 5th dynas- Zabshali, 201
ty of —, 11-12, 120-23, 171, 202-6, and Zabum, Table II
Table II; — period, 189 Zambia, 126-27 and Table II
Uru-KA-gina, 150, 179-80, and Tables I-II; Zarikum, 150
—- period, 187-88; — texts, 148, 184-85 Zi-u-sud-ra, 59-60/76, 103
Ur-ur, 12,106-7,159-60,176-77, and Tables Zuqaqip, 15-21, 78-79
I-II
Ur-Utu(k), 51, 116-17, 205-6, and Table Zuzu, 181
. . . .-alimma(k), 72
Ur-Zababa(k), 13, 105, 106-9, 110-11, 143, . . . .bi-MU§4-MAs, 102-3
160-61, 163, 168, 178-79, and Tables I -
II gi, 100-101, 172, 176, and Tables I-II
Ur-zaged, 181-82 . . . .kidunnu, 72
Ush, 139 . . . .-lugal, 102-3
"Osl-watar, 108-9, 159-61, 168, and Tables - 1 u - g a 1, 102-3
I-II . . . .ni, 102-3
''usurper'' in Agade dynasty, 28 . .uggalgim, 86-87
Utu, 21, 84-85, 142-43, 151 . . . .zi, 102-3
C PRINTED 1
IN U S A J
oi.uchicago.edu
K 11624
CUNEIFORM T E X T OF J AND OF E P I C FRAGMENT K 11624

oi.uchicago.edu

Sume

Uploaded by

Copyright:

Available Formats

Sume

Uploaded by

Document Information

Original Title

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

Sume

Uploaded by

Copyright:

Available Formats

oi.uchicago.

T H E ORIENTAL INSTITUTE OF THE UNIVERSITY OF CHICAGO

JOHN ALBERT WILSON & THOMAS GEORGE ALLEN • EDITORS

THE ORIENTAL INSTITUTE OF THE UNIVERSITY OF CHICAGO

ASSTRIOLOGICAL STUDIES . NO. 11

THE UNIVERSITY OF CHICAGO PRESS • CHICAGO • ILLINOIS

Internet publication of this work was made possible with the

International Standard Book Number: 0-226-62273-8

TO O. E. RAVN, TO THE MEMORY OF EDWARD CHIERA

LIST OF ABBREVIATIONS xiii

IV. COMPOSITION 128

xii TABLE OF CONTENTS

The Uruk Source 169

PLATE. CUNEIFORM TEXT OP J AND OF EPIC FRAGMENT K 11624 at end

ITT Istanbul. Asan Atika Muzeleri. Inventaire des tablettes de Tello

PBS X I 1 CHIERA, EDWARD. Lists of personal names from the temple

WZKM Wiener Zeitsehrift fiir die Kunde des Morgenlandes (Wien,

6 THE SUMERIAN KING LIST

P5 CBS 19797.10 Published by Poebel, PBS V, No. 5.11 Transliteration

dynasty summary. 15 P 6 was found in Nippur by the Babylonian Ex-

U CBS 14220. Published by Legrain, PBS XIII, No. I.16 Transliteration

L2 CBS 14223. Published by Legrain, PBS XIII, No. 2. Transliteration

S BM 108857. Last published by Gadd, The Early Dynasties of Sumer

10 THE SUMERIAN KING LIST

SUI Published by Scheil in RA XXXI (1934) 150 f. Transliteration and

K K 8532+K 8533+K 8534. Published by L. W. King, Chronicles con-

12 THE SUMERIAN KING LIST

G Published by H. de Genouillac, Fouilles frangaises d'el-^Akhymer. Pre-

DERIVATION FROM A SINGLE ORIGINAL

14 THE SUMERIAN KING LIST

The sections in which we have relatively many manuscripts overlapping so

tion in the position of the 3d dynasty of Kish (Ku(g)-Baba). We may begin

18 THE SUMERIAN KING LIST

stated to be the son of his predecessor: "A reigned x years; B, son of A,

the repetition of Atab as Atabba will be satisfactorily explained.

To a division of our sources very similar to that indicated by the Atabba

26 THE SUMERIAN KING LIST

Instead of r i m u § d u m u § a r r u - k t n m u 9 i - a5 the scribe would thus

28 THE SUMEKIAN KING LIST

the erroneous copy. Instead we must assume that it represents a different

The occurrence of the Atabba variants could be observed in five of our

1. Formula for change of dynasty: A ki giSt u k u 1 b a - a n - s i g n a m - l u g a l -

g [ u - t u - u m k i ] (!)52 n a m - l u g a l - b i u n u g k i - § & [ b a - t u m ] ; viii 7-8:

47: g u l - l a d n i d a b a - a n - n a - r d a1 s i k i l 5 8 m u 960 l - a 5 ; ii 7-8: k a - l f -

change to singular 1 u g a 1 and i n - a 8 . Occurrences: obv. 7: 6 l u g a l - e - n e

2. Formula for dynasty total: x l u g a l m u - b i y m u 1 - a 6 . Occurrences:

i 1: [N. m u x] i n - a 6 ; ii 2: m b a 1 i fr . . . . [mu x i n - a5] ; ii 3: m e n - m e n -

(RA IX [1912] 112-13 i 1-6), which contains our phrase: g u - t i - u m ki 1u

c) In WB, P«, Li, and J(?):

usage in which the numeral precedes.88 Application of later grammatical

40 THE SUMERIAN KING LIST

The last form, x l u g a l - e - n e m u y i n - a k - e § , i s probably due to

ATABBA VARIANTS AGADE VARIANTS

44 THE SUMERIAN KING LIST

We may next look at Possibility II, which, like Possibility I, presupposes

WB P2 P3 P5 S Su7~Su^ L,? (or L, here?)

46 THE STJMERIAN KING LIST

WB P2 P3 P5 5 Su, S u ^ L, (or, less likely, L, here)

48 THE SUMERIAN KING LIST