1975 Book IntersexualityInTheAnimalKingd

Intersexuality
in the Animal Kingdom
Edited by R. Reinboth
With 221 Figures
Springer-Verlag
Berlin Heidelberg New York 1975
Professor Dr. Rudolf Reinboth
Institut fUr Zoologie
der Johannes Gutenberg-Universitat
0-65 Mainz, SaarstraBe 21
ISBN-13 978-3-642-66071-9 e-ISBN-13: 978-3-642-66069-6

001: 10.1007/978-3-642-66069-6
This work is subject to copyright. All rights are reserved, whether the whole
or part of the material is concerned, specifically those of translation, reprinting,
re-use of illustrations, broadcasting, reproduction by photocopying machine
or similar means, and storage in data banks.
Under § 54 of the German Copyright Law, where copies are made for other
than private use, a fee is payable to the publisher, the amount of the fee to be
determined by agreement with the publisher.
© by Springer-Verlag Berlin' Heidelberg 1975.
Softcover reprint of the hardcover 1st edition 1975
Library of Congress Cataloging in Publication Data. Symposium on Inter-

sexuality in the Animal Kingdom, Akademie der Wissenschaften und der
Literatur, 1974. Intersexuality in the animal kingdom. Bibliography: p.lncludes
index. 1. Hermaphroditism--Congresses. 2. Sex--Cause and determination--
Congresses. I. Reinboth, Rudolf, ed. II. Title. [DNLM: 1. Hermaphroditism--
Congresses. 2. Sex determination--Congresses. 3. Sex reversal, Gonadal--
Congresses. WJ712 S 986i 1974]. OP267.S94. 1974. 591.1'6'67. 75-4585
The use of registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general
use.
Offsetprinting: Julius Beltz, Hemsbach/Bergstr. Bookbinding: BrOhlsche
Universitatsdruckerei, GieBen.
Preface
When Richard Goldschmidt' coined the term "intersexuality" in 1915,

he intended it to apply to normally dioecious species which exhibit
some kind of mixture between male and female characters. However, as
knowledge of the bewildering variability present in the sexual orga-
nization of members of the animal kingdom has increased, the original
meaning of the word has changed. Today many authors define inter-
sexuality as "the presence of both male and female characteristics,
or of intermediate sexual characteristics, in a single individual".2
This more extensive and widely accepted concept justifies the title
of our book •.
Among all the anatomical and physiological features of living organisms

the reproductive system has a unique importance for the perpetuation
of the species. Conversely, reproductive processes are of little or
no account for the viability of the individual. Therefore, within the
framework of general biology reproduction has all too often been looked
at solely from the point of view of genetics. Lively discussions about
genotypic versus phenotypic sex determination long dominated the sci-
entific literature on sexuality in animals; this one-sided emphasis
has tended to obscure many important facets of an organism's ability
to reproduce.
Recent developments in current biological research have brought the

classic problem of sex differentiation into focus again, and the rapid
progress being made in comparative endocrinology has added a new di-
mension to the study of reproductive biology.
Whatever is understood by intersexuality, the term always implies

some deviation from the usual gonochoristic pattern and acknowledges
the indisputable fact that sex depends on a strict dualism of "male-
ness" and "femaleness", at least at the level of the gametes. The
basically disjunctive character of the two types of sexual potency
means that the occurrence of intersexual phenomena links the two
extremes, so offering an insight into nature's differentiation be-
tween "male" and "female" and the processes involved.
Such ideas were the leitmotif for a symposium that brought together
research workers from such varied biological disciplines as endo-
crinology, developmental physiology, genetics, cytology, ecology
and evolutionary theory. The papers presented there have been edited
for this book, which gives a fairly detailed picture of the problems
'Biol. Zbl. 35, 565 (1915).

2Atz ,J.W.: In: Intersexuality in Vertebrates Including Man (C.N.
Armstrong and A.J. Marshall, eds.). New York: Academic Press, 1964.
VI
and objectives of current research on intersexuality. No claim is

made to present an exhaustive review of current knowledge, but the
reader will find ample information in the literature cited.
At first sight the heterogeneity of the contributions is rather

startling. Some authors reported details of original studies where-
as others tried in different ways to provide a synoptic review of
certain topics. Studies on intersexuality are being carried out on
various groups in quite different directions. For example, in teleost
fishes surprising discoveries of a wide array of different types of
natural intersexuality have aroused the curiosity of research workers
in different disciplines, whereas in other classes (e.g. turbellaria,
echinoderms, etc.) the scope for study is more limited because less
is known about the biology of these animals. The reader may look in
vain for information on some taxonomic classes in which intersexual
organization is either the rule (trematodes and cestodes) or fairly
cornmon (nematodes) because of the few recent findings available.
Professor Nigon of the University of Lyon, France, an expert on
nematodes, wrote to me: "Since my own studies on sexuality in nema-
todes (of which the investigation of intersexuality represented only
a limited part), this subject does not appear to have been pursued
further anywhere in the world. I am aware of only sporadic observa-
tions in this field ... which cannot be compared with the active
experimental studies being carried out on other invertebrates."
Despite such unevenness, there was unanimous agreement among the

contributors that the exchange of ideas about intersexuality was
stimulating and deserved a wider audience. Moreover, the historical
subdivisions of the biological disciplines have not favoured attempts
to examine intersexuality as a highly complex phenomenon, requiring
joint efforts from various angles.
No single theme is likely to emerge as the "essence" of this book,

but one may risk the opinion that the problem of "bisexual potency",
a term introduced by Max Hartmann in 1923,3 is an important reference
point. Answers have yet to be found to the question as to which pro-
cesses are involved in inducing (bi-?) potential (germ) cells to stop
mitotic divisions and to enter a development (by meiosis or similar
events) which leads irreversibly to the formation of clearly polarized
male and female gametes.
The critical reader is also asked to keep in mind that one of the
unavoidable obstacles facing the editor of such a symposium is that
English is not the mother tongue of all contributors. Both publisher
and editor made great efforts to achieve an evenness in grammar and
style, by revising and sometimes even retranslating. Nevertheless,
rapid publication called for an acceptable compromise that did not
allow for "smoothing the rough edges".
Please note that the index is merely meant to supplement the key words
of the article titles. Terms basic to the general contents (e.g. gonado-
genesis, germ cells, gametogenesis, ovary, testis, gonoducts, sex dif-
ferentiation, etc.) have deliberately been omitted because they are
dealt with in almost every contribution.
3Studia Mendeliana Brlinn, 203 (1923).

~I
The Symposium on Intersexuality in the Animal Kingdom was held in the

Akademie der Wissenschaften und Literatur at Mainz on 8-12 July 1974.
The participants are much obliged to the hosts, who provided a most
pleasant environment for our discussions. We acknowledge the generous
financial help granted by the Deutsche Forschungsgemeinschaft, Bonn,
the Akademische Auslandsamt der Universit~t Mainz, the government of
Rheinland-Pfalz, the Vereinigung der Freunde der Universit~t Mainz,
the Humboldt-Stiftung, Bonn, Farbwerke Hoechst AG, and C.H. Boehringer
Sohn, Ingelheim. Members of the staff of the Akademie der Wissenschaf-
ten and the Zoological Institute of the University of Mainz gave us
broad support. Dr. J.A.P. Mehl as copyeditor has worked intensively
and constructively on most of the manuscripts. Without the personal
engagement and cooperation of my secretary, Mrs. A. Reichow, the
whole project would never have materialized.
The publisher met all our interests in a very efficient and always
friendly and sympathetic way.
To all persons and institutions who lent us their support I am deeply

bound in gratitude.
March 1975 R. Reinboth

Contents
INVERTEBRATES
Sex and Sex Determination in Coelenterates •••.•..•.••••.••....••

P. Tardent
Sexual Differentiation in the Fissiparous Strain of Dugesia

gonoaephaZa • • . • • • • • • . • • • • • • • • • • • • • • • • • • • • . • • • • • • • • • • • • • • • • • • • • • 14
Th. Lender and C. Brian90n
Sexuality and Neurosecretion in Freshwater Planarians •.•••.•.••• 20

M. Grasso
Sex Differentiation in Regeneratingd' I ~ Nemertine Chimeras ••.••• 30

J. Bierne
Sex Reversal in the SyUinae (Polychaeta: Annelida) 41

M. Durchon
Mutual Influence on the Sexual Differentiation in the Protandric

Polychaete Ophryotroaha pueriUs •• • • • • • • • • . • • • • • • • • • • • • • • • • • • • • • • 48
H.-D. Pfannenstiel
Neurosecretory Phenomena during Reproduction in Oligochaeta .••.. 57

H. Herlant-Meewis
Study of Sex Control of Gametogenesis by Organ Culture in the

Oligochaete Annelid Eisenia foetida f. typiaa Sav. .•••..••••..... 64
M. Lattaud
Factors of Primary Sexual Differentiation in the Simultaneous

Hermaphrodite Eisenia foetida (Oligochaeta: Lurnbricidae) •...•.• 72
J.C. Relexans
Sex-Determination in BoneUia ••••••••••••••••••.••.••••••••••.••• 84

R. Leutert
Hermaphroditism and Gynandromorphism in Malacostracan Crustacea. 91

H. Charniaux-Cotton
Temperature-Sensitive Intersexuality and Its Determinism in

Orahestia gammareZZa Pallas .•.•.••..•.•••.•....••.••••..••..••.. 106
T. Ginsburger-Vogel
Parasite-Induced Castration and Intersexuality in Insects ••.••.. 121

w. Wiilker
Hermaphroditism in Insects. Studies on Iaerya purahasi •••••••••.••• 135
M. Royer
x
Sex Specific Cell Differentiation in Different Types of

Intersexes of Lymantria dispar L. . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
G.C. Mosbacher
Structural and Endocrinological Aspects of Hermaphroditism in

Pulmonate Snails, with Particular Reference to Lymnaea
stagnaZis (L.) • . . . • • . . • • . • . . . . . . . . . . • . . • • • . . . . • . . . . . . . • . • . • . . . 158
J. Joosse
Protandric Hermaphroditism in Prosobranch Gastropods ........... 170

S. Le Gall and W. Streiff
The Sexuality of Pelecypod Molluscs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

P. Herlin-Houtteville and P.E. Lubet
Hermaphroditism in Echinoderms. Studies on Asteroids ............ 188

R. Delavault
VERTEBRATES
The Gonadal and Adenohypophysial Functions on Natural Sex

Reversal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . 201
S.T.H. Chan, Wai-sum 0, and S.W.B. Hui
Ovarian and Testicular Intersexuality in Two Protogynous

Mediterranean Groupers, EpinepheZus aeneus and EpinepheZus guaza ... 222
J. Brusl~ and S. Brusl~
Hermaphroditism in Cichlid Fishes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

H.M. Peters
An Unusual Approach to Experimental Sex Inversion in the Teleost

Fish, Betta and Maeropodus .....••......•.•••.......•...•....... 236
P. Becker, H. Roland, and R. Reinboth
The Possible Significance of Sex-Chromatin for the Determination

of Genetic Sex in Ambisexual Teleost Fishes . . . . . . . . . . . . . . . . . . 243
J.A.P. Mehl and R. Reinboth
Sex Determination and Differentiation among Uniparental

Homozygotes of the Hermaphroditic Fish RivuZus marmoratus
(Cyprinodontidae: Atheriniformes) . . . . . . . . . . . • . . . . . . . . . . . . . . . . 249
R.W. Harrington, Jr.
Protogynous Hermaphroditism in Fishes of the Family Scaridae .... 263

J.H. Choat and D.R. Robertson
Ecology and Physiology of Sex Reversal in Anthias squamipinnis

(Peters), (Teleostei: Anthiidae) .....•...............•....... 284
L. Fishelson
The Evolution of Hermaphroditism in Fishes . . . . . . . . . . . . . . . . . . . . . . 295

C.L. Smith
Unisexual Female Offsprings in the Salamander, PZeurodeZes waUUi

Michah . . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . 311
A. Collenot
XI
Autoradiographic Study on the Mechanisms of Testosterone-

Induced Sex-Reversal in Rana Tadpoles . . . . . . . . . . . . . . . . . . . . . . . . 318
E. Vannini, A. Stagni, and F. Zaccanti
Temperature and Sex Differentiation in Embryos of Two Chelonians,

Emys orbicula:r>is L. and Testudo graeca L . . . . . . . . . . . . . . . . . . . . . . . . 332
C. Pieau
Evolution of Parthenogenetic Species of Reptiles . . . . . . . . . . . . . . . . 340

Ch.J. Cole
Intersexuality in Birds. Study of the Effects of Hybridization

and Post-Embryonic Ovariectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
L. Gomot
Some Effects of Embryonic Gonad and Non-Gonadal Grafts on the

Development of Primary Sexual Characteristics in the Chick ... 375
J. Thiebold
Intersexuality of the Genital System and "Free-Martinism" in

Birds . . . . . . . . . . . . . . . . . . . . . c • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 382
H. Lutz and Y. Lutz-Ostertag
The Prenatal Development of Bovine Freemartins . . . . . . . . . . . . . . . . . . 392

A. Jost, J.P. Perchellet, J. Prepin, and B. Vigier
The Role of Androgens in Sexual Differentiation of Mammals ...... 407

F. Neumann, W. Elger, H. Steinbeck, and K.-J. Graf
Recent Studies on the Intersexual Programming of the Genetic Rat

Male Pseudohermaphrodite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
A.S. Goldman
Chromosomes and Sex Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

U. Mittwoch
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447

List of Contributors
BECKER, Peter, Dr., c/o Celamerck GmbH & Co. KG, 6507 Ingelheim (Rhein),
W. Germany
BIERNE, J., Prof., Faculte des Sciences de l'Universite de Reims,

Laboratoire de Biologie generale, 51062 Reims Cedex, France
BRIANCON, C., Dr., Faculte des Sciences, Laboratoire de Biologie

animale, 91405 Orsay, France
BRUSL~, Jacques, Prof., Centre Universitaire de Perpignan, Faculte des

Sciences,. Laboratoire de Biologie marine, 66000 Perpignan, France
BRUSL~, Solange, Maitre-Assistant, Centre Universitaire de Perpignan,

Faculte des SCiences, Laboratoire de Biologie marine,
66000 Perpignan, France
CHAN, S.T.H., Dr., Department of Zoology, University of Hong Kong,

Hong Kong
CHARNIAUX-COTTON, Helene, Prof., Universite Paris VI, Laboratoire

Sexualite et Reproduction des Invertebres, 75230 Paris Cedex OS,
France
CHOAT, J.H., Dr., University of Auckland, Department of Zoology,

Auckland, New Zealand
COLE, Charles J., Dr., American Museum of Natural History, Department

of Herpetology, New York, NY 10024, USA
COLLENOT, Alain, Prof., Universite Paris VI, 75230 Paris Cedex OS,
France
DELAVAULT, R., Prof., Universite d'Orleans, Laboratoire de Biologie

cellulaire et animale, 45045 Orleans Cedex, France
DURCHON, Maurice, Prof., Universite des Sciences et Techniques de

Lille, Biologie Animale, 59650 Villeneuve-d'Ascq, France
ELGER, Walter, Dr., c/o Schering AG, Dept. Endokrinpharmakologie,

1000 Berlin 65, W. Germany
FISHELSON, Lev, Prof. Dr., Tel-Aviv University, Department of Zoology,

Tel-Aviv, Israel
GINS BURGER-VOGEL , T., Maitre-Assistant, Laboratoire de Genetique

evolutive du CNRS, 91190 Gif sur Yvette, France
GOLDMAN, Allen, Associate Professor, M.D., The Children's Hospital of

Philadelphia, Division of Experimental Pathology, Philadelphia,
PA 19104, USA
XIV
GOMOT, Lucien, Prof., Laboratoire de Zoologie et d'Embryologie de la

Faculte des Sciences, 25030 Besan90n, France
GRAF, Klaus-JUrgen, Dr., c/o Schering AG, Dept. Endokrinpharmakologie,

GRASSO, Mario, Prof. Dr., Istituto di Zoologia, 40126 Bologna, Italy
HARRINGTON, Robert W., Jr., Dr., Florida Medical Entomology Laboratory,

Florida Division of Health, Vero Beach, FL 32960, USA
HERLANT-MEEWIS, Henriette, Prof., Universite Libre de Bruxelles,

Faculte des Sciences, 1050 Bruxelles, Belgium
HERLIN-HOUTTEVILLE, Paulette, Dr., Universite de Caen, Laboratoire de

Zoologie, 14000 Caen, France
HUI, W.B., Dr., University of Hong Kong, Department of Zoology,

Hong Kong
JOOSSE, J., Prof. Dr., Free University, Department of Biology,

Amsterdam-Buitenveldert, Netherlands
JOST, Alfred, Professeur au College de France, Universite Pierre et

Marie Curie, Laboratoire de Physiologie comparee, 75005 Paris,
France
LATTAUD, Claude, Dr., Universite Paris VI, Laboratoire Sexualite et

Reproduction des Invertebres, 75230 Paris Cedex 05, France
LE GALL, So lange , Dr., Universite de Caen, Laboratoire d'En~ocrinologie

et d'Embryologie experimentales, 14000 Caen, France
LENDER, Theodore, Prof., Universite Paris Sud, Laboratoire de Biologie

animale, 91405 Orsay, France
LEUTERT, Rudolf, Dr., Zoologisches Institut der Universitat ZUrich,

8006 ZUrich, Switzerland
LUBET, P., Prof., Universite de Caen, Laboratoire de Zoologie,

14000 Caen, France
LUTZ, Hubert, Prof. Dr., Universite de Clermont, Biologie animale,

63170 Aubiere, France
LUTZ-OSTERTAG, Yvonne, Dr., Universite de Clermont, Biologie animale,

63170 Aubiere, France
MEHL, John A.P., Dr., Divison of Sea Fisheries, Cape Town 8001,
South Africa
MITTWOCH, Ursula, Dr., Galton Laboratory, Department of Human Genetics

and Biometry, University College London, London NW1 2HE, Great
Britain
MOSBACHER, Georg Christian, Prof. Dr., Universitat des Saarlandes,

Fachbereich Biologie, 6600 SaarbrUcken 11, W. Germany
NEUMANN, Friedmund, Prof. Dr., c/o Schering AG, Dept. Endokrinpharma-

kologie, 1000 Berlin 65, W. Germany
xv
0, Wai-sum, Dr., University of Edinburgh, MRC Unit of Reproductive

Biology, Department of Obstetrics and Gynaecology, Edinburgh EH1
2QW, Great Britain
PERCHELLET, Jean-Pierre, Chercheur, Universite Pierre et Marie Curie,

Laboratoire de Physiologie comparee, 75005 Paris, France
PETERS, Hans M., Prof. Dr., Institut fur Biologie III, Lehrstuhl
Zoophysiologie, Abteilung Verhaltensphysiologie, 7400 Tubingen,
W. Germany
PFANNENSTIEL, Hans-Dieter, Dr., Zoologisches Institut der Technischen

Universitat, 3300 Braunschweig, W. Germany
PIEAU, Claude, Dr., Laboratoire Pasteur, 95110 Sannois, France
pRtPIN, Jacques, Assistant, Universite Pierre et Marie Curie, Labora-

toire de Physiologie comparee, 75005 Paris, France
REINBOTH, Rudolf, Prof. Dr., Institut fur Zoologie der Johannes

Gutenberg-Universitat, 6500 Mainz, W. Germany
RELEXANS, J.-C., Dr., Universite de Bordeaux I, Institut de Biologie

Animale, 33405 Talence, France
ROBERTSON, David Ross, Dr., Smithsonian Tropical Research Institute,

Balboa, Canal Zone
ROLAND, Heidrun, Dr., 1000 Berlin 45, Pottsdarner Str. 57a, W. Germany
ROYER, Michel, Dr., Universite d'Orleans, Laboratoire de Biologie

cellulaire et animale, 45045 Orleans Cedex, France
SMITH, C. Lavett, Dr., American Museum of Natural History, Department

of Ichthyology, New York, NY 10024, USA
STAGNI, Anna, Prof. Dr., Istituto di Zoologia, 40126 Bologna, Italy
STEINBECK, Hermann, Dr., c/o Schering AG, Dept. Endokrinpharmakologie,

STREIFF, W., Prof., Universite de Caen, Laboratoire de Zoologie,

14000 Caen, France
TARDENT, Pierre, Prof. Dr., Zoologisches Institut der Universitat

Zurich, 8006 Zurich, Switzerland
THIEBOLD, Jean, Prof., Universite Louis Pasteur, Laboratoire de Zoo-

logie et d'Ernbryologie experimentale, 67000 Strasbourg, France
VANNINI, Enrico, Prof. Dr., Istituto di Zoologia, 40126 Bologna, Italy
VIGIER, Bernard, Charge de recherches CNRS, Universite Pierre et Marie

Curie, Laboratoire de Physiologie comparee, 75005 Paris, France
WULKER, Wolfgang, Prof. Dr., Biologisches Institut I (Zoologie) der

Universitat Freiburg, 7800 Freiburg (Breisgau), W. Germany
ZACCANTI, Francesco, Prof. Dr., Istituto di Zoologia, 40126 Bologna,

Italy
Sex and Sex Determination in Coelenterates
P. Tardent
A. General Remarks
The Coelenterata, which include the 3 classes Anthozoa (sea-anemones

and corals), Scyphozoa (jelly-fishes) and Hydrozoa (hydroid polyps
and hydromedusae) constitute an extremely heterogenous phylum, not
only with respect to their morphology (Hyman, 1940) and developmental
habits (Berrill, 1961; Siewing, 1969; Mergner, 1971; Campbell et al.,
1974), but also with regard to their sexual reproduction.
The Coelenterata offer a rich selection of different types of life

cycles. The simplest is that of the Anthozoa, where the sessile or
semi-sessile polyps reproduce both sexually and asexually (by budding
or spontaneous fission). The eggs that are or shed into the water, or
kept in brood-chambers, develop directly or through an intermediate
planula-larva (Korn, 1966; Widersten, 1968) into a polyp (Nyholm,
1943, 1949). In all 9 orders of this class there are gonochoristic and
hermaphroditic species or groups of species. The gonads are imbedded
in the acellular mesoglea of the entodermal septa which subdivide the
voluminous gastric cavity into a variable number of radial chambers
(Hyman, 1940). In hermaphrodites the testes and ovaries can be situ-
ated side by side in the same septum (Nyholm, 1943).
-- ©' Fig. 1. The metagenet-

ic life-cycle of Podo-
cOY'yne carnea M. Sars
(Hydrozoa, Hydroidea).
1 fertilized egg,
2 free-swimming pla-
nula larva, 3 planula-
\
larva attached to the
substratum, 4- 7 part
of a polymorphic asex-
I
ual polyp colony,
4 autozoids (feeding
polyps), 5 nematocyte-
bearing spiralzoid,
6 young gonozoid pro-
~ ducing medusae buds,
7 gonozoids with a
/ fully differentiated
young medusa, 8 free-
swimming adult medusa
with gonads (Go) at-
tached to the manu-
brium. (Modified from
Frey, 1967)
2
On the other hand, the basic type of the life cycles (Fig. 1) of the
recent Scyphozoa and Hydrozoa is metagenetic and includes two alter-
nating generations: that represented by the asexually reproducing
benthic polyps (scyphopolyps and hydroid-polyps); and that of the
free-swimming scypho- or hydromedusae, which reproduce sexually. In
addition, some hydromedusae produce asexual offspring by means of
budding (Russell, 1953; Werner, 1958). In both classes this classical
alternation of asexual and sexual generations undergoes various modi-
fications, leading to a partial or total suppression of one or the
other of the two generations (Kuhn, 1914). Wherever the medusa is sup-
pressed the sexual functions are transferred to the remaining polyp
generation.
Most Scyphozoa and Hydrozoa are gonochoristic, i.e. dioecious. This

means that a polyp or a clone of polyps will always produce medusae
of one sex only and that the state of sex determination is transmitted
from one subsequent asexual generation to another. This stability of
the state of determination speaks in favour of a genotypic mechanism
although, for the present moment, we lack any reliable information in
support of this statement (the chromosomes of the coelenterates are in
general very small, and no true genetic experiments have been performed
so far). Besides gonochorism, there exist in both classes cases of
temporary and normal hermaphroditism (see below) .
B. The Origin of Germ Cells
In coelenterates there is, as already emphasized by Weismann (1883)

and Klihn (1914), nothing comparable to a true "germ-line" (Keimbahn)
The gametes derive from the so-called interstitial cells (I-cells),
which represent an omnipresent population of small basophilic cells
situated in the ecto- and entoderm of both polyps and medusae (Fig.
2a). The literature (Tardent, 1954; Weiler-Stolt, 1960; Brien, 1966)
offers a broad selection of different views about the origin and func-
tion of these cells, which are capable of moving freely in the inter-
cellular spaces of the epithelia (Tardent and Eymann, 1959; Tardent
and Morgenthaler, 1966) and passing through the mesoglea from one
layer to the other (Glatzer, 1971). From studies in Hydrozoa (parti-
cularly Hydra) it is known that the I-cells can differentiate into a
number of various somatic cells, including nematocytes (Slauterback
and Fawcett, 1959; Lentz, 1966), neurones (Burnett and Diehl, 1964),
glandular cells and others (Fig. 3). On the other hand, they will,
when the polyp or medusa initiates gametogenesis, become oogonia and/
or spermatogonia. There is no evidence whatsoever for the presence
amongst the interstitial cells of two different lineages i.e. a somatic
and a gametic population. In Hydra it is known (Brien, 1966) that mor-
phogenetic events like budding or regeneration on one hand, and gameto-
genesis on the other, are competing for the same type of omnipotent
cells to the extent that in some species the simultaneous occurrence
of these events is excluded. It is also a fact that, when gametogenesis
has occurred in Hydra, the entire population of these cells is used up
in the neighbourhood of the gonads (Fig. 2b). If there were 2 separate
lineages of morphologically undistinguishable I-cells, at least the
somatocytic population (not being involved in gametogenesis) should
persist. In Hydra fusca L. (Brien, 1966) a continuous gametogenetic
activity can lead to a complete exhaustion of the polyp, because there
are no more I-cells available for the replacement of somatic elements,
such as nematocytes and nerve cells. Brien (1966) calls this a "gametic
crisis".
3
Fig. 2 a-d. Histology of Hydra attenuata (Pall.). (a) Part of a cross-

section (7 ~m) through the body wall of a polyp showing the ectoderm
(EKT) and the entoderm (ENT) with the acellular mesoglea eM) in be-
tween. Note the interstitial cells (IZ) and the nematocytes (NC) be-
tween the ectodermal epithelio-muscular cells (EMC). ( DC = digestive
cells, CC = glandular cells). x 445 . (b) Cross section (7 ~m) through
the body wall of a male polyp after spermatogenesis had taken place.
Note the absence of interstitial elements (I-cells and nematocytes)
in the ectoderm. x 393. (c) Cross section through a hermaphroditic
gonad showing oogenesis and spermatogenesis occurring side by side
(OC = oocytes, OC = oogonia, S C = spermatocytes). x 217. (d) Cross
section through the body wall of a male polyp 4 days after having been
irradiated with 6500 r. Note the complete absence of I-cells and nema-
tocytes. x 440.
(DC = digestive cells, EKT = ectoderm, EMC = epithelio-muscular cells,
ENT = entoderm, CC = glandular cells, IZ = interstitial cells, M =
mesoglea, NC = nematocytes, OC = oocytes, OC = oogonia, SC = spermato-
cytes)
4
2~
~
,
.. / ,/ I
I
/
/
Fig. 3. Schematical drawing showing the possible performances of inter-

stitial cells (1) of Hydr a as to their abilities of somatic and gametic
differentiation.
1 Interstitial cell (I-cell).
2-8 EctodeY'l7la l c e U types:
2 epithelio-muscular cell,
3 sensory nerve cell,
4 neuron
5 desmoneme,
6 atrichous isorhiza,
7 holotrichous isorhiza,
8 stenothele ( 5- 8 = nematocytes)
9-11 Entodermal ce l l types:
9 digestive cell (entodermal epithelio-muscular cell) (the broken
arrow indicates that it is not certain yet, whether digestive
cells of the entoderm are derived from I-cells) ,
10 gland cell,
11 mucuous cell.
12-1 3 game tes:
12 oocyte,
13 sperm.
In my view the I-cells do not necessarily represent a continuous lin-

eage of cells, the origin of which reaches back in the early ontoge-
netic history of the individual. They appear relatively late in devel-
opment in the ecto- or endoblast of the larva, where they first pro-
5
duce nematocysts (Weiler-Stolt, 1960). It is still uncertain whether

all I-cells which are propagated from one asexual generation to an-
other, or from the polyp to the medusae, derive from this original
stock or not. We know, on the other hand, that under particular exper-
imental conditions somatic cells can, by a process of de-differentia-
tion, return to the state of I-cells, the descendants of which may
later become oogonia or spermatogonia. As shown by Burnett et al.
(1966), isolated entoderm of Hydra (lacking interstitial cells) is
capable of reconstituting a normal polyp, which later will produce
gametes. Consequently, the somatic cells of the entoderm, such as
digestive and glandular cells, have not only the ability of replacing
the missing somatic part of the ectoderm, but also the lacking gono-
cytes. In the view of these and other facts, the problems concerned
with the initiation of gametogenesis and the determination of sex are,
at least in the Hydrozoa, closely related to those of somatic determi-
nation and differentiation. The somatic portion of the gonad, if pres-
ent, does not seem to playa role in this process, because - as will
be shown later - the potential oogonia or spermatogonia are determined
as such, before they reach the gonad or the site of gonad formation.
In this maze of possible developmental pathways every indifferent I-cell

is first confronted with the choice between somatic and gametic differ-
entiation. Irt the first case, it is then faced with a series of alter-
natives such as nematoblasts, neuroblasts, etc. (Fig. 3). In the second
case, determination has to decide whether the cell is to become an
oocyte or a spermatocyte. How this decision is brought about is still
a matter of speculation. Wherever we deal with species exhibiting a
clear-cut gonochoristic behavior, we may postulate a chromosomal mech-
anism of sex-determination. In the cases of functional hermaphroditism,
or of regularly or accidentally occurring sex inversions (Tardent,
1966a), we must assume that determination of the sexual status occurs
under the influence of "environmental" factors.
C. Sex Determination in Hydra
Most of the observations and experiments aiming at the identification

of these factors have been performed on various species of the fresh-
water Hydra (Hydrozoa, Hydroidea). Amongst this genus there are true
gonochorists (H. fusea L.); unbalanced gonochorists (H. attenuata Pall. ) ;
and hermaphrodites (H. viridis L., H. eireumeineta, Schulze) species,
the gametogenesis of which has been described in detail by numerous
authors (Brien, 1966; Zihler, 1972).
According to a model offered by Burnett (1962, 1967), the somatic as

well as the gametic determination and differentiation of cells in Hydra,
are subjected to the control of 2 antagonistically acting substances:
an inducer produced by the neurons (Burnett, 1967; Schaller, 1973;
Schaller and Gierer, 1973); and an inhibitor, the origin of which is
not yet known for certain. The relevant information for determining
the fate of a differentiating I-cell resides, according to this model,
in the relative concentrations of these two factors at a given time
and a given position of the cell. The permanent morphogenetic activi-
ties of the polyp (Brien, 1953), resulting in a morphological steady-
state, in which cell-replacement and proliferation have to compensate
for continuous losses of cells at both extremities of the animal, are
polarizing the distribution of these 2 substances, one of which is
said to be more diffusable than the other. According to the axial dis-
tribution of these substances, sex determination would be polarized in
6
the sense of the dis to-proximal axis of the polyp. In fact, in hermaph-
rodi tes, such as H. viridis L., the testes appear in the sub tentacular
region of the body column above the so-called female region, which more
or less coincides with the budding region. This model considers sex-
determination as being an integrated part of the somatic differentia-
tion pattern. In my view there is no reason to dissociate both events
(somatic and gametic differentiation) from one another and to postulate
separate mechanisms of determination, I believe that the model, as pro-
posed by Burnett (1967), can in its simplest form not be applied to
sex-determination - as will be shown by the following observations and
experiments.
Hydra attenuata Pall. is fundamentally a gonochoristic species (Brien

and Renier-Decoen, 1951; Tardent, 1966a). When culturing isolated indi-
viduals for many months I observed (Tardent, 1966a) that most of them
would remain faithful to their functional state of sexual differentia-
tion throughout many subsequent sexual periods. This stability expres-
ses itself also in the fact that asexually produced buds and their
offspring show, when reaching sexual maturity, the same sex as that of
the polyp from which they originated (Fig. 4 A, B).
Fig. 4 A-C. Hydra atte-

Days 225Kd 230 9 9
o nuata Pall. (A and B)
A B C Stability of the state
2 of sexual determina-
4 tion in asexual off-
6 spring of a male (A)
r-- polyp and a female (B)
8
polyp (the black ver-
10 - tical bars indicate
12 the duration of a pe-
14
-- riod of spermatogene-
sis and the black
16 circles represent Lhe
18 eggs produced by the
20 female polyps). (C)
Spontaneous sex in-
22
version as it occurred
24
..
in a single asexual
26 bud produced by a male
polyp. (The numbers of
28 I
•
I
the clones from which
30
32
d d (5
I I
(5 d d
.~ the polyps were iso-
lated are given on top
9 9 9 9 9 9 of each graph). (From
~ '------v-----"'
Buds Buds Tardent, 1966b)
For unknown reasons, however, such relatively stable individuals would

suddenly undergo an inversion of their sex (Fig. 4 C). These spontane-
ous inversions which occurred in both directions, male to female and
vice-versa, were mostly associated with a transitional functional
hermaphroditism lasting for one sexual period only (Tardent, 1966a).
In these cases oocytes and spermatocytes would appear side by side
(Fig. 2c).
Already Goetsch (1922), Wiese (1953), Brien (1962, 1963, 1966) and
others, had shown that in balanced and unbalanced gonochoristic species
of Hydra a similar inversion of the sexual status can be brought about
experimentally by hetero-sexual parabiosis. All these experiments agreed
7
in the finding that when 2 complementary fragments of the same species

(but representing both sexes) are grafted together, the female partner
exposed to the male portion will always be masculinized. The parabio-
tically induced sex-inversion is, in this case, invariably a one-way
affair.
Fig. 5 A and B. Masculinization of female halves by complementary male

halves, following parabiosis in Hydra attenuata (Pall.). (A) Grafting
procedure by which heterosexual chimaera were made in a reciprocal
axial arrangement. (B) Results: The sexual status of entire, non-
grafted male and female controls and of their asexually produced buds
(upper 2 rows); the sexual status of heterosexual chimaera and that
of their buds (lower 2 rows). (From Tardent, 1968)
As shown in Fig. 5, the masculinizing effect is not polarized (as sug-

gested by Burnett, 1967), since proximal male portions representing
a part of the budding region and the stalk act, as well as distal frag-
ments including the hypostome, which is considered as being a morpho-
genetic organizer. Masculinization of the female partner is also ob-
tained, when relatively small male fragments such as the hypos tome
8
alone or the stalk of the polyp, are grafted onto complementary female
partners (Fig. 6). Here again, proximal and distal fragments act the
same way. In practically all cases the sex-inversion, induced by graft-
ing a male fragment onto a female, remains stable (Fig. 7). Following
transplantation, the heterosexual chimaera remain males throughout
many subsequent gametogenetic periods. The same is true for the asexu-
ally produced offspring which are males too, even if the larger portion
of the heterosexual chimaera originally has been female (Fig. 7).
3
? 0
rf· 0
~. 29
? 1
9
?- 0
Fig. 6 A and B. Masculinization of large female portions by relatively

small male fragments following parabiosis (Hydra attenuata Pall.).
(A) Grafting procedure by which heterosexual chimaera were made from
unequal but complementary male and female fragments. (B) Results: The
sexual status of entire, non-grafted male and female animals and of
their buds (upper 2 rows); the sexual status of the heterosexual chi-
maera and that of their buds (lower 2 rows). (From Tardent, 1968)
These results suggest, that this trans-determination of the female

part must be brought about by "something" moving in from the adjacent
male fragment. This "something" may be cells (potential male gonocytes)
or a hormone-like masculinizing substance. In order to obtain full
success, i.e. a masculinization of all female fragments, parabiosis
between complementary male and female components must last for at least
67 hours. This was revealed by an experiment (Tardent, 1968), in which
the duration of heterosexual parabiosis was limited, the grafts being
separated again at various intervals after parabiosis had been estab-
lished by grafting. This finding can, of course, not decide whether
masculinization is brought about by a diffusing substance or by migrat-
ing cells. From other experiments, in which radioactive labeling was
combined with the grafting technique (Tardent and Morgenthaler, 1966)
9
Fig. 7. The postopera-

tional sexual behavior
of 2 heterosexual chi-
222 228 maeras (cf. Fig. 6) and
their buds, demonstrat-
ing the stability of the
sex inversion induced
d by parabiosis (Hy d ra atte-
Buds Buds
nua t a Pall.) (hatched
1. 2. 3. 4. bars = periods of sper-
I
I
I ! matogenesis; black cir-
II !
I
cles = eggs). (From
Tardent, 1968)
I I
I
I
I
I
I
I
d d d cf
we know that interstitial cells and nematocytes move rapidly (a matter

of few hours) from one fragment to another.
A rather rude method for eliminating interstitial cells consists of

exposing a Hydra (or fragments of it) to sublethal dosis of X-rays
(Zawarzin, 1929; Evlakhova, 1946; Brien and van de Eeckhoudt, 1953)
or mustard gas (Figi, 1969). As a consequence of this treatment and
due to the absence of I-cells, which seem to be particularly sensitive
to X-rays (Fig. 2d), no more nerve cells and no nematoblasts are pro-
duced and the animal is doomed - unless it is put in parabiosis with
a non-treated healthy animal from which it received new cells (Tardent
and Morgenthaler, 1966).
Therefore, it was of interest to know how male fragments, deprived of

their interstitial cells by X-ray treatment (Fig. 2d), would act in
parabiosis with complementary female portions. The results of these
experiments, in which the male halves had been irradiated with 6500 r
four days before being grafted to a complementary untreated female
fragment, are given in Fig. 8. They show that the irradiated male com-
ponents fail to exert their masculinizing effect as compared with the
non-irradiated controls. Again, these results cannot be considered as
being conclusive with respect to the above mentioned alternative, cells
or substance. There are certainly no I-cells moving from the irradiated
male fragment into the opposite female partner, but we must also take
into consideration the fact, that no I-cells means no more nerve-cell-
production and consequently, no more or at least considerably less,
neurosecretory products ("substance").
When exposing whole males of Hydra attenuata Pall. to sublethal doses of

X-rays (1200 and 2400 r) that fail to knock out the entire I-cell popu-
lation, but reduces it considerably, a relatively high percentage of
such treated males (13 and 71%) become females (Fig. 9) in the sexual
10
periods following irradiation (Tardent, 1968). The same is true when

whole male polyps are treated with mustard gas (0.005 and 0.006%). In
this case, up to 81% of the males become females following treatment
(Figi, 1969). On the other hand, various concentrations of the cyto-
staticum Colcemid failed to exert a comparable effect (Figi, 1969).
Fig. 8. The sexual status of

heterosexual chimaeras com-
posed of complementary halves
(cf. Fig. SA) in which the
male partner had been irra-
diated (6500 r, arrows)
4 days before transplanta-
tion. The controls consist
of chimaeras, in which (as
in Fig. 5B) the male portion
had not been irradiated
(Hydra attenuata Pall.). (From
Tardent, 1968)
27
Fig. 9. The sexual sta-

tus of whole male polyps
of Hydra attel1uata Pall.
following treatment with
sublethal doses of X-
? rays. The question marks
indicate those animals,
which failed to produce
gonads after treatment
and therefore could not
be identified with re-
spect to their status of
sex determination
o. 1200. 2400, 3600,
In taking into consideration all these facts one is inclined to think

that the population density of the I-cells, i.e. the potential gono-
cytes, somehow playa role in the process of sex determination. A low
density level seems to characterize the female, a high density level
the male state of determination. All results presented here could be
interpreted in this way. However, I think that population density as
such cannot be attributed a key role in the determinating events.
I rather think that the primary information deciding whether a gono-
11
cyte is to become an oocyte or spermatocyte, is dependant on the con-

centration of a neurosecretory substance, in the sense that a high
concentration would mean spermatogenesis, while a relatively low con-
centration would favor oogenesis. This substance mayor may not be
identical with the inducing substance known to promote morphogenetic
events (Schaller, 1973; Schaller and Gierer, 1973). As this substance
can now be isolated in fair amounts it will be of interest to investi-
gate its effect on sexual differentiation.
In the light of this working hypothesis, the causal connection between

the number of interstitial cells on one hand and the concentration of
this still hypothetical neurosecretory product on the other, could be
seen as follows: As nerve cells, which are supposed to produce this
substance, are known to derive by differentiation from interstitial
cells (Burnett and Diehl, 1964), a dense population of this cell type
would permit a larger amount of nerve-cells to be produced, and more
neurons are expected to produce larger amounts of neurosecretory pro-
ducts. As proposed before, this situation would characterize the male
state of determination, while fewer I-cells, i.e. fewer nerve cells
and accordingly, smaller amounts of substance, would create a milieu
favoring oogenesis.
This view, based on the quantitative action of one substance only, is

nothing more but a working hypothesis which needs to be thoroughly
checked by quantitative histological and biochemical procedures. I am
inclined to think that in coelenterates sex-determination is, as in
other lower invertebrates, also subjected to the control of neurogenic
factors. The difficulty in uncovering the identity of these factors
resides in the fact that the nervous system of these primitive animals
is nothing more than a network of loosely interconnected neurons
(Bullock and Horridge, 1965), and that neurosecretion as such is diffi-
cult to study, particularly when quantitative data are needed.
References
Berrill, N.J.: Growth, Development and Pattern. San Francisco-London:

W.H. Freeman and Company 1961.
Brien, P.: La perennite somatique. Biol. Rev. 28, 308-349 (1953).
Brien, P.: Contribution a l'etude de la biologie sexuelle: Induction
gametique et sexuel+e chez les Hydres d'eau douce par des greffes
en parabiose. Bull. Acad. Roy. Belg. Cl. Sci. 48, 825-847 (1962).
Brien, P.: Contribution a l'etude de la biologie sexuelle chez les
Hydres d'eau douce. Induction gametique et sexuelle par la methode
des greffes en parabiose. Bull. Biol. France Belg. 97, 213-283
(1963) .
Brien, P.: Biologie de la reproduction animale. Blastogenese, gameto-
genese, sexualisation. Paris: Masson and Cie. 1966.
Brien, P., Reniers-Decoen, M.: La gametogenese et l'intersexualite
chez HydPaattenuata (Pall.). Ann. Soc. Roy. Zool. Belg. 82, 285-327
(1951) .
Brien, P., van den Eeckhoudt, J.P.: Bourgeonnement et regeneration chez
les Hydres irradiees par les rayons x. C. r. Acad. Sci. (Paris)
237, 756-758 (1953).
Bullock, T.H., Horrid~e, G.A.: Structure and Function in the Nervous
Systems of Invertebrates. San Francisco-London: W.H. Freeman and
Company 1965.
Burnett, A.L.: The maintenance of form in Hydra. S¥IDP. Soc. Study
IDevelop. Growth ~, 27-52 (1962).
12
Burnett, A.L.: Control of polarity and cell differentiation through

autoinhibition - a model. In: EXp. Biol. Med. 1, 125-140. Basel-
New York: S. Karger 1967.
Burnett, A.L., Davis, L.E., Ruffing, T.E.: A histological and ultra-
structural study of germinal differentiation of interstitial cells
arising from gland cells in Hydra viridis. J. Morph. 120, 1-8 (1966).
Burnett, A.L., Diehl, N.A.: The nervous system in Hydra.---I. Types,
distribution and origin of nerve cells. J. expo Zool. 157, 217-226
(1964). -
Campbell, R.D.: Cnidaria. In: Reproduction of marine Invertebrates
(eds. A.C. Giese and J.S. Pearce) Vol. 1, 133-199. New York-London:
Academic Press Inc. 1974.
Evlakhova, V.F.: Form-building migration and regenerative material in
Hydra. C.R. (Doklady) Acad. Sci. URSS 53, 369 (1946).
Figi, H.: Die Wirkung von Colcemid und Dichloren auf das sexuelle und
vegetative Verhal ten von Hydra attenuata Pall. Thesis, Department of
Zoology, University of Zurich, Switzerland (1969).
Frey, J.: Die Entwicklungsleistungen der Medusenknospen und Medusen
von Podoeoryne earnea M. Sars nach Isolation und Dissoziation.
Wilhelm Roux' Archiv 160, 428-464 (1967).
Glatzer, K.H.: Die Ei- und Embryonalentwicklung von Corydendriwn paras i-
tiewn mit besonderer Berlicksichtigung der Oocyten-Feinstruktur
wahrend der Vitellogenese. Helgolander wiss. Meeresunters. 22,
213-280 (1971). --
Goetsch, W.: Hermaphroditismus und Gonochorismus bei Hydrozoen. Zool.
Anz. 21, 6-18 (1922).
Hyman, L.H.: The Invertebrates: Protozoa through Ctenophora. Vol. I.
New York-London: Mc Graw-Hill Book Company, Inc. 1940.
Korn, H.: Zur ontogenetischen Differenzierung der Coelenteratengewebe
(Polyp-Stadium) unter besonderer Berlicksichtigung des Nervensystems.
Z. Morphol. Oekol. Tiere 57, 1-118 (1966).
Kuhn, A.: Entwicklungsgeschichte und Verwandtschaftsbeziehungen der
Hydrozoen. 1. Teil: Die Hydroiden. Ergebn. u. Fortschr. der Zoolo-
gie 4,1-284 (1914).
Lentz, T.L.: The cell biology of Hydra. Amsterdam: North-Holland Pul:Jl.
Co. 1966.
Mergner, H.: Cnidaria. In: Experimental Embryology of Marine and Fresh
Water Invertebrates (ed. G. Reverberi), pp. 1-84. Amsterdam-London:
North-Holland Publ. Co. 1971.
Nyholm, K.G.: Zur Entwicklung und Entwicklungsbiologie der Ceriantha-
rien und Aktinien. Zool. Bidr. Uppsala 22, 85-248 (1943).
Nyholm, K.G.: On the development and dispersal of Athenaria aetinia with
special reference to HaZeampa duodeeimeirrata M. Sars. Zool. Bidr.
Uppsala 27, 467-505 (1949).
Russell, F.S.: The Medusae of the British Isles. Anthomedusae, Lepto-
medusae, Lirnnomedusae, Trachymedusae and Narcomedusae, pp. 1-530.
Cambridge University Press 1953.
Schaller, H.C.: Isolation and characterization of a low-molecular-
weight substance activating head and bud formation in Hydra. J.
Embryol. expo Morph. 29, 27-38 (1973).
Schaller, H.C., Gierer, ~: Distribution of the head-activating sub-
stance in Hydra and its localization in membranous particles in
nerve cells. J. Embryol. expo Morph. 29, 39-52 (1973).
Siewing, R.: Lehrbuch der vergleichenden~ntwicklungsgeschichte der
Tiere. Hamburg-Berlin: Paul Parey Verlag 1969.
Slautterback, D.B., Fawcett, D.W.: The development of the cnidoblast
of Hydra. An electronrnicroscope study of cell differentiation.
J. biophys. biochem. Cytol. 5, 441-452 (1959).
Tardent, P.: Axiale Verteilungsgradienten der interstitiellen Zellen
bei Hydra und TUbuZaria und ihre Bedeutung flir die Regeneration.
Wilhelm Roux' Archiv Li£, 593-639 (1954).
13
Tardent, P.: Regeneration in the Hydrozoa. Biol. Rev. 38, 293-333

(1963) .
Tardent, P.: Zur Sexualbiologie von Hydra attenuata (Pall.). Rev. suisse
Zool. 73, 357-381 (1966a).
Tardent, ~: Experimente zur Frage der Geschlechtsbestimmung bei Hydra
attenuata (Pall.). Rev. suisse Zool. 73, 481-492 (1966b).
Tardent, P.: Experiments about sex determination, in Hydra attenuata Pall.
Develop. Biol. 17, 483-511 (1968).
Tardent, P., Eymann; H.: Experimentelle untersuchungen uber den rege-
nerationshemmenden Faktor von Tuhularia. Wilhelm Roux' Archiv 151,
1-37 (1959). -
Tardent, P., Morgenthaler, U.: Autoradiographische untersuchungen zum
Problem der Zellwanderungen bei Hydra attenuata (Pall.). Rev. suisse
Zool. 73, 468-480 (1966).
Weiler-Stolt, B.: tiber die Bedeutung der interstitiellen Zellen fur
die Entwicklung und Fortpflanzung mariner Hydroiden. Wilhelm Roux'
Archiv 152, 398-455 (1960).
Weismann, A.: Die Entstehung der Sexualzellen bei den Hydromedusen.
Jena: Gustav Fischer 1883.
Werner, B.: Die Verbreitung und das jahreszeitliche Auftreten der
Anthomeduse Rathkea octopunctata M. Sars, sowie Temperaturabhangig-
keit ihrer Entwicklung und Fortpflanzung. Helgolander wiss. Meeres-
untersuch. 6, 137-170 (1958).
Widersten, B.:-On the morphology and development in some cnidarian
larvae. Zool. Bidr. Uppsala 37, 139-182 (1968).
Wiese, L.: Geschlechtsverhaltnisse und Geschlechtsbestimmung bei SuB-
wasserhydroiden. Zool. Jahrb. Abt. Zool. Qi, 55-83 (1953).
Zawarzin, A.A.: Rontgenologische Untersuchungen an Hydren. I. Die
Wirkung der Rontgenstrahlen auf die Vermehrung und Regeneration von
Pelmatohydraoligactis. Wilhelm Roux' Archiv. 115, 1-26 (1929).
Zihler, J.: Zur Gametogenese und Befruchtungsbiologie von Hydra.
Wilhelm Roux' Archiv l£2, 239-267 (1972).
Sexual Differentiation in the Fissiparous Strain of
Dugesia gonocephala
Th. Lender and C. BrianQon
A. Introduction
Most of the freshwater planarians are simultaneous hermaphrodites.

Some species show an alternation of sexual reproduction and agamic
reproduction. The first generally takes place in winter, agamic repro-
duction in summer. Fissions may cease during sexual activity. For some
species, especially Dugesia gonocephala and D. tigrina, two strains can
be distinguished, one in which both sexual and fissiparous individuals
alternate regularly and an exclusively fissiparous strain.
In fissiparous planarians sexual differentiation can be induced by

various methods: either by grafts as in D. tigrina (Kenk, 1941) and
D. gonocephala (japonica) (Okugawa, 1957) or by feeding of D. gonocephala
with sexually mature planarian tissues (Grasso and Benazzi, 1973).
The sexual state, induced by grafting, arises from neoblasts of the

sexual type which are in the graft and which are able to differentiate
in the host. The graft possibly contains some substances able to act
on the host regeneration cells in order to influence them to differ-
entiate in the sexual cells. This is Okugawa's hypothesis and the
experiments of Grasso and Benazzi support this hypothesis.
In grafting experiments, Brian90n studied the effects of various tis-

sues upon sexual differentiation in fissiparous strains of D. gonocephala
(Lender and Brian90n, 1974).
B. Materials and Methods
For the last 15 years D. gonocephala, originating from Northern France,

Alsace and Paris, has been cultured in the laboratory. Some animals
have a gonopore and produce cocoons all year long. These animals are
sporadically fissiparous. Other animals without gonopores are fissipa-
rous. The fission is cyclical, temperature and illumination are factors
which condition these fluctuations. Between Hay and September a few
specimens (5 - 10%) do occasionally become sexual. The laid cocoons al-
ways turned out to be completely sterile.
Testes, ovaries, brain and parenchyma were grafted into planarians

without gonopores. Histological preparations were made after the pla-
narians were fixed in Bouin solution, then cut in sagittal sections
and stained with hemalum picro-indigo carmine.
15
c. Results
I. Anatomy of the Genital Apparatus of Sexual Planarians
The specimens have a gonopore and an entire genital apparatus. There

are three areas: (1) ventrally, and behind the brain, there are the
two ovaries. They are connected with the common antrum by two oviducts
which also collect the yolk from the yolk glands; (2) numerous testes,
which are situated dorsally or ventrally according to species, form
small gonads - from the ovary to the copulatory apparatus, and they are
connected with the copulatory apparatus by two ciliated spermiducts;
(3) the copulatory apparatus and the gonopore are postpharyngeal. The
reproductive system appears about the third month after hatching, in
the following order: ovary, testes and copulatory apparatus - being
Fissiparous animals without gonopores are not asexual. They often have
large ovaries, but oogenesis ceases before meiosis. Testes and copula-
tory apparatus are never visible.
II. Sexual Differentiation by Grafts (Table 1)
1. Grafts of Testes
Testes grafts produced a gonopore in three quarters of the hosts after

approximately 11 weeks. Some planarians were cut across behind the
grafts. In these cases, the gonopore only appeared in fragments with
the graft. Histological investigation showed the grafted testes. In
the host, the ovary became mature (Fig. 1). Near the graft, small and
young testes without spermatozoons were visible (Fig. 2). The copula-
tory apparatus was well developed and normal (Fig. 3). In this experi-
ment the rate of fission was low. Fission occurred in 4 of the 37 pla-
narians in which a gonopore has appeared.
Table
Graft material No. of grafts No. of gonopores
Testes 46 37
Ovary of sexual strain 28 19
Ovary of fissiparous strain 18 0
Brain 31 0
Parenchyma of sexual strain 25 0
Parenchyma of asexual strain 16 0
2. Ovary Grafts
Mature ovaries taken from sexually mature planarians were grafted on

planarians without gonopores. A gonopore appeared in two thirs of the
hosts. However, it took approximately 2 weeks longer than the testicu-
lar grafts for the gonopore to appear. Histological investigation
showed one normal copulatory apparatus, with the testes maturing in
the graft (Fig. 4), and young immature testes in the host tissues near
the graft (Fig. 5). The grafted ovaries remained and the host ovaries
were well developed and beginning their differentiation. When the pla-
narian was cut across, the fragment without a graft never differen-
tiated sexually.
16
Figs. 1 - 3
17
In these experiments, the fission rate is higher than in testes-graft

experiments. Fission occurred in 10 of the 19 planarians in which a
gonopore appeared. But fission occurred during the first weeks of the
experiment, at least 6 weeks before the gonopore appeared.
With immature ovary grafts from fissiparous planarians, the gonopore

never appeared.
3 .. Non-sexual Tissue Graft
Brain or caudal parenchyma from mature planarians was grafted on pla-

narians without gonopores. Sexual differentiation was never induced.
The ovaries of the host did not develop and there were no testes or a
copulatory apparatus.
The rate of fission was normal with parenchymal grafts. The rate of
fission was high with brain grafts (81%). This result was normal as
neurosecretion controls. asexual reproduction (Lender, 1973).
C. Discussion
The graft experiments on D. gonocephala showed that sexual differentia-

tion could be induced in fissiparous planarians. However, sexual tis-
sue was necessary for induction as has been shown by Kenk, Okugawa,
Grasso and Benazzi. The host ovaries began to develop, testes and
copulatory apparatus also differentiating. However, the different
parts of the genital apparatus only developed in the area where they
normally could be found, especially the copulatory apparatus which is
always postpharyngeal whatever the position of the graft may be.
There are three areas from the head to the tail: a female territory
with ovaries, a male territory with testes, and a territory with a
copulatory apparatus. The female territory exists even in fissiparous
planarians.
Studies are proceeding to observe whether the other territories can

exist in fissiparous planarians but which are unable to differentiate
corresponding sexual organs. Appearance of testes in the ovary graft
can be explained if a part from the testicular area has been trans-
planted with the ovaries.
Sexual differentiation was due to the testes and not to a transfer of

neoblasts coming from the graft. This is Okugawa's hypothesis and in
Grasso's and Benazzi's experiments, this migration does not occur.
Testes produce the ovarian maturation of fissiparous planarians, the
~ Fig. 1. Ovary of a fissiparous planarian, grown after grafting of

testes. DC digestive cells; Fd dorsal face; OC oocytes
Fig. 2. Testes developed in the parenchyma of a fissiparous planarian
implanted with testes. Fd dorsal face; SPI spermatocyte I; SPII sper-
matocyte II; SPG spermatogonium; SPT spermatid
Fig. 3. Copulatory apparatus developed in the postpharyngeal region of
a fissiparous planarian grafted with testes. Ag common antrum; Ch cp
copulatory bursa; Cn bursal canal; Fd dorsal face; Fv ventral face;
Pg common gonopore; Ph pharynx; Pn penis; Pp mouth; Vs spermiducal
vesicle
18
differentiation of testes and the differentiation of the copulatory

apparatus in the host. When obtaining sexual differentiation with
ovary grafts, regenerate testes are always differentiated and the
appearance of the gonopore is delayed. The differentiation of the co-
pulatory apparatus is directly conditioned by the testes. Vandel (1920)
suspected this in PolyceUs cornuta and this was proven in Dugesia lugu-
bris by Fedecka-Bruner (1968).
Fig. 4. Testes developed in a graft of ovaries implanted on a fissi-

parous planarian. DC digestive cells; Fd dorsal face; Ov ovary im-
planted; T testes developed in the graft
Fig. 5. Young testes developed in the parenchyma of a fissiparous
planarian implanted with ovaries. Fd dorsal face; M mitosis in sper-
matogonium; SPG spermatogonium
19
These experiments showed that there is also a antagonism between sexual

maturity and fission. The fissiparous planarians with a testes graft
have a low rate of fission (10%). After ovary graft of sexually mature
planarians, fission is higher (50%). However, the fissions occurred
during the first weeks of the experiment, when the testes and the co-
pulatory apparatus were missing. With a parenchyma graft, the rate of
fission was normal, whereas a brain graft promoted fission (81%). This
is new evidence of the action of the brain on fission.
Benazzi (1973) claimed that the opposition between sexual maturation

and fission was due to a genetical difference. In fissiparous plana-
rians some genes would be repressed, which would jnhibit the sexual
state. Grasso and Benazzi claimed, that the derepression was due to
neurosecretion in the crushed tissues. Our brain graft experiments
and Okugawa's experiments did not prove this hypothesis. The induction
of sexual maturation seems due to testes. Their activity acts at a
distance, so that we can presume secretion of a substance. However,
we know nothing about its nature or its origin. The mechanism which
promotes the appearance of the first testes is also not clear.
References
Benazzi, M.: Fissioning in planarians from a genetic standpoint. In:

Biology of the Turbellaria, Libbie H. Hyman Memorial Volume. New
York: McGraw-Hill Book Company 1973.
Fedecka-Bruner, B.: ~tudes sur la regeneration des organes genitaux
chez la Planaire Dugesia Zugubris. III. Regene-ration de l' appareil
copulateur. Bull. Biol. Fr. Belg. 102, 3-44 (1968).
Grasso, M., Benazzi, M.: Genetic and physiologic control of fissioning
and sexuality in planarians. J. Embryol. expo Morph. 30, 317-328
(1973) . -
Kenk, R.: Induction of sexuality in the asexual form of Dugesia tigrina.
J. expo Zool. ~, 55-69 (1941).
Lender, Th.: The r.ole of neurosecretion in freshwater planarians. In:
Biology of the rurbellaria, Libbie H. Hyman Memorial Volume, pp.
460-475. New York: McGraw-Hill Book Company 1973.
Lender, Th., Brian90n, C.: Induction de la maturation sexuelle chez
la Planaire scissipare Dugesia gonocephaZa (Turbellarie, Triclade).
J. Embryol. expo Morph. 32, 159-168 (1974).
Okugawa, K.J.: An experimental study of sexual induction in the asexual
form of Japanese fresh-water planarian Dugesia gonocephaZa (Duges).
Bull. Kyoto GakugeiUniv. 11, 8-27 (1957).
Vandel, A.: Le developpement de l'appareil copulateur des Planaires
est sous la dependance des glandes genitales. C.R. Acad. Sci. Fr.
170, 249-251 (1920).
Sexuality and Neurosecretion in Freshwater Planarians*
M. Grasso
In planarians, the induced development of the hermaphroditic genital

apparatus, by grafting the anterior body third of a sexual planarian
into the two posterior body thirds of an asexual planarian, has been
demonstrated by Kenk (1941) and Okugawa (1957). The investigations of
Lender (1952) have shown the morphogenetic influence of the cerebral
plexus for the regeneration of ocular spots in planarians.
In my experiments (Grasso, 1959, 1963), I have demonstrated that in

Dugesia lugubris the ablation of the sole cephalic region (inclusive of
the cerebral plexus) by means of a cross-cut in front of the ovaries,
had no effect upon the hermaphroditic genital apparatus, which was
preserved in decapitated planarians during the short time necessary
for complete regeneration of the head. In fact, the ablation of both
the cerebral plexus and the anterior tract of the posterior ventral
nerve cords (carried out by a cut behind the level of the ovaries) ,
results in complete disappearance of the testes on the posterior body
piece. This disappearance lasted until the removed area, with cerebral
plexus and ovaries, had attained complete regeneration.
After a similar experiment, Brandi and Ghirardelli (1963) observed

that the posterior piece remains permanently deprived of genital appa-
ratus if cephalic regeneration is inhibited by an amphetamine treat-
ment.
In PolyceUs nigra, decapitation results in the disappearance of testes

in the decapitated fragment (Grasso, 1964). In this case, however, for
the cross-cut to be effective it must be performed at a more caudal
level behind the whole row of marginal eyes which, in this species,
border the head so that together with the brain, a much longer tract
of posterior ventral nerve cords is removed. Moreover, in P. nigra,
a complete regeneration of the removed area also results.
The influence of the cerebral plexus on the development and preserva-

tion of the hermaphroditic genital apparatus seems to receive further
confirmation from experiments on planarians made bicephalous through
proper operations. In D. Zugubris I observed (Grasso, 1963) that in each
of the two heads, the development of two ovaries and numerous testes
took place. The ablation of one head alone, behind the ovaries, re-
sulted merely in the regression of the testes which were closer to the
wound. Removal of both heads resulted instead, in the disappearance
of all testes (Fig. 1).
These findings have been confirmed by Fedecka-Bruner (1967a,b; 1968).

Fedecka-Bruner also stated that the regression of the genital appara-
tus in D. lugubris is but partial, if the cerebral plexus alone is re-
moved; non-existent if only one of the two posterior ventral cords is
-This research was financed by the "Consiglio Nazionale delle Ricerche"

(Italy)
21
removed and total if both such cords are cut off leaving the brain
untouched. The posterior ventral nerve cords, to whose initial tract
the ovaries are attached seem, therefore, to be even more important
than the cerebral plexus for the development and preservation of the
hermaphroditic genital apparatus.
A B c o
Fig. 1 A-D. Dugesia lugviJris. Big spots represent the ovaries, small
spots the testes. (A) T-shaped cut for obtaining bicephalous condition.
(B) bicephalous specimen. (C) unilateral decapitation behind the ova-
ries determines regression of a few testes near the wound. (D) bilat-
eral decapitation determines regression of all the testes
All the results obtained by Grasso (1963) through the above described
operations have been fully confirmed by Teshirogi and Fujiwara (1970),
who reached the same interpretation, carried out by independent re-
search on a Japanese planarian, Bdellocephala brunnea (Teshirogi, 1963).
In planarians, the regeneration of all organic apparatus (including

the genital one), comes from particular pluripotential undifferentiated
cells called "neoblasts"" which are capable of active migration into
the parenchyma - to differentiate in the most varied ways according
to the body level in which they at last are placed (Wolff and Dubois,
1948; Dubois, 1949; Stephan-Dubois and Gusse, 1970; Sengel, 1960;
Lender and Gabriel, 1961, 1965; Cecere et al., 1964). Thus, the exis-
tence of some substances responsible for the different inducing actions
which regulate the various sorts of cellular differentiation may be
inferred.
With the purpose of testing the hypothesis of a nervous origin of at

least a part of such substances responsible for the genital apparatus
differentiation, I started a series of investigations in order to as-
certain what is the exact relationship between the nervous system and
the gonads in planarians. By using the silver impregnation method ac-
cording to Cajal in D. lugviJris and P. nigra (Grasso, 1965a, 1966a),
I was able to demonstrate that ovaries and testes are surrounded by
a network of nerve cells and nerve fibers. These histological data
strongly suggest a functional interdependence between gonads and ner-
vous system, so explaining the reasons by which the differentiation
22
and maturation of the hermaphroditic genital apparatus in planarians

requires careful preservation of the morphological integrity of the
central nervous system. The hypothesis of an existing neurohormone is
becoming more and more pressing.
Lender and Klein (1961, 1962) were the first to demonstrate the pres-
ence of neurosecretion in planarians, - in the cerebral plexus of
P. nigra and Dendrocoelum lacteum. Subsequently, Vendrix (1963) and Ude
(1964) described neurosecretory cells not only in the brain, but also
along the posterior ventral nerve cords of P. nigra, Dugesia gonocephala
and Dendrocoelum lactev.m. Lender (1964) likewise, reported the existence
of neurosecretory cells in P. nigra, Dendrocoelum lacteum, Dugesia lugubris,
Dugesia tigrina and Dugesia gonocephala and besides admi tting an intervention
of neurosecretion in the regeneration processes of the posterior parts
of these amimals' bodies, was the first to suggest that neurosecretion,
by acting as a "gonado-stimuline", may help the sexual maturation pro-
cesses.
By means of the usual histological methods for the study of neurose-

cretion (Gomori-Bargmann's chrome-hematoxylin-phloxine, Adams-Sloper's
Alcian blue modified by Herlant, and especially Gabe's paraldehyde-
fuchsin), I could also verify the presence of large amounts of neuro-
secretory cells all along the nervous system, not only in sexually
mature specimens of Dugesia lugubris (Grasso, 1965b) and P. nigra (Grasso,
1966a), but also in specimens of an agamic strain of D. tigrina with
no genital apparatus (Grasso, 1965c). Moreover, in the two above-
mentioned sexual species I was able to observe that the networks of
nerve cells surrounding the ovaries and testes stained selectively
with the dyes used for neurosecretion. Thus I could definitely suggest
the hypothesis that this is the place of production and transport of
substances active on the germ cells, functioning as "gonado-stimulines".
It is, therefore, not surprising that the existence of neurosecretory

cells occur even in the agamic strains of planarians. It is by now
widely demonstrated (Bondi and Pascolini, 1966; Liotti et ai, 1966;
Kaplonska, 1967; Liotti and Rosi, 1968; Lender and Zghal, 1969; Sauzin-
Monnot, 1972) that in planarians not only the genital apparatus matu-
ration, but also very many other events are regulated by neurosecre-
tory processes - in particular, the regeneration phenomena; those of
agamic reproduction by fission; and, according to Ude (1964), also
those of hydro-saline balance regulation, which are essential for the
survival of freshwater animals.
The importance of neurosecretion for the regeneration of some species

of planarians at both adult and embryonic stages has been demonstrated
by Grasso (1965d, 1966b,c). It seems safe to suppose that in these
organisms the neurosecretory cells do not produce a single neurosecre-
tion, but many kinds of neurosecretions. Some of these (exclusive to
sexual strains) are endowed with the already mentioned "gonado-stimu-
line" activity, others instead (common to both sexual and agamic
strains) are endowed with other activities among which that of "neo-
blasto-stimuline"-important in regeneration and agamic processes.
A "gonado-stimuline" action could be attributed to the neurosecretions

elaborated by the above described nerve networks, which surround the
ovaries and testes in sexual planarians. The existence of such a peri-
ovarian and peritesticular neurosecretion has been fully confirmed
by electron microscope observations (Grasso and Quaglia, 1970a,b;
1971). These findings have a much higher specificity than those, pre-
viously named, obtained by histological staining. These ultrastruc-
tural investigations have demonstrated the existence, around the go-
23
nads, of networks of nerve cells whose fibers, throughout the matura-

tion of gametes, are rich in electron-dense neurosecretion granules
surrounded by a membrane. These granules are comparable in structure:e
and size (500 - 2000 A) to the neurosecretory granulations already
described by Morita and Best (1965), and later on confirmed by other
authors in the central nervous system of various species of planarians.
It has been observed that in periovarian and peri testicular nerve net-
works there are many neurosecretory granules present while both male
and female gametes are maturing which are scarce when the maturation
process is completed. Attention has been drawn to the role of neuro-
secretory granules in the cisternae of Golgi's apparatus and in the
perikaryon of nerve cells. Here, the granules seem to migrate along
the axons adjacent to the gonads, finally discharging into the inter-
cellular spaces and into the interstices among the neurons. This occurs
through a mechanism of "exocytosis" similar to that described by Bunt
(1969) for the neurosecretory systems of Crustaceans. The neurosecre-
tion thus liberated would therefore enter into direct contact with the
germ cells, inducing gametogenesis in them (Fig. 2).
Fig. 2 A and B. Electron microscope aspect of nerve fibers near gonads.

Note the neurosecretion granules surrounded by a membrane. (A) x 40,000;
(B) x 50,000
On the other hand, it is well known that among the Tricladid Paludicola
of the family Planariidae, there are species in which reproduction
occurs exclusively by gametes. Others exist in which there are periods
of sexual reproduction and periods of agamic reproduction by fission,
according to seasonal cycles. In many other species there is even a
coexistence of sexual and agamic strains (de Beauchamp, 1961). Agamic
reproduction usually occurs by a process called "architomy" , which
consists of a simple cross-division, almost always at a retropharyngeal
level, with no preliminary regenerative phenomena (Vandel, 1921, 1922;
Kenk, 1937; Benazzi, 1938; Stagni and Grasso, 1965).
24
According to Benazzi (1938) the agamic reproduction of fissiparous

strains, although influenced in its rhythm by environmental conditions,
depends essentially on genetic factors. These would not only induce
reproduction by fission, but one of their earliest effects would be
that of inhibiting the development of the whole hermaphroditic genital
apparatus.
To this genetic aspect of agamic reproduction in planarians can now

be added one of a physiological and perhaps also, of an endocrinolo-
gical nature. Recent investigations (Grasso, 1971, 1972; Grasso and
Benazzi, 1973) have in fact demonstrated that specimens of agamic
fissiparous strains of the species Dugesia gonocephaZa and Dugesia tigrina
attain, in a very high percentage of cases, the sexual state, if fed
with a mixture obtained by crushing sexually mature individuals of the
species PoZyceZis nigra - a planar ian which reproduces exclusively by
gam~tes.
After nearly two months of this treatment it could be ascertained that

almost all the animals stopped reproducing by fission. For this reason
they then grew in size, very soon reaching conspicuous dimensions. In
many cases this phenomenon is followed by the appearance, in the ven-
tral retropharyngeal area, ofa copulatory apparatus often provided
with its oWn genital pore. Furthermore, as the breeding proceded,
numerous ovigerous capsules (the characteristic "cocoons") were found.
They had the typical peduncles of the genus Dugesia but have always
turned out to be completely sterile.
The histological examination of specimens fed with the above mentioned

crushed tissues has clearly demonstrated the development of two hyper-
plasic ovaries in the usual body areas. These ovaries had many oocytes
which did not, however, develop beyond a limited stage of growth. Nu-
merous testicular nodules rich in spermatogonia were also present,
occasionally with spermatocytes present. Generally, a more or less com-
plete copulatory apparatus was present, showing the typical characters
of those existing in sexual strains of the above mentioned species
(Figs. 3 - 5).
It is important to remark that the agamic strains of Dugesia gonocephaZa
and Dugesia tigrina used in these experiments, had been bred in our
laboratory for several years. They have been fed on either live Tubifex
or beef heart, without ever having shown any tendency to reach a sexual
state in any season of the year. The effects obtained are no doubt due
to the result of a sexualizing activity brought about by the type of
nourishment to which the animals have been experimentally subjected.
The specimens sexualized by means of feeding show, however, the same
abortive characteristics of gonads and reproductive sterility found
in the few specimens of the agamic strains which on rare occasions,
spontaneously attain the sexual state. These are the ex-fissiparous
animals described by Benazzi (1938).
The sexualizing effects achieved by feeding methods are similar to the

ones obtained by Kenk (1941) and Okugawa (1957), who grafted into the
two posterior thirds of the body of planar ian agamic strains the an-
terior third of the body of planarian sexual strains. The same authors
report to have unsuccessfully tried to induce a sexual state in plana-
rian agamic strains by feeding them in a way similar to my own method.
Their failure could probably be due to the short duration of the treat-
ments. Vandel (1921) also unsuccessfully tried to obtain the same
result by injecting the parenchyma of agamic planarians with crushed
tissues of sexually mature planarians.
25
Fig. 3. Hyperplasic ovaries of Dugesia gonocephala in cross-section. x 140

Fig. 4. Cross-section of testis with many spermatogonia. x 350
Fig. 5. Cross-section of copulatory apparatus. x 140
26
In sexualized strains of Dugesia gonocephala I have been able to verify

that the sexual state obtained lasted for several months. There was
no resumption of reproduction by fission, even after prolonged treat-
ment with crushed tissues of PoZ-ycelis nigra was stopped and replaced
by a normal feeding based on live Tubifex, beef meat, or even after
feeding on crushed tissues from specimens of the same agamic strain
of D. gonocephala. The latter finding is interesting, as it seems to
signify that in tissues of agamic strains sex-inhibitory substances
are not present. However, in asexual strains of Dugesia tigrina the sex-
ual state is lost after about one month of feeding on crushed tissues
of P. nigra, and multiplication by fission is resumed.
All our present knowledge about the phenomena of sexual maturation in

planarians suggests the hypothesis that the sexualizing substance,
contained in the feeding mixtures used in our experiments, consists
of a neurosecretion acting as a "gonado-stimuline" (according to the
name proposed by Lender, 1964). It could be elaborated by the nervous
systems of planarian sexual strains, but not of planarian agamic
strains. My latest research work on this subject (Grasso, 1973) demon-
strated that the whole nervous system in sexual strains, seems to pro-
duce such a neurosecretion. An equally strong sexualizing activity is
produced, no matter which mixture of crushed tissues (irrespective of
their origi~) from the anterior, median or posterior third of the body
of sexual specimens of Polycelis nigra is used.
For the purpose of isolating this sexualizing substance, a first at-

tempt has recently been made (Grasso et al., 1975) with good results.
About 600 specimens of Polycelis nigra (4 gm) were homogenized in 20 ml
bicarbonate saline solution. The homogenate thus obtained was centri-
fuged at 800 g for 10 min. The first supernatant was removed and cen-
trifuged at 18,000 g for 20 min and the second supernatant was centri-
fuged at 105,000 g for 120 min. Thus, we obtained 4 fractions, the first
containing the sediment at 800 g, the second that at 18,000 g, the
third that at 105,000 g and the fourth containing the last supernatant.
These single fractions were administered for 10 weeks to 4 groups af
5 planarians each, belonging to the usual Dugesia gonocephala agamic
strain. Only the first and second fractions have proved to be effec-
tive in inhibiting the continuation of agamic reproduction by fission,
and in inducing the sexual state. After 5 weeks two animals of each
group were killed. Histological examination showed that the specimens
belonging to the first and second groups were characterized by a com-
plete sexualization induced by the alimentary fractions, whereas those
of the third and fourth groups showed no sign of an hermaphroditic
genital apparatus. Moreover, the ultrastructural examination of each
single fraction of crushed feeding substance has shown that only in
fractions 1 and 2 are detectable (mixed up with fragments of other
cellular structures), formations surrounded by an irregularly shaped
membrane. These may be considered as the "neurosecretosomes" (isolated
nerve endings, Labella and Sanwal, 1965; Bindler et al., 1967; Norstrom
and Hansson, 1972) which are fairly rich in neurosecretory granules
(Figs. 6 and 7). The presence of such structures, in only the first
two fractions of crushed feeding tissues, supports the hypothesis that
the sexualizing substance may be identified with a neurosecretion.
Investigations are proceding to test the validity of this suggestion

and also to attempt to isolate and characterize the biochemical nature
of the neurosecretions endowed with sexualizing powers.
27
Fig. 6. Neurosecretosomes in fraction 1. x 100,000

Fig. 7. Neurosecretosomes in fraction 2. x 100,000
References
Benazzi, M.: Ricerche sulla riproduzione delle planarie Tricladi Palu-

dicole con particolare riguardo alla moltiplicazione asessuale.
M. Accad. Linc~i ser. VI, 7, 33-89 (1938).
Bindler, E., Labella, F.S., Sanwal, M.: Isolated nerve endings (neuro-
secretosomes) from the posterior pituitary. Partial separation of
vasopressin and oxytocin and the isolation of microvesicles. J. Cell
BioI. 34, 185-205 (1967).
Bondi, C.~Pascolini. R.: Osservazioni sulla neurosecrezione in Dugesia
lugubris e i suoi rapporti con la rigenerazione. Acta Medica Romana
4, 1-6 (1966).
Brandi, L., Ghirardelli, E.: L'azione del cervello sulla rigenerazione
delle gonadi di Dugesia lugubris. Rend. Accad. Naz. Lincei 12, 120-
125 (1963).
Bunt, A.H.: Formation of coated and "synaptic" vesicles within neuro-
secretory axon terminals of the Crustacean sinus gland. J. Ultra-
struct. Res. 28, 411-421 (1969).
Cecere, F., Grasso, M., Urbani, E., Vannini, E.: Osservazioni auto-
radiografiche sulla rigenerazione di Dugesia lugubris. Rend. 1st. Sci.
Camerino 5, 193-198 (1964).
De Beaucharnp~ P.: Classe des Turbellaries. Turbellaria (Ehrenberg,
1831). In: Grasse, P.P.: Traite de Zoologie (ed. Masson), Vol. IV,
pp. 35-312. Paris: Masson 1961.
Dubois, F.: Contribution a l'etude de la migration des cellules de
regeneration chez les Planaires dulcicoles. Bull. Biol. Fr. Belg.
~, 213-283 (1949).
Fedecka-Bruner, B.: Sur quelques conditions du maintien, de la rege-
neration et de la maturation des testicules chez la planaire Dugesia
lugubris. Bull. Soc. Zool. France ~, 287-293 (1967a).
28
Fedecka-Bruner, B.: ~tudes sur la regeneration des organes genitaux

chez la planaire Dugesia ZU(fubY'is. I. Regemeration des testicules
apres destruction. Bull. Biol. Fr. Belg. 101,255-319 (1967b).
Fedecka-Bruner, B.: ~tudes sur la regeneration-des organes genitaux
chez la planaire DU(fesia ZU(fubY'is. II. Regeneration de l' appareil
copulateur. Bull. Biol. Fr. Belg. 102, 3-44 (1968).
Grasso, M.: Fenomeni rigenerativi e apparato genitale in Dugesia lU(fubY'is.
Boll. Zool. 26, 523-526 (1959).
Grasso, M.: L'organizzazione dell'apparato genitale ermafrodita in
esemplari rigenerati e in esemplari bicefali di Dugesia lU(fubY'is.
Rend. Accad. Naz. Lincei 35, 101-104 (1963).
Grasso, M.: Rigenerazione cefalica e apparato genitale in Polyeelis
nigY'a. Rend. Accad. Naz. Lincei 37, 203-206 (1964).
Grasso" M.: Rapporti tra sistema nervoso periferico e gonadi nell'
apparato geni tale ermafrodi ta di Dugesia lugubY'is. M:lnit. Zool. Ital. 73,
172-131 (1965a).
Grasso, M.: Presenza e distribuzione delle cellule neurosecretrici in
DU(fesia lugubY'is. Monit. Zool. Ital. 73, 182-187 (1965b).
Grasso, M.: Dimostrazione di cellule neurosecret>::ici in Dugesia tigY'ina.
Rend. Accad. Naz. Lincei 38, 712-714 (1965c).
Grasso, M.: Prime indagini sul significato funzionale della neuro-
secrez ione in Dugesia Zugubris e DU(fesia tigY'ina. Bol:!.. Zool. 32,
1037-1049 (1965d).
Grasso, M.: Rapporti fra sistema nervoso, gonadi e neurosecrezione in
PolyeeUs nigY'a. Riv. Biol. 1.2., 157-172 (1 966a) .
Grasso, M.: Neurosecrezione e rigenerazione in embrioni di DU(fesia
lU(fubY'is. Boll. Zool. 33, 371-380 (1966b).
Grasso, M.: Sui fenomeni~i neurosecrezione durante la rigenerazione
di dischetti isolati di Dugesia lugubY'is. Arch. Zool. ital. .2.1, 327
- 336 (1966c).
Grasso, M.: Esperimenti sul controllo della maturazione sessuale in
ceppi agami di Planarie. Boll. Zool. 38, 532-533 (1971).
Grasso, M.: Ulteriori dati sull'induzione-della sessualita in planarie
di ceppo agamo nutrite con planarie sessuate. Boll. Zool. 39, 627
-628 (1972).
Grasso, M.: Further observations on the induction of sexuality in
agamous Planarian strains. Boll. Zool. 40, 393-397 (1973).
Grasso, M., Benazzi, M.: Genetic and physiologic control of fissioning
and sexuality in Planarians. J. Embryol. expo Morph. 30, 317-328
(1973). -
Grasso, M., Montanaro, L., Quaglia, A.: Studies on the role of neuro-
secretion in the induction of sexuality in planarian agamic strain.
J. Ultrastruct. Res. (1975, in press).
Grasso, M., Quaglia, A.: Studies on neurosecretion in planarians.
I. Neurosecretory fibres near the testes of DU(fesia lugubY'is. J. Sub-
micro Cytol. 2, 119-125 (1970a).
II. Observations on the ovaries of Dugesia lugubY'is. J. Submicr. Cytol.
2, 1 27 -1 32 ( 1 97 Ob) •
III. Neurosecretory fibres near the testes and ovaries of PolyeeUs
nigY'a. J. SUbmicr. Cytol. 3, 171-180 (1971).
Kaplonska, J.: Neurosecretion in Tricladida with the reference to
seasonal variations and changes occurring in the course of regene-
ration. Zool. Pol. 17, 73-95 (1967). .
Kenk, R.: Sexual and asexual reproduction in EuplanaY'ia tigY'ina (Girard).
Biol. Bull. 73, 280-294 (1937).
Kenk, R.: Induction of sexuality in the asexual form of Dugesia tigY'ina
(Girard). J. expo zool. 87, 55-69 (1941).
Labella, F.S., Sanwal, M.: Isolation of nerve endings from the poste-
rior pituitary gland. Electron microscopy of fractions obtained by
centrifugation. J. Cell BioI. ~, 179-193 (1965).
29
Lender, Th.: Le role inducteur du cerveau dans la regeneration des

yeux d'une Planaire d'eau douce. Bull. Biol. Fr. Belg. 86, 140-215
(1952).
Lender, Th.: Mise en evidence et role de la neurosecretion chez les
Planaires d'eau douce (Turbellaries, Triclades). Ann. Endocr. Paris
25, 61-65 (1964).
Lender, Th., Gabriel, A.: Le comportement des neoblastes pendant la
regeneration de la Planaire Dugesia lugubris (Turbellarie, Triclade).
Bull. Soc. Zool. Fr. 86, 67-72 (1961).
Lender, Th., Gabriel, A.~Les neoblastes marques par i'uridine tritiee
migrent et edifient le blasteme de regeneration des Planaires d'eau
douce. C.R. Acad. Sci. 260, 4095-4097 (1965).
Lender, Th., Klein, N.: Mise-en evidence de cellules secretrices dans
le cerveau de la Planaire Polycelis nigra. Variation de leur nombre
au cours de la regeneration posterieure. C.R. Acad. Sci. 253, 331
-333 (1961).
Lender, Th., Klein, N.: Les cellules neurosecretrices de Dendrocoelwn
lactewn. Bull. Soc. Zool. Fr. 87, 380-381 (1962).
Lender, Th., Zghal, F.: Influencedes conditions d'elevage et de la
neurosecretion sur les rhythmes de scissiparite de la race asexuee
de Dugesia gonocephala. Ann. Embryol. Morph. 2, 379-385 (1969).
Liotti, F.S., Bruschelli, G., Rosi, G.: Variazioni dell'attivita neuro-
secretoria durante la rigenerazione in esemplari di Dugesia lugubris
trattati 0 non con simpamina. Riv. Biol. 59, 353-384 (1966).
Liotti, F.S., Rosi, G.: Osservazioni sui diversi stadi del ciclo neuro-
secretorio di Dugesia lugubris. Riv. Biol..§.i, 419-432 (1968).
Morita, M., Best, J.B.: Electron microscopic studies on Planaria.
II. Fine structure of the neurosecretory system in the planarian
Dugesia dorotocephala. J. Ultrastruct. Res. 13, 396-408 (1965).
Norstrom, A., Hansson, H.A.: Isolation and characterization of neuro-
secretory granules of the rat posterior pituitary gland. Z. Zell-
forsch. 129, 92-113 (1972).
Okugawa, K.I.: An experimental study of sexual induction in the asexual
form of Japanese fresh-water Planarian, Duges-ia gonocephala (DugeS).
Bull. Kyoto Gakugei Univ. 11, 8-27 (1957).
Sauzin-Monnot, M.J.: ~tude ultrastructurale du tissu nerveux et des
produits de secretion nerveuse, au cours des premieres heures de
regeneration de la planaire Polycelis nigra (Turbellarie-Triclade)
au niveau de la blessure. Ann. Embryol. Morph. 5, 257-265 (1972).
Sengel, C.: Culture in vitro de blastemes de regeneration de Planaires.
Stephan-Dubois, F., Gusse, M.: Migration des neoblastes en cas de
regeneration anterieure, chez la Planaire Dendrocoelwn lactewn. C.R.
Sci. Biol. 164, 1825-1829 (1970).
Teshirogi, W.:'Transplantation experiments of two short pieces of a
fresh-water Planarian Bdellocephala brunnea. Jap. J. Zool. li, 21-48
(1963).
Teshirogi, W., Fujiwara, H.: Some experiments on regression and dif-
ferentiation of genital organs in a fresh-water planar ian Bdello-
cephala brunnea. Sc. Rep. Hirosaki Univ. 11, 38-49 (1970).
Ude, J.: untersuchungen zur Neurosekretion bei Dendrocoelwn lactewn Oerst.
(Plathelminthes-Turbellaria). Z. wiss. Zool. 170, 223-255 (1964).
Vandel, A.: La regeneration des glandes genitales chez les Planaires.
C.R. Acad. Sci. 172, 1072-1074 (1921).
Vandel, A.: Recherches experimentales sur les modes de reproduction
des Planaires Triclades Paludicoles. Bull. Biol. Fr. Belg. 55,
343-518 (1922).
Vendrix, J.J.: Existence de cellules neurosecretrices chez Polycelis
nigra (Ehrenberg) et Dugesia gonocephala (Duges) (Triclades Paludicoles)
Characteristiques cytologiques et histochimiques. Bull. Soc. Liege
3-4, 293-303 (1963).
Wolff, E., Dubois, F.: Sur la migration des cellules de regeneration
chez les Planaires. Rev. Suisse Zool. 22, 218-227 (1948).
Sex Differentiation In Regenerating o/e;;. Nemertine Chimeras
J. Sierne
A. Introduction
In a symposium devoted to intersexuality in the animal kingdom, it

would be appropriate to discuss the two aspects of this sex status
in the invertebrate phylum of Nemertinea: the natural and functional
intersexuality (or bisexuality) in hermaphroditic worms and the ab-
normal, pathological or experimental, intersexuality (or sex reversal
stage) in dioecious animals.
Within the developmental and physiological scope of this paper, we

shall take into consideration only the second aspect, as experimental
approaches have not been reported on rarely occurring bisexual species
of Nemertines.
With this object, we shall attempt to relate and interpret male/female

parabiotic experiments since the existence of sex inductive substances
and/or sex hormones has been previously demonstrated by such experi-
ments in vertebrate groups (well documented reviews by Burns, 1961,
and Gallien, 1973).
It is now known that this subject received a great deal of attention

sixty years ago because of the pioneer works of Lillie (1916) and
Keller and Tandler (1916) on the role of sex hormones in the differen-
tiation of mammalian sex organs and ducts as demonstrated by their
studies on the freemartin, a parabiotic "experiment of nature" as
Lillie called it.
Our own investigations, as reported in this paper, involve a series of

experiments on the freemartin effect in which transformation in the
opposite directions from ovaries to testes was induced by means of
parabiosis between male and female nemertine worms.
B. Materials
We used as experimental material Lineus ruber and L. sanguineus collected

on the shore of the Channel and L. lacteus from the Mediterranean. In
the laboratory they were kept in a cold room maintained at 12° ± 1°C.
Controls as well as experimental specimens were put separately in

small glass bottles, each containing 50 cl of sea water. They were
fed with calf liver once a week. The day after feeding the remains of
food and the faeces were removed together with the sea-water as a whole
and fresh sea-water was added. Special care was taken to prevent con-
tamination of this medium by predatory Protozoans.
31
Lineus ruber (Heteronemertini) was chosen as the principal object of

investigation because its characteristics are quite unique and rele-
vant for our own purpose.
First, the worms of this species are more common. They are also more
resistant to adverse effect than other species, and they are easily
kept in laboratory conditions.
Secondly, sex dimorphism is particularly distinct in these strictly

gonochoristic animals. It has been generally assumed by zoologists
that the sexes in the Nemertines of the family Lineidae are indistin-
guishable without gonad examination. While this is true for most spe-
cies of the genus Lineus it was recently discovered that in at least
one species, L. ruber, the sexes are truly dimorphic (Bierne, 1970a).
In the female L. ruber worm, specific basophilic dermal glands are
scattered allover the body, while in the male worm, special cutaneous
gland cells are found only ventrally. Furthermore, in the male, sperm
ducts are located ventrally, below the testes, carried between the
intestinal diverticula along the sides of the body while, in the fe-
male, oviducts are situated dorsally of the ovaries. The position of
the sperm pores on the ventral surface of the body is an indication
that internal fertilization takes place since such a position permits
the male to 'press its body against the open oviducts located dorsally
in the female.
A final consideration in the selection of L. ruber as a material of

study is that these nemertine worms have been used previously for dem-
onstrating an endocrine control of their reproductive processes by a
brain hormone (Bierne, 1964, 1966, 1970a, 1973).
c. Methods
Animals that were to undergo an operation were anesthetized for 15 min

in distilled water containing 8% magnesium chloride and then stretched
over a bed of paraffin on the stage of a dissecting binocular micro-
scope where they were cut with a micropenknive or a razor blade frag-
ment.
For grafting, the anesthetized nemertine pieces that were to be joined

were put wound by wound over a wet filter paper circle laid in a Petri
dish for 24 hrs. Separation of pieces was prevented by a prolongated
anesthesia. The details of the grafting technique have been previously
described elsewhere (Bierne, 1970a).
All the experiments involved the procedure of joining in parabiosis

male and female parts of two adults divided longitudinally into right
and left halves (Fig. 1). With such a procedure male and female pieces
were brought together to form a single worm. The expression d' / ~ chimera
has been chosen for such an "allophenic" (term coined by Mintz, 1970)
and heterosexual worm. It is evident that only histocompatible allo-
geneic and xenogeneic combinations (Langlet and Bierne, 1973; Bierne
and Langlet, 1974) have allowed this work.
For investigating sex differentiation, a new and complete organogenesis

of all the sexual characteristics was induced in d' / ~ chimeras called
to regenerate after castration and removal of somatic sexual charac-
teristics by esophageal transection (cf. Bierne, 1968). Indeed, the
amputation of the posterior genito-intestinal region is always followed
32
by its regeneration, one side of the regenerate developing from genet-

ically male, the other from genetically female tissue.
Moreover, the brain extirpation was performed simultaneously with

esophageal transection for a small number of experiments in order to
study the influence of the brain hormone (gonad inhibiting factor:
G.I.F.) upon the regeneration of sexual characteristics in chimeras
(cf. Bierne, 1970a).
-- 0
0- -- 0
0
0
_0
0
-
• • 0 0 0
• 0 • • 0 0
• •
-•
0 0 0
0 • 0 0
Fig:. 1. Grafting proce-
• 0 0 0
dure by which sagittal
• 0
• 0 0
c! / 9- chimeras were made
• 0
• 0 0
• from halves of male and
• 0 • 0 0 female nemertine worms.
White circles: testes;
~ ~ black circles: ovaries.

(From Bierne, 1968)
D. Results
I. Regenerating Normal d' /9- Chimeras
A series of 26 regenerating d'/9- chimeras formed from L. ruber specimens

was used for this study. Of these 26 heterosexual worms, 6 were killed
at 6/9 months, 17 at 10/20 months and 3 at 60 months of regeneration.
Thes e 3 periods correspond to three different sexual states observed
in the regenerates (Table 1).
1. Transitory Gynandromorphous Effect
During the initial stages of gonadal differentiation in the d'/~ regen-

erates (6/9-month regenerating chimeras) an autonomous sex organogen-
esis is observed, with small ovaries and testes developing on their
33
respective side. The first sexual status of chimeras is a phenotypic

dissymetry of juvenile gonads superposed upon the genetic sexual dis-
symetry of worms as for bipartite gynandromorphous animals. However,
the differentiation of testes starts sooner than the differentiation
of ovaries. Thus, a slight male.dominating effect is early obtained
(Bierne, 1968, 1970b).
Table 1. Sex differentiation in regenerating.if/~ chimeras
I. Transitory gynandromorphous effect (autonomous sex organogenesis)

II. Freemartin effect (masculinization of the ~ part)
III. Very late effect (feminization of the if part)
Period Sex characteristics
of rege-
No.of neration Gonads Genital tracts Dermal gland s
chime- (in
Effect ras months) if part ~ part if part ~ part if part ~ par t
I 6 6 - 9 if ~ if ~ - -
II 4 10 - 13 if <:f' and if if ~ and if - -
2 10 - 13 if if if if - -
5 13 - 18 if if if if if -
4 16 - 20 0' if if ~ and cf' cf' -
1a 18 if if if and ~ ~ and cf' cf' -
1b 20 cf' ~ if ~ cf' 0
+
III 3 60 ~ ~ if ~ andif ~ ~
aOnly in this one case, oviducts developed in the male part of the
regenerating chimera.
bExceptional permanent and complete gynandromorphous effect in a normal
if/~ chimera.
2. Freemartin Effect
Subsequently, in 9/20-month regenerating chimeras (Fig. 2), masculini-

zation of the ovaries occurs and may lead to complete sex reversal.
The female gonads become neo-testes. Furthermore sperm ducts may de-
velop in addition to or may replace oviducts.
However, when testes and neo-testes are mature, male-specific dermal

gland cells develop only in male halves of chimeras. In contrast, no
male-specific as well as female-specific reproductive dermal glands
are observed in female halves, the cutis of which remains sexless
(Bierne, 1968, 1970a).
3. Very Late Effect
Ultimately, in 3 very old regenerated chimeras (5 years after the eso-

phageal transection) I have found recently that there was an effect
opposite to freemartin, produced in the same specimens in which the
masculinization of the female part had been previously observed by
removal of pieces at the appropriate time.
34
Fig. 2. Two 10-month reg e n-

erating if/~ chimeras made fr o m
complementary anterior halves
of male and f e male Li neus ruber,
in a reciprocal sagittal ar-
rangement. Notice the allo-
phenic pigmentation of each
specimen
Indeed,if/~ chimeras which remained bipartite allophenic specimens

for the pigmentation possessed only ovaries and showed secondary fe-
male characteristics (basophilic cutaneous glands) in the two side
halves of each heterosexual worm. In contrast s p erm duc t s in the male
parts and new sperm ducts and / or ov iducts in the female parts remained
on their respective side (Fig. 3).
Therefore, this very late effe ct shows:

- that masculinized female halves may b e come femal e again according
to their starting sex;
- that male halves may be feminized (gonads and s e c o ndary somatic
characteristics) .
II. Regenerating Interspecific if/~ Chime ras
In histocompatible interspecific chimeras (1/2 d' L. sanguine us / 1/2

~ L. lacteus) in which gonads developed after 20 months of posterior
regeneration (Fig. 4), the freemartin effect has be e n also obtained.
Fig. 5 shows gonadal aspects commonly. observed in ma sculinized female
and male halves of such heterosexual 'xenogeneic worms.
III. Regenerating Brainless d'/ ~ Chimeras
In L. r uber, ablation of the anterior end or decerebration of worms

regenerating the gonado-intestinal region of the body greatly accel-
erates gonadal maturation and the differentiation of somatic sexual
35
characteristics, namely, gonoducts and special dermal gland cells.

The brain inhibitory influence upon sexual differentiation operates
in both sexes. However, it is more effective upon ovaries than testes.
Furthermore, other experiments show that this brain control of matu-
ration is exerted by a hormone called the gonad inhibiting factor,
G.LF., originating from the cerebral ganglia (Bierne, 1964, 1966,
1970a) .
\
Fig. 3. Very late effect (feminization) in a 5-year regenerating if/~
chimera. Gonads and sex ducts are indicated by arrows
An additional experimental series has been performed to answer the ques-

tion if the brain is involved in the sex reversal effects obtained in
regenerating if /~ chimeras.
The sexual status of 5 brainless chimeras, killed at 6 months of pos-

terior regeneration, was compared with that of 5 regenerating normal
chimeras killed at the same times (controls).
36
Fig. 4. A 20-month regenerating inter-

specific d'/Cf chimera (male half: Lineus
sanguineus from the Channel; female
half: Lineus lacteus from the Medi ter-
ranean Sea) showing the xenophenic
pigmentation of the worm
No gonads or indifferentiated gonadal primordia have been observed in

three controls. In the two others, the transitory gynandromorphous
effect, i.e. very small ovaries and testes developing in respective
female and male halves, occurred in accordance with the previous ex-
periments.
On the other hand, in all regenerating brainless chimeras the sexual

differentiation was precocious.
The upper part of the regenerated body showed testes and ovaries in
gametogenesis, sperm ducts and oviducts, male-specific ventral glands
and female-specific cutaneous glands (permanent and complete gynandro-
morphous effect, Fig. 6).
The freemartin effect was obtained in the posterior part of the same
regenerates: testes and ovotestis below which were testes and neo-
testes, male-specific cutaneous cells and sexless cutis (Fig. 6).
E. Discussion
I. Gonads
In order to explain the freemartin effect obtained in reg e nerating

d'/Cf nemertine chimeras two kinds of phenomena may occur:
- migration of determinate male cells from male halves to female
halves and corre lative inhibition of the determinate female cells;
37
Fig. 5 a-d. Gonads of the male (a) and female (b, c and d) halves in
the same regenerating chimera as shown in Fig. 4
- transfer of a diffusible androgenic substance from male halves to

female halves.
In the same way, two interpretations are possible for the very late
effect (feminization):
- migration of determinate femal e cells from female halves to male
halves and correlative inhibition of the determinate male cells;
- defici e ncy of the androgenic substance in the old h e terosexual worms.
It is obvious that none of the above hypotheses finds confirmation or

is rejected by our current results.
However it appears reasonable to retain Lillie's hormonal interpreta-

tions (1916) as the basic explanation because this interpretation is
in any case a necessity in order to explain somatic effects and because
of the modern theories of sex differentiation in other animal groups
as elaborated by Charniaux-Cotton (1965) for crustaceans.
38
cf
OVIduct
~~ --~~~~~~\; ~
\ 2001'
cf
sperm duet
mole gland sperm duct
Fig. 6. Permanent gynandromorphous effect, in the anterior part (above)

and freemartin effect in the posterior part (below) of the regenerated
body of a 6-month regenerating brainless if/~ chimera. Notice the ab-
sence of reproductive gland cells in the cutis of the posterior female
half. (From Bierne, 1970a)
II. Sex Ducts
Since sperm ducts may develop in addition to or may replace oviducts

in genetic female halves of the heterosexual worms that show the free-
martin effect, it is only logical to assume that the primordia of the
male ducts are present in the genetic female tissues but can be ex-
pressed only in the presence of an androgenic factor. In addition,
this factor apparently exerts an inhibiting influence on the primordia
of the female ducts.
39
III. Secondary Sexual Characteristics
After complete masculinization of the gonads and sex ducts in regener-

ating ~/~ chimeras, male-specific ventral glands are always absent in
the genetic female halves. This lack of responsiveness of female cuta-
neous target cells suggests sex differences in the reactivity of male
and female cutaneous targets.
In mammals, some observations suggest also that the genetic constitu-

tion of tissues influences the sensitivity to hormonal mediators.
Tissues from male fetuses seem to be more sensitive to male hormone
than tissues from female fetuses (Jost, 1967; Price and Pannabecker,
1959) .
On the other hand, after feminization of the old if/~ chimeras, female-
specific cutaneous glands are present in the genetic male halves. This
indicates that the target dermal cells are present in the male but
develop only in the absence of testes, i.e. in the absence of the
androgenic factor. The female-specific gland differentiation is per-
haps inhibited by the androgenic factor in the ordinary course of male
development.
IV. Brain Influence
According to the presence or the absence of the gonad inhibiting cere-

bral factor (G.I.F.) during the sex differentiation of regenerating
if/~ chimeras, the sexual phenotype of anterior parts of regenerates
shows a masculinization of the female halves (freemartin effect) or
corresponds to the sex organogenetic autonomy (gynandromorphous effect) .
The freemartin effect must be due to the masculinizing influence of

the androgenic factor produced by the male part and the gynandromor-
phous effect must be attributed to the fact that the early differen-
tiated ovaries and female cutaneous glands are no longer able to react
to this factor when it is produced.
F. Conclusion
The theories of sex differentiation as elaborated by Jost (1953) in

mammals and Charniaux-Cotton (1965) in crustaceans indicate that the
normal initial male differentiation depends upon a "male organizing"
or "androgenic" substance secreted by special male cells. In its ab-
sence as in normal females the first sex characteristics differentiate
along female lines, i.e. according to the neutral sexual state.
The present work introduces an element of unity into the concept of

the mechanism of sex differentiation in gonochoristic animals since
the above theories, explaining the sex phenomenon on the basis of an
androgenic factor, may be supported by the results of our experiments
in nemertine worms.
For masculinization to occur in female halves of regenerating iff~

chimeras, a diffusible androgenic substance emanated from the male
part must be probably transferred to the female part.
The late feminization in male halves of old regenerated ~/~ chimeras

is believed to be due to the deficiency of masculinizing substance in
the old heterosexual worms.
40
The gynandromorphous effect in brainless if/9 chimeras is believed to be

due to the fact that the early differentiated female characteristics
are no longer able to react to the androgenic factor when it is pro-
duced by the male part.
Therefore, nemertines show evidence of vertebrate-like and crustacean-
like sex control mechanisms characterized by the first event: the auto-
nomous initial female organogenesis and the male organogenesis induced
by an androgenic factor.
Some of the organisms which have attracted the least attention from
biologists may perhaps contribute to the understanding of biological
control mechanisms in general and to sex differentiation in particular.
References
Bierne, J.: Maturation sexuelle anticipee par decapitation de la fe-

melle chez l'Heteronemerte Lineus ~er Muller. C.R. Acad. Sci. (Paris)
259, 4841-4843 (1964).
Bierne, J.: Localisation dans les ganglions cerebroides du centre regu-
lateur de la maturation sexuelle chez la f emelle de Lineus rUber Mul-
ler (Heteronemertes). C.R. Acad. Sci. (Paris) 262, 1572-1575 (1966).
Bierne, J.: Facteur androgene et differenciation du sexe chez la Nemerte
Lineus ~er Muller. L'effet "free-martin" dans la parabiose hetero-
sexuee. C.R.Acad. Sci. (Paris) 267, 1646-1648 (1968).
Bierne, J.: Recherches sur la differenciation sexuelle au cours de l'on-
togenese et de la regeneration chez le Nemertien Lineus rUber Muller.
Ann. Sci. Nat. Zool. Biol. animale (Paris) 12, 181-298 (1970a).
Bierne, J.: Aspects experimentaux de la differenciation sexuelle chez
Lineus rUber (Heteronemertes). Bull.Soc.Zool.France 95, 529-543 (1970b)
Bierne, J.: Contrale neuroendocrinien de la puberte chez le male de
Lineus rUber Muller (Heteronemerte). C.R. Acad. Sci. (Paris) 276,
363-366 (1973). -
Bierne, J., Langlet, c.: Recherches sur l'immunite de greffe chez les
Nemertiens du genre Lineus. Etude de la reponse primaire ~ la trans-
plantation heterospecifique. C.R. Acad. Sci. (Paris) 278, 1445-1447
(1974). -
Burns, B.: Role of hormones in the differentiation of sex. In: Sex and
Internal Secretion (ed. W.C. Young), pp. 76-158. Baltimore: Williams
and Wilkins 1961.
Charniaux-Cotton, H.: Hormonal control of sex differentiation in inver-
tebrates. In: Organogenesis (eds. R.L. de Haan and H. Ursprung),
pp. 701-740. New York: Holt, Rinehart and Winston 1965.
Gallien, L.: Differenciation et organogenese sexuelles des Metazoaires.
Paris: Masson 1973.
Jost, A.: Problems of fetal endocrinology: the gonadal and hypophyseal
hormones. Recent Progr. Hormone Res. 8, 379 (1953).
Jost, A.: Modalities in the action of androgens on the foetus. Res. on
Steroids 3, 207-220 (1967).
Keller, K., Tandler, J.: tiber das Verhalten der Eihaute bei der Zwil-
lings-Trachtigkeit des Rindes. Wien. Tierarztl. Wschr. 3, 513 (1916).
Langlet, C., Bierne, J.: Recherches sur l'immunite de greffe chez les
Nemertiens du genre Lineus. Evolution de transplants homospecifiques
et heterospecifiques. C.R. Acad. Sci. (Paris) 276, 2485-2488 (1973)
Lillie, F.: The theory of the free-martin. Science 43, 611 (1916).
Mintz, B.: Gene expression in allophenic mice. In: Control mechanisms
in the expression of cellular phenotype (ed. M.A. Padykula), pp. 15
-41. New York: Academic Press 1970.
Price, D., Pannabecker, R.: Comparative responsiveness of homologous
sex duct and accessory glands of fetal rats in culture. Arch. Anat.
Micr. Morph. EXp. 48 bis, 223-243 (1959).
Sex Reversal in the Syllinae (Polychaeta: Annelida)
M. Durchon
A. Introduction
Among the syllidians, most of the Syllinae reproduce by stolonization.

Sexualization, in both the male and female occurs in a certain number
of posterior segments, the number being relatively constant for each
particular species and in direct relationship to the total numbers of
segments in each animal (Durchon, 1959). This sexualized posterior
part later develops a head, undergoes a series of muscular modifica-
tions, and acquires swimming chaetae on each parapod. These form the
stolon which, ultimately, breaks away from the earlier part or stock,
thus freeing the gametes at spawning time, and then dies. The stock
later re-develops and is capable of further stolonizations.
Observation of sexual stolonizations (both natural and experimental)

in the laboratory, by a selective ablation of the proventriculus, has
made it possible to assert the existence of successive hermaphroditism
in several species of Syllinae (Durchon, 1951, 1959; Junqua, 1957;
Wissocq, 1963-1964).
The species studied were found at Cherbourg (SyUis amica); Algiers

( SyUis vittata, TrypanosyUis zebra) ; and Roscoff (SyUis prolifera, S. spon-
gicola, S. variegata).
In the laboratory the animals were isolated in Petri dishes, in com-

plete darkness and at a fixed temperature, and were left unfed. The
water in which they were being reared was changed twice a week by close
observation through a stereoscopic microscope. Genital development was
observed both in SyUis undergoing natural stolonization and in those
in which stolonization had been experimentally induced by selective
ablation of the proventriculus. Each series of experiments involved
at least 25 animals.
C. Results
The results obtained in the course of natural and experimental stolo-

nizations and those observed in the context of genital development
under the influence of different temperatures have been recorded.
42
I. Genital Evolution during Natural Stolonization
A change of sex was observed in ·the majority of cases which underwent

several stolonizations, irrespective of which species were involved.
The general tendency was toward the male sex and only infrequently
toward the female.
1. Sex Reversal of the Protandric Type
This kind of successive hermaphroditism was observed in the following,

species: SyHis ami ca. S. vittata. S. prolifera (Table 1), S. spongicola.
Trypanosyllis zebra.
In S. amica the proportion of the sexes remained constant; i.e. 2/3

male, 1/3 female. Observation of sexual development showed that ap-
proximately 20% of the animals. underwent two successive stolonizations,
while 2/3 of the females changed into males and only 1/3 of the females
become males.
This masculinization was clearly demonstrated in the case of S. vittata;

at the fifth stolonization the sex ratio was 90% males and 10% females.
Observations of each animal showed that at the fourth stolonization
all the females became males, but only a small percentage of the males
(11 - 15%) became female; these were the only females present during
the last two stolonizations. Development toward the female sex was
observed in over twenty cases of stolonization (Fig. 1).
However, following the second natural stolonization in S. prolifera,

masculinization was complete.
Fig. 1. Sexual de-

velopment in Syllis
vittata during five
natural stoloniza-
tions
2. Protogynous-type Sex Reversal
This type of hermaphroditism was observed in S. variegata by Wissocq

(1963-1964). In this species, all 24 cases having shown a male ten-
dency during the initial stolonization, released secondary female
stolons.
Table 1. Sexual development during natural stolonization
1st stolonization 2nd stolonization 3rd stolonization 4th stolonization 5th stolonization
Species if <j> if <j> d' <j> d' <j> d' <j>
SyUis
arnica 66 % 34 % 67 % 33 %
Syllis
rrolifera 67 % 33 % 100 %
Syllis
vittata 54 % 46 % 61 % 39 % 72 % 28 % 88 % 12 % 90 % 10 %
--~
Table 2. Sexual inversion during experimental stolonization, in Syllis arnica
1st stolonization 2nd stolonization 3rd stolonization
Sex d' <j> d' <j> d' <j>

percentage
63 % 37 % 83 % 17 % 100 %
Sex reversal
if ---->-<j> 10 % o %
Sex reversal
<j>~d' 67 % 100 %
~-
--
Table 3. Sexual inversion during experimental stolonization, after

proventriculus ablation, in TrypanosyUis zebra
Sex d' <j> if <j> d' <j>

percentage
d' --'> <j> 79 % 21 % 87 % 13 % 100 %
Sex reversal
8 % o %
Sex reversal 90 % ./>0
<j> --'> 100 % W
44
II. Genital Development during Experimental Stolonization
It is well known (Durchon, 1957, 1959) that the selective ablation of

the proventriculus causes permanent stolonization in the Syllinae. It
has, therefore, been possible to follow genital development of treated
animals during successive stolonizations, outside the period of natural
reproduction. The formation of the first stolon occurs at variable
intervals, according to the species involved and the temperature of the
water in which they are breeding. In S. o;nica, S. vittata, S. prolifera and
T. zebra, the average delay is one month, at 20°C.
In all the species studied (Tables 2, 3, 4, 5) it is clear that from

the moment of the first stolonization, the percentage of males is al-
ways greater than that of the females. However, masculinization is
rapid. This fact is obvious from the percentage of males which may
reach 100% at the third stolonization (5. arnica, s. prolifera, T. zebra)
and also from the low percentage of males changing into females at the
third stolonization (10% of s. prolifera, 8% of T. zebra). This was also
confirmed from the high percentage of females again undergoing a change
toward the male at the second stolonization (67% of s. o;nica, 90% of
T. zebra). Masculinization is more rapid during experimental stoloni-
zation than during natural reproduction. An example of this case can
be seen for S. vittata, in Table 5.
III. Influence of Temperature
It is clear that temperature plays a particularly important role in

the sexualization of SyUinae. A certain temperature is needed for
stolonization to commence and to proceed to sexualization. This tem-
perature varies according to the species. Under natural conditions,
s. arnica reproduces in summer. Those which were gathered in winter and
placed in a temperature of 20°C, underwent a "natural" stolonization
in 10% of the cases. At 25°C each one was sexualized within 15 days
and was fully mature after a month. s. spongicola, gathered in winter
at Roscoff and bred at 20°C, also sexualized rapidly (Wissocq, 1963-
1964) .
Temperature also affects the direction of the sexualization, either

toward male or female. Higher temperatures encourage masculinization.
From 100 observations Wissocq (1963-1964), has shown that at temper-
ature of 20°C sexualization is entirely toward the male.
After ablation of the proventriculus, the speed of induced stoloniza-

tion in s. arnica was in direct relation to the degree of temperature
(1 month at 20°C, 1 week at 30°C). However, and most importantly, the
percentage of male stolons at the first stolonization was in direct
proportion to the temperature (Table 6).
D. Discussion
The Syllinae show sexual variations throughout stolonizations, whether

.natural or induced. The primordial germinal cells may, therefore, be
distinguished both in the male or female sense. This can be shown in
two ways: it is possible to demonstrate hetero-sexualization using
sections from the same animal, bred in isolation and at the same tem-
perature (20°C). This has been shown in S. vittata and s. arnica. In two
experiments, 7 examples out of 20 for the first, and 4 out of 20 for
Table 4. Sexual development during experimental stolonization, in Syllis prolifera
d' 'i' d' 'i' d' 'i'

64 % 36 % 100 % 100 %
----
Table 5. Sex reversal during natural and experimental stolonization, in SyUis vittata
1st stolonization 2nd stolonization 3rd stolonization 4th stolonization

d' 'i' d' 'i' d' 'i' d' 'i'
Natural
stolonization 54 % 46 % 61 % 39 % 72 % 28 % 88 % 12 %
Experimental
stolonization 88 % 12 % 96 % 4 % 95 % 5 % 100 %
Table 6. Influence of temperature on masculinization during experimental stolonization, in SyUis arnica
Stolonization and sex
Temperature 1st 2nd 3rd

d' 'i' d' 'i' d' 'i'
20°C 63 % 37 % 83 % 17 % 100 %
25 °c 72 % 28 % 92 % 8 % 100 %
30°C 81 % 19 % 94 % 6 %
Table 7. Hetero-sexualization of "pieces" from an animal
Species Anterior "piece" Median "piece" Posterior "piece"
SyUis vittata, 2d' 3d' 6d'

(7/20) 5'i' 4'i' 1 'i'
SyUis arnica 4d' 2d' Od' -t>
01
(4/20) 2'i' 4'i?
46
the second have displayed stolonization in differing sexual directions

and in different segments: anterior, median and posterior of the same
animal (Table 7).
An example of genuine hermaphroditism has been noted in three other

species: T. zebra, S. spongicola and S. arnica . Oocytes and spermatozoa
existed side by side in the same segment of a stolon (Fig. 2). This
observation is another argument in favour of a lack of sexual differ-
entiation in the primordial germ cells.
Fig. 2. Hermaphroditism in Sylli s arnica .

o Oocytes; S Spermatozoa
Of all the factors which may influence sexual differentiation, temper-

ature seems to be among the most important. It has been established
that the higher the temperature of the surrounding water, the higher
is the percentage of male stolons. Hauenschild (1953) came to the same
conclusion with results obtained from another Syllidian species, G~ea
clavata .
These temperature effects have also been demonstrate d in other inver-

tebrates, the Oligochaeta and the Arthropoda in particular.
Relexans (1970) demonstrated that with embryos of Eisenia f oetida obtained

from cocoons at a temperature of 32 °C, there was a degeneration of the
germinal cells (67% of the cas e s observed) and of sexual inversion
(12% of testicular feminization, 8% of ovarian masculinization). The
application of sudden heat (34 °C) over a period of several days, caused
feminization of the testicles, especially when this was carried out
during the first half of the development of the embryo. This applica-
tion of heat clearly affected the primordial germinal cells.
Ginsburger-Vogel (1973) showed by experiments carried out at a stable

temperature of 17 ° C, that in Orchestia gamrnareUa (intersexual males and
females gathered at La Penze, near Roscoff), the phe nomenon of thely-
genia was linked with intersexuality and affected by temperature. This
may be explained by total sex-inversion giving rise to females.
Dobzhansky (1930) has shown that the intersex triploid, Dr osophila mela-
nogas ter , is more readily feminized at 28 °C than at 15 °C. However,
Bergerard (1961a,b, 1967) established that raising of the temperature
encouraged the breeding of males and intersexes in Car ausius morosus ,
a species both parthenogenetic and thelytocus.
47
The influence of other factors has also been established. Durchon

(1959) established that during the 2nd and 3rd stolonizations T. zebra
formed stolons in newly-grown tails, which were nearly always male.
If these tail buds are regularly resected, the genital products which
form very slowly in at most five segments are (in contrast to what has
already been shown) generally female. Sexual differentiation in the
male direction seems therefore to be related to a coefficient of rapid-
ity in sexualization. The same conclusion can be reached by comparing
natural with experimental stolonizations. Once stolonization has arisen,
whether by ablation of the pharyngeal region or by selective ablation
of the proventriculus, the differentiation of the stolons occurs more
rapidly than under natural conditions. Correspondingly, even from the
first stolonization, the percentage of male stolons obtained is very
high.
In summarizing the results of different experiments (the effect of

different temperatures and experimental stolonization), it is clear
that the percentage of males increases with the spread of the process
of sexualization. It must be emphasized, however, that the Syllinae
have no gonads, outside their period of sexualization. These only de-
velop around the genital vessel when stolonization began. Therefore,
it appears that in these annelid polychaetes, sexual differentiation
is epigeneti~. This topic is now the subject of further research.
References
Bergerard, J.: In tersexuali te 'experimentale chez Carausius morosus Br.

(Phasmidae) . Bull. bioI. France-Belgique 2.2, 273-300 (1961a).
Bergerard, J.: Analyse de la periode sensible a la temperature pour
la differenciation sexuelle de Carausius morosus Br. (Phasmidae ). C. R.
Acad. Sci. 253, 2149-2151 (1961b).
Bergerard, J.: Arnbisexualite et intersexualite chez les Insectes.
Ann. BioI. 6,259-269 (1967).
Dobzhansky, T.: Genetical and environmental factors influencing the
type of intersexes in Drosophila melanogaster. Amer. Nat . .§.i, 261-271
(1930) .
Durchon, M.: Stolonisation et hermaphrodisme successif chez SyUis arnica
Quatrefages. Arch. Zool. expo et Gen. 88, N. et R., 96-100 (1951).
Durchon, M.: Role du proventricule dans le-determinisme de la stolo-
nisation chez les Syllidiens (Annelides Polychetes). C.R. Acad.
Sci. 244, 1883-1886 (1957).
Durchon,~: Contribution a l'etude de la stolonisation chez les
Syllidiens (Annelides Polychetes): I. Syllinae. Bull. bioI. France-
Belgique 93, 156-219 (1959).
Ginsburger-Vogel, T.: Determination genetique du sexe, monogenie et
intersexuali te chez Orchestia gammareUa Pallas (Crustace Arnphipode
Talitridae ). These Doctorat d' Etat, Uni versi te de Paris VI, no.
C.N.R.S. AO 7966 (1973).
Hauenschild, C.: Neue Versuche liber phanotypische Geschlechtsbestim-
mung bei Polychaeten. Naturwissenschaften 6, 186-189 (1953).
Junqua, C.: Stolonisation et polycephalie exp§rimentales chez Trypano-
syllis zebra Grube (Annelide Polychete). Ann. Sci. Nat. Zool. 11,
Ser. 19,59-68 (1957). -
Relexans, J.C.: Mise en evidence experimentale chez Ie Lombricien Eisenia
foetida d'une segregation precoce de la lignee germinale par applica-
tion d'une temperature elevee au cours du developpement embryonnaire.
C.R. Acad. Sci. 270, Ser. D, 977-980 (1970).
Wissocq, J.C.: Nouveaux cas d'inversion sexuelle chez les Syllidiens
(Annelides Polychetes). Mem. Soc. Nat. Sci. Nat. et Math. de Cher-
bourg 51,6. Ser., l, 105-109 (1963-1964).
Mutual Influence on the Sexual Differentiation in the
Protandric Polychaete Ophryotrocha puerilis
H.-D. Pfannenstiel
A. Introduction
The polychaete, Ophryotrocha pueriUs, is considered to be the typical

example for permanent potential hermaphroditism. Sex determination in
this species was formerly considered to be of an epigenetic nature.
This opinion was overruled by selection experiments, which revealed
that genetic factors playa part in sex determination (Muller, 1962).
Therefore, i t is likely to.qualify sex in Ophryotrocha pueriUs, as de-
termined by polyfactors (Bacci, 1965). Males and females of Ophryotrocha
do not show pexual dimorphism. The gametes themselves mark the differ-
ence between males and females. They develop from undifferentiated
gonia into oocytes or spermatozoa. The inducing conditions for this
change in the direction of differentiation were investigated during
the last three years (Pfannenstiel, 1971, 1973, 1975). The results
shall be briefly reported, as they are necessary for the understanding
of the influences which two individuals can exert on each other. The
experiments were carried out with individually raised and kept worms.
B. Results
I. Sexual Differentiation in Isolated Individuals
When Ophryotrocha pueriUs has reached a length of about 10 segments,

spermatozoa appear in the coelomic fluid and this marks the beginning
of the short juvenile male phase. At a length of 16 segments the germ
cells start to proliferate oocytes and the female phase is reached.
Another sex-reversal, into a secondary male, does not normally take
place in isolated females. The coelom of isolated females always con-
tains up to three generations of oocytes, which may be distinguished
by their sizes. Having reached the final size of about 120 ~m, the
oldest set of oocytes are not shed, but resorbed. Isolated females
never lay egg masses.
The upper jaw of o. pueriUs appears in two different forms. In juve-

nile individuals, the primitive jaw (p-jaw) looks like a small pair
of pincers, while the definite upper jaw (k-jaw) looks like a pair of
forcipated claws. From time to time, the p-jaw is replaced by a new
one, but only under certain circumstances is the p-jaw replaced by the
k-jaw. The definite jaw is a permanent structure and is not replaced
repeatedly. This process is similar to the moulting in insects. The
controlling mechanism in both cases may be similar: one condition being
responsible for the changing or moulting of the jaw, while another
determines the structure of it.
49
II. Sexual Differentiation in Couples
When two isolated females meet (Fig. la,b), the suppression of the
shedding of oocytes is set aside and one individual produces an egg
mass within 3 to 5 days (Fig. lc). The contact to another individual
also induces the replacement of the p-jaw by the definite k-jaw. With-
in 8 days this individual changes to the male phase and starts produc-
tion of spermatozoa. Both individuals now represent a couple and con-
tinually bring forth fertilized eggs (Fig. ld). At irregular intervals
both individuals may change sex (Fig. le). These sex reversals are
called simultaneous, since they occur within a relatively short time.
It is possible, however, that one individual marks the start of sex
reversal and the other responds by reversing its own sex. Normally,
the individual functioning as the male at that moment, starts to pro-
liferate oocytes instead of spermatozoa and thereby causes the other
female to switch to the male phase. This couple also brings forth fer-
tilized eggs.
isolated Couple isolated
a b
/
c
)1
8J(r'"
t 'iI
'"':.-0
r'
d I Fig. 1 a-g. The
"Paarkultureffekt".
~ ~
)
(a and b) Isolated
females, (c) couple,
one female has shed
oocytes, replaced
its jaw and produced
spermatozoa, (d)
e
/
couple brings forth
~
fertilized eggs,
(e) simultaneous sex
reversal in both
partners, (f and g)
partners separated,
f 9 male changes to the
female phase
50
Whenever a female becomes a secondary male, or more precisely, when

it touches with another female - it develops the k-jaw and never
changes it, even if separated from its partner. The subsequently
separated individuals either stay female (Fig. 1f) or become female
(Fig. 19). Hartmann and Huth (1936), who first studied sex determina-
tion in O. puerilis, characterized these observations by the term "Paar-
kultureffekt". The results of Muller (1962) and may own results (Pfan-
nenstiel, 1971, 1973), had some share in accurately defining this term.
III. Endocrine Control of Oogenesis
The "Paarkultureffekt" was the starting point for further investiga-

tions. Firstly, an attempt was made to analyse the conditions for the
regularly occurring change of sex in the ontogeny of O. pueriUs. Since
the reversal of sex in a female (induced by another female) is an in-
version of the first alternation of sex, the nature of the mutual in-
fluence must contradict the conditions which lead to the first produc-
tion of oocytes. Numerous striking experiments with other polychaetes,
as well as oligochaetes, showed a more or less complicated endocrine
controlling mechanism of gametogenesis, maturation and spawning and
also of epitoky in some species (Durchon, 1970; Clark and Olive, 1973;
Hauenschild,1974, 1975). The family Nereidae has especially attracted
the attention of many research workers. In nereids, a slowly decreasing
brain hormone titer causes the development and maturation of oocytes
and finally, metamorphosis of the heteronereis. The regeneration ca-
pacity depends on high concentrations of the brain hormone.
Gametogenesis in O. puer'iUs was also suspected to be controlled by a

humoral factor. Some experiments proved this assumption to be partially
true.
Decapitated females shed their oocytes, or resorb them at once and

thereafter produce spermatozoa. The p-jaw is not replaced. Thus, the
effect of another female could be partially imitated by decapitation.
Juvenile males, without their prostomia, did not reach the female
phase - even when they sould have started oocyte production within
the next days.
The effects of decapitation in females were prevented by implanting

prostomia from female donors. The hosts did not shed their oocytes
but continued to produce new ones. The implantation of female prosto-
mia into the coelomic cavity of decapitated males offered proof that
the resulting proliferation is stimulated by a hormone. Males with
implanted prostomia started to generate oocytes immediately.
IV. Mutual Influence on Sexual Differentiation
The "Paarkultureffekt" must, therefore, be based upon mutual influence

on the hormone system of the partner. Hartmann and von Lewinsky (1942)
claimed the so-called "Eistoffe" to be the active substance, which was
released into the seawater and exerted an influence on the partner.
In a simple experiment, Muller (1962) demonstrated that the mutual
influence is only exerted when the two partners are in direct contact.
Two females, separated by a fence which allowed water circulation but
prevented contact of the individuals, did not show the "Paarkultur-
effekt".
In culture bowls of O. puerilis mucous traces are left by the worms.

Although these mucous traces are recognized and followed by the indi-
viduals, they are not responsible for the mutual influence. This was
51
demonstrated in an experiment in which females were alternatively

changed in their culture vessels. The individuals were left for some
time (8 hrs up to 4 daysr in the bowl of their partner, but had no con-
tact with it. Not in a single case could the "Paarkultureffekt" be ob-
served. As soon as the two individuals met and had the chance of con-
tacting each other, one of them changed sex. Although the mucous traces
are marked individually they do not bear the information necessary for
sex reversal. The mutual influence must, therefore, be effected by
other means.
In another·series of experiments, the length of time was investigated

for which two females need to be in contact - until the normal effects
of mutual influence are determined. Shedding of the oocytes, formation
of the k-jaw and the start of spermatogenesis are considered to be the
normal effects of mutual influence. Normally, only one individual of
a newly meeting couple of females sheds its oocytes, changes its jaw
and finally brings forth spermatozoa. Only in a few cases did these
effects occur in both partners. Then, either the determination or the
realization of sex reversal was disturbed. The differences between
determination and realization, as far as the "Paarkultureffekt" is
concerned, will be discussed later. Consequently, isolated females
were brought together and the resulting couples were separated after
different times ranging from 4 hrs up to 5 days. The occurrence of
egg masses, k-jaws and sex reversals in the ex-partners was registered
(Fig. 2). Almost regularly, both sex-partners showed a certain amount
Couple 2
o
010
~o
""'0°
separated
after
20 couples for each experiment
Fig. 2. Couples consisting of two females separated after different

times (4 hrs up to 6 days); oocyte shedding, occurrence of k-jaw and
sex reversals are registered (%) for both partners (1, 2); average
time in days for occurrence of egg masses (4.0), k-jaw (5.4) and
spermatozoa (7.8)
52
of reaction. Reactions in both partners seemed to occur a little more

frequently than in normal couples which were afterwards not separated.
The first occurrence of sex reversal was noticed in individuals which
had been in contact with each other for 2 days. The changes of sex in-
creased with the contacting time and reached a maximum at 4 days. There-
fore, separation of the partners after 4 days could no longer prevent
sex reversal in at least one of them. The same correlation exists be-
tween contacting time and replacement of the p-jaw by the definite one.
The production of egg masses does not reach 100%, as the oocytes are
not always shed but sometimes resorbed. This is quite normal for the
"Paarkultureffekt". In partners separated after 5 days, the number of
sex reversals and k-jaws decreases slightly. This is not unusual, since
even in undisturbed couples of two females, sex reversal in one part-
ner does not occur in every case. Strictly speaking, the 100% sex re-
versals and k-jaws in partners which were separated after 4 days, oc-
curred by chance. Another result of this experiment must be mentioned.
The average times in days for the occurrence of the three effects were
calculated. After 4.0 days the oocytes were shed or resorbed. The k-jaw
appeared after 5.4 days while spermatozoa were first observed after
7.8 days. Although these times clearly show an order of succession,
the events may occur irrespective of sequence in an individual, or in-
completely. This means that there were individuals which only shed
their oocytes but did not change sex or their jaw. Others replaced the
jaw but did not change sex. Even sex reversals, without previous re-
placement of the jaw, were obtained.
Nevertheless, all these results justify the assumption that the con-
tact of two females has to last for at least about 4 days in order to
induce sex reversal and/or the other effects of a normal "Paarkultur-
effekt". For a longer maintenance of the male phase in one of the
partners, the contact has to be prolonged. Even secondary males start
oogenesis again when the couple is separated. Then, only the k-jaw is
a permanent reminiscence of the former contact with another female.
The normal course of sexual differentiation with the already mentioned
simultaneous sex alternations in both partners of a couple, requjres
permanent mutual influence.
With the following experiments, an attempt was made to localize that

part of the body which acts as a transmitter or receiver for the mu-
tual influence. Couples were formed in which one partner was deprived
of a certain part of its body.
In one experiment, a couple consisted of an intact and a decapitated

female (Fig. 3a). The inhibition of oocyte shedding never ceased in
the intact partner, but, as expected, did in the decapitated one.
The intact partner did not change sex or its jaw. The decapitated
partner changed to the male phase after shedding of its oocytes but
without replacement of its jaw. Sex reversal in decapitated individ-
uals is due to the lack of the prostomium, as shown before. Moreover,
no mutual influence was found in this experiment. This result suggested
the prostomium to be either the transmitting or the receiving site for
the mutual stimulus.
Consequently, in another experiment the decapitated partner was sup-

plied with an implanted prostomium from a female partner. However,
the result was the same as in the previous experiment (Fig. 3b). No
mutual influence was realized. Both individuals did not shed their
oocytes but kept them in the coelomic fluid and produced new ones.
K-jaws were not developed and spermatogenesis was also not observed.
The implanted prostomium proved to be competent for brain hormone
supply, since it prevented the effects of decapitation in its host.
Nevertheless, this prostomium was not competent for the "Paarkultur-
53
effekt". A possible explanation for these results is, that the prosto-
mium has to be in situ or must be located outside, or at the surface,
of the host.
Fig. 3 a and b. Couples of two females, with one intact and one ampu-
tated partner. (a) One partner decapitated, no mutual influence, (b)
decapitated partner with an implanted prostomium from a female donor,
no mutual influence
Two more experiments were carried out in order to test this assumption.
The posterior segments of one partner were amputated, leaving only
5 segments (Fig. 4a), while the other female was left intact. The re-
sult was a qualitatively normal "Paarkultureffekt". Whereas a change
to the male phase was obtained in 76 % of the ante rior fragments, in
54
the remaining 24% the intact female showed sex reversal with oocyte
shedding and jaw replacement. Quantitatively, the result was somewhat
abnormal. In couples which consist of two intact females of about the
same size, the number of sex reversals among the smaller and larger
females was not related to size. The following experiment yielded the
clearest result. Only the isolated prostomium of a female donor was
given to an intact female as a "partner" (Fig. 4b). In 25% of these
"couples", the prostomia induced shedding of the oocytes and spermato-
genesis in the intact female. The induction of a k-jaw, however, was
not observed. A possible action of the intact female was not observed
in the isolated prostomium. These results indicate that the prostomium
is both the transmitter and the receiver of the stimulus for the "Paar-
kul tureffekt" .
~ n = 25 couples
...........
..........
J, ..........
~i
T!!
76% 24%
~
n = 20 couples
........
........
J, ........
75%
't
0:
25%
Fig. 4 a and b. Couples of two females, with one intact and one ampu-
tated partner. (a) One partner's posterior segments are amputated -
qualitatively normal "Paarkultureffekt", (b) one partner consists only
of a female's prostomium, mutual influence, no k-jaw induced
55
C. Conclusions
There are three different events which formally characterize the

"Paarkultureffekt". When two females meet, they at first have to rec-
ognize each other as females. Thereafter, they have to determine which
of them will change sex. Finally, sex reversal has to be brought into
effect. The most likely mutual influence in a couple is based on the
action of a pheromone which is transmitted during contact of the part-
ners. The releasing site seems to be located on the prostomium, where
chernoreceptors possibly act as receivers of the pheromone. There are
definite nervous connections between these chernoreceptors and the neu-
rosecretory brain cells, which are believed to produce the ootropic
hormone. In this way, external impulses of the pheromone possibly con-
trol oogenesis, by suppression or derepression of the release of the
ootropic hormone. On that basis, mutual influence is thought to be
exerted.
The following model is based on the understanding that a pheromone

exists and acts in the mentioned way. A female does not only produce
the ootropic hormone (the existence of which is evident), but also a
pheromone. The release or production of the pheromone may becorre-
lated to oogenesis. The strength of the pheromone impulse of a female
may therefore change from time to time, according to the stage of
oogenesis. One of the two meeting females will certainly exert a
stronger pheromone impulse than the other. By this assumption, two of
the three main events characterizing the "Paarkultureffekt": the rec-
ognition of the partners and the determination of the secondary male,
can be explained. The individual with the lower release of pheromone
at that moment, will change sex - since its hormone production is
suppressed. Whenever the action of the pheromone decreases, the sec-
ondary male will restart its hormone production, proliferate oocytes
and thereafter release the pheromone itself. Thereby, it suppresses
oogenesis in the partner by inhibition of the hormone production.
This partner, now lacking the ootropic hormone, will at once prodnce
spermatozoa.
This model must still be considered as speculative. There is until now,

no direct evidence for the existence of such pheromone, although there
is some proof for its action. The replacement of the jaw and the con-
trol of oocyte shedding are not included. Further investiga~ions are
necessary on sexual differentiation in O. puerilis.
References
Bacci, G.: Sex discrimination and genic balance of Ophryotrocha puerilis,

a hermaphrodite polychaete worm. Nature (Lond.) 207, 1208-1209
(1965) .
Clark, R.B., Olive, P.J.W.: Recent advances in polychaete endocrinol-
ogy and reproductive biology. Oceanogr. Mar. Biol. Ann. Rev. 11,
175-222 (1973). -
Durchon, M.: Determinisme endocrine de la maturation sexuelle chez
les annelides polychetes. Bull. Soc. Zool. France 22, 489-509
(1970) .
Hartmann, M., von Lewinsky, G.: Untersuchungen tiber die Geschlechtsbe-
stirnmung von Ophryotrocha puerilis. III. Die stoffliche Natur der ver-
m!nnlichenden Wirkung "starker" Weibchen ("Eistoffe"). Zool. Jb.
Abt. allg. Zool. u. Physiol. &2.., 1-12 (1942).
56
Hauenschild, C.: Endokrine Beeinflussung der geschlechtlichen Entwick-

lung einiger Polychaeten. Fortschritte der Zoologie, 22. Heft 2/3.
Stuttgart: Gustav Fischer 1974.
Hauenschild, C.: Die Beteiligung endokriner Mechanismen an der ge-
schlechtlichen Entwicklung und Fortpflanzung von Polychaeten. Verh.
Dtsch. Zool. Ges. ~ (1975, in press).
Muller, H.: tiber die Sexuali tat des Polychaeten o-phryotrocha puerilis,
ihre Determination und ihren EinfluB auf Drlisentatigkeit und Kau-
apparatentwicklung. Z. Morph. Okol. Tiere 52, 1-32 (1962).
Pfannenstiel, H.-D.: Zur sexuellen Differenzierung des Borstenwurms
OphryotY'ocha puerilis. Naturwissenschaften 58, 367 (1971).
Pfannenstiel, H.-D.: Zur sexuellen Differenzierung von Ophryotrocha
puerilis (Polychaeta: Eunicidae). Mar. Biol. 20, 245-258 (1973).
Pfannenstiel, H.-D.: The role of the prostomium for the sexual differ-
entiation in the polychaetes Ophryotrocha puerilis CLAP. MECZ. 1869
and Ophryotrocha notoglandulata Pfannenstiel 1972. Pubbl. Staz. Zool.
Napoli, Suppl. (1975, in press).
Neurosecretory Phenomena during Reproduction in
Oligochaeta
H. Herlant-Meewis
A. Introduction
It is known that all organisms possess neurosecretory cells producing

various substances which play an important part in physiology.
The different neurosecretory cells of the nervous ganglia of the in-

vertebrates have been widely described and knowledge of their ultra-
structure has added vital information. However, the role of these
neurosecretions still remains puzzling, due to the fact that their
exact nature is unknown. However, it has been established that in
insects, the secretions are peptidic and that a brain extract contains
several substances (Gersch and Ude, 1960, 1967; Ishizaki and Ishikawa,
1967) .
A substance inhibiting oocyte development, has been obtained from

brain extracts of polychaetes (Cardon, 1970) but the nature of this
substance is also unknown.
Lacking biochemical knowledge, it is only possible to use experimental

methods of removals, grafts, regenerations, organ cultures, etc., in
an attempt to elucidate this problem. The selective destruction of
neurosecretory cells is only possible in those cases where different
types of cells are distinctly grouped (Girardie, 1970). However, in
most cases (and particularly in the Oligochaeta) where the different
cells are closely mixed in the brain and where a complete assortment
is found in every ganglion of the nervous chain, their removal is im-
possible. Moreover, the destruction of the nervous glanglia in Oligo-
chaeta is rapidly followed by their regeneration and the reconstitution
of the neurosecretory cells.
The study of the neurosecretory role in reproduction, raises another

difficulty. Most Oligochaeta are simultaneous hermaphrodites, both
genital tracts functioning at the same time. It is, therefore, diffi-
cult to relate the activity of certain cells to anyone of the sexual
functions. Consequently, cases of dissociation in the functioning of
the two genital tracts have been examined in this paper.
B. Results
During the last years, the distribution of the different types of

neurosecretory cells in the cerebral and ventral ganglia of adult lay-
ing Eisenia foetida has been studied. I have demonstrated by paraldehyde-
fuchsin of Gomori-Gabe stains the presence of two posterior caps, con-
sisting of "a" cells (Herlant-Meewis, 1955).
58
59
More recently (Herlant-Meewis and de Vries-Schoumacker, 1973), Adams

and Sloper's (1955) technique, modified by Herlant (1958) (oxidized
Alcian blue, haemalum, phloxine), have been used and this has confirmed
the presence of the two kinds of "a" cells, described by Aros et al.
(1965) .
Type 1 cells were rather small and angular, and their cytoplasm was
irregularly granular. They were found at the periphery of the two
posterior caps. Type 2 cells were bigger and rounder and had a more
uniform appearance. They filled the dorsal surface of the ganglia and
extended down laterally and posteriorly under the type 1 cells.
Type 2 cells were particularly numerous and large. At this time, they
appeared to be very active, but it was impossible to relate them more
specifically either to oogenesis or to spermatogenesis. Therefore,
young worms (200 - 300 mg) reaching puberty, deserve special attention.
It has been shown in Eisenia, that oogenesis started a few days after
birth, when the first pre-meiosis appeared, whereas spermatogenesis
began only at puberty, i.e. 5 - 6 weeks later. At birth, only type 1
"a" cells were found in the brain. The second type appeared when the
seminal vesicles contained only spermatogonial follicles, and the
worms weighed about 200 mg. It was only at the onset of spermatogene-
sis that these cells became abundant. However, a third type of very
large neurosecretory cell is found in young specimens. These are rather
rounded and in the staining procedure, only the cytoplasm is lightly
stained, the granules are not stained.
Neurosecretory cells have been studied in various species of Oligo-

chaeta using the electron microscope, particularly by Scharrer and
Brown (1961) and Rohlich et al. (1962) in Lumbricus terrestris; Oozaki
(1966) in Eisenia foetida; Takeuchi (1967, 1968) in Pheretia comrrrunissima;
Gersch and Ude (1967) in Enchytraeus albidus; and Gallissian and Girardie
(1972) in Eophila dollfusi. These authors all confirmed the presence of
several types of neurosecretory cells in the adult worms and have de-
scribed them: (1) during the secretory cycles (Aros et al., 1965);
(2) during different periods of the sexual life (Gallissian and Girar-
die, 1972); and during regeneration (Gersch and Ude, 1967).
I have studied the ultrastructure of the brain of young Eisenia and of

adult worms. They were fixed for 30 min in phosphate-buffered glutar-
aldehyde solution (2,5%) and post-fixed for 1 hr in osmium solution
( 1 %) •
Electron microscopy confirmed the existence of the three types of neu-

rosecretory cells in the young worm.
Type 1 cells (Fig. 1) contained electron dense-granules of various

sizes (1500 - 2700 .8.), budding off at the periphery of the Golgi com-
plex (Scharrer and Brown, 1961). These granules were irregularly dis-
persed in small clusters. Type 1 cells contained numerous frail and
ribboned mitochondria, while the ergastoplasmic cisterns and free
ribosomes were abundant. These cells closely resembled type 2 described
by Oozaki (1966) and type 1 (Gallissian and Girardie, 1972)
.... Fig. 1-3. Neurosecretory cells of Eiseniafoetida (200 mg). x 1600

Fig. 1. Type 1: Gr Granules, issued from active Golgi apparatus (G)
and maturating in the cytoplasm. N Nucleus; Ly Lysosome
Fig. 2. Type 2: Irregular granules. E Peripheral ergastoplasm; N Nucleus
Fig. 3. Type 3: Weak electron dense granules (arrows) issued from
ergastoplasm. C Smooth or partially granular cisternae
60
Type 2 ceils (Fig. 2) were distinguished by the presence of smaller

irregular granules. Such granules were often oblong, the largest hav-
ing a length of about 1500 A. They also appeared at the level of the
Golgi complex, and the ergastoplasmic lamellae often filled the peri-
phery of the cell. The mitochondria were less oblong than in type 1
cells. The distinctive features of these cells corresponded exactly
to type 2 described- by Gallissian and Gerardie, and perhaps to type 3
described by Oosaki.
In the adult, these two types of cells were abundant and closely mixed,
and filled with secretory granules which were not always easy to dis-
tinguish. It seems, that when the granules are mature and accumulating
in the cell, they acquire a more uniform size and volume (Takeuchi,
1967). However, it was easier to distinguish these cells in the young
specimens of 200 mg, where there were mostly type 1 cells.
Type 3 cells (Fig. 3) have only been observed up to now in very young
worms. They were very big and contained large droplets of weak electron
density (about 2000 ~). The cytoplasm showed important areas of ergasto-
plasm, relatively few mitochondria and some small dyctiosomes. The
quantity of ribosomes lining the cisterns was very variable: they may
be quite numerous on both sides of them; becoming scarce or non-existent
on one face; or leave some membranes perfectly smooth. In the areas
where the ribosomes were most numerous, there were distended cisterns
where dense material accumulated. This might be a site of formation of
granules which are directly released, without passage of the secretion
through the Golgi complex. These granules seemed to reach their defini-
tive size inside the ergastoplasm from which they emerged, surrounded
by a membrane and crowned with ribosomes.
Takeuchi (1967, 1968) seems to have discovered similar cells in Pheretima

and he proposed the same explanation as to the origin of the granules.
In an Eisenia individual of 300 mg, the cytoplasm of these cells was
filled with granules, while large areas of ergastoplasm had disappeared.
However, some ergastoplasmic cisterns may be seen isolated among the
granules. Numerous lysosomes appeared, within which the granules were
destroyed. It appeared that such penomena, when increasing, led to the
complete lysis of these cells which then disappeared from the brain.
At that moment, cells in the course of destruction were frequently
observed along the neurilemma. These might perhaps, correspond to the
last stage of type 3 cells.
Type 3 cells may produce a metabolic hormone, having a role in the

young worm's growth up to puberty. Oosaki (1966) described these cells
under the name of type 1, although he did not mention the age of the
worms he observed. They contained homogenous granules, which were
moderately electrorr-dense. and measured 2000 - 3000 ~.
It is possible that the "a" cells described by Gersch and Ude (1967)
might be connected with the type 3 cells. The granules of these cells
measured 1100 - 2500 ~, and part of the cytoplasm was filled by an
important ergastoplasm very rich in ribosomes. The Golgi formations
were small and the lysosomes often numerous. The author studied the
development of these cells during regeneration, where they seemed to
playa significant role. During the formation of the secretion, the
ergastoplasmic cisterns seemed distended and the dictyosomes were
small. However, the secretion did appear in their cisterns. The author
concluded that the secreting droplets were of Golgian origin. It is
not impossible that two types of secretion might exist in this cell:
the neurosecretion originating in the ergastoplasm, as well as neuro-
transmitters released by the Golgi apparatus as small vesicles. This
61
hypothesis may then help in explaining Takeuchi's figures. To substan-

tiate this interpretation, it will be necessary to study younger worms,
where type 3 cells begin to develop their secretion.
In Eisenia the ordinary neurones may, in some cases, be confused with

neurosecretory cells. In the cerebral ganglia of Oligochaeta, these
cells are considered as associating neurons. They produce catechol-
amines or serotonin "neurotransmitters", that have been studied by
Myhrberg (1972). Of varying size, mostly unstained by the histological
method outlined, they may nevertheless show light green stained areas,
or be more or less phloxinophilic.
In electron microscopy, such neurons also showed similar characteris-

tics. They may contain any number of the small umbilicated granules
originating at the level of the Golgi apparatus, which possibly cor-
respond to the catecholamines (de Robertis, 1962). These granules are
stained light green with Alcian blue after oxidation. At the same time,
some neurons contain such granules and light vesicles. Large neurons
loaded with mitochondria were also observed with the electron micro-
scope. These might correspond to the "M" cells described by Gersch
and Ude. The neurophils of the young worms contained fibers lined with
the various granules, corresponding to those described in the three
types of neurosecretory cells and in the ordinary neurons.
It was thought that the type 2 cells, appearing at the moment of pu-
berty and thereafter strongly activated, might play a role in sperma-
togenesis. To test this hypothesis, the brain of worms bearing anoma-
lies at the level of the male genital tract, were examined.
Recently, the breeding ground, composed of humus and rotting leaves

was replaced by manure of known origin. The adult worms raised on this
ground lost weight in some generations (down to 450 - 500 mg), whereas
in those raised on humus, the weight reached 600 - 700 mg. These worms
still retained their external sexual characters, mated and deposited
cocoons. However, a reduction in the number of cocoons and of the
young worms born from these was observed. Therefore, their fecundity
seemed to have lessened.
Anomalies in spermatogenesis were observed in the genital tracts of

these worms, while oogenesis was only slackened. The male genital
ducts and particularly, the funnels of the sperm ducts, became smaller
- while the ovisacs and oviducts remained normal. The seminal vesicles
showed many spermatocytes blocked at meiosis. There were a few cells
continuing spermatogenesis and, in extreme cases, the seminal vesicles
only contained spermatognonia and spermatozoa. This was similar to the
aspect described in fasting worms, or after removal of their brains
(Herlant-Meewis, 1966b), but in these cases the female tractus remains
working. It has not yet been possible to explain the reason for the
blocking of meiosis - it may be a result of the feeding medium. This
problem is being investigated at present.
The genital tract of 20 growing worms of the same origin were examined
to determine the moment when the anomaly appeared in the worm. In the
young worms, prior to puberty, the male and female genital tracts de-
veloped normally, while neither the testicles nor the ovary showed any
modification. At puberty, when the worms weighed about 250 to 300 mg,
the puberty tubercles appeared, but only after some delay. What might
be termed a "meiotic crisis" started in the seminal vesicles of these
worms. Indeed, a great number of spermatocytes started meiosis simul-
taneously. Some of them continued their spermatogenetic development,
while most of them appeared to be blocked and later degenerated. The
62
testes continued to free spermatogonial follices. The ovary had a

normal development. Therefore, the influence the food had on young
worms at the moment of puberty was to stop meiosis. In the brain,
type 2 cells were empty and all were less active, as spermatogenesis
became slower.
Another case of dissociation in the activity of the two genital tracts

was investigated in an oligochaete worm, Alma errrini, collected in Zaire
and showing a functional successive hermaphroditism (Khalaf-el-Duweini,
1950, 1953). When breeding for the first time the worm acts as "male".
At the moment it possessed well developed claspers and spermatogenesis
was very active. However, it mates with an older worm, at the female
stage - possessing reduced copulatory appendages and a prominent cli-
tellum. This older worm had maturing eggs and its male tract was re-
duced.
When the examples had well-developed claspers, the testes were active
and the seminal vesicles contained all the stages of spermatogenesis.
However, some mature oocytes were already found in the ovaries, but
the oviducts were not yet open exteriorally. At this stage, the brain
contained numerous type 1 and type 2 cells. The latter were large and
very active.
Such worms were kept fasting in water for a month. During that period,
the copulating organs were reduced and disappeared completely. These
specimens underwent a complete degeneration of the male genital tract,
the testes became much smaller and all the cells of the male tissue
retrogressed. The ovary, however, was unchanged. In the brain, the
type 2 cells became progressively empty, much smaller in size and the
neurosecretion accumulated in the neurophil.
These first observations support the hypothesis that the type 2 cells
may playa role in spermatogenesis.
References
Adams, C.W.J., Sloper, J.C.: Technique for demonstrating neurosecretory

material in the human hypothalamus. Lancet 1, 651-652 (1955).
Aros, B., Vigh, B., Teichman, I.: The structure of the neurosecretory
system of the earthworm. Symp. Biol. Hung. 5, 303-316 (1965).
Cardon, C.: Procede de fractionnement des gangIions cerebroides de
Nereis diversicolor Miiller en vue de l'isolement de l' hormone inhibi-
trice de sexualisation. Bull. Soc. Zool. France 95, 543-549 (1970)
(1970) . -
De Robertis, E.: Ultrastructure and function in some neurosecretory
systems. Proc. Third Internat. Conf. on Neurosecretion. Mem. Soc.
Endocr. 11., 3-20 (1962).
Ehinger, B., Myhrberg, H.E.: Neuronal localization of Dopamine, Nor-
adrenaline, and 5-Hydroxytryptamine in the central and peripheral
nervous system of Lumbricus terrestris (L). Histochim. 28, 265-275
(1971). -
Gallissian, A., Girardie, J.: Etude histologique et ultrastructurale
des cullules nerveuses du cerveau d' Eophila doUfusi Tetry (Lumbri-
cide). C.R. Acad. Sci. Paris 275, 79-82 (1972).
Gersch, M., Ude, J.: Die Isolierung neurohormonaler Faktoren aus dem
Nervensystem der Kiichenschabe Periplaneta americana. Z. Naturforsch.
12, 319-323 (1960).
63
Gersch, M., Ude, J.: Elektronenmikroskopische Untersuchungen zur Dyna-

mik neurosekretorischer Zellen von Enchytraeus (Oligochaeta). Z.
Zellforsch. 81, 374-389 (1967).
Girardie, A.: Neurosecretion cerebrale chez les Acridiens. Bull. Soc.
Zool. France 95, 783-802 (1970).
Herlant, M.: L'hypophyse et le systeme hypothalarno-hypophysaire du
pangolin (Manis tricuspis Rof. et Manis tetradactyZa L.). Arch. Anat.
Micr. Morph. Exp. 47, 1-23 (1958).
Herlant-Meewis, H.: Neurosecretion chez les Oligochetes. Bull. Acad.
R. Belg. V 11, 500-508 (1955).
Herlant-Meewis, H.: Les cellules neurosecretrices de la chaine nerveuse
d' Eisenia !,oetida. Z. Zellforsch. 69, 319-325 (1966a).
Herlant-Meewis, H.: Evolution de l'appareil genital d'Eisenia foetida
au cours du jeune, de la regeneration posterieure et a la suite de
l'ablation de ganglions nerveux. Ann. Soc. R. Zool. Belg. ~, 189-
240 (1966b).
Herlang-Meewis, H., de Vries-Schournacker, H.: Les cellules neurosecre-
trices des ganglions cerebroides d' Eisenia foetida. Gen. Compo End.
20 (1973, in press).
Ishizaki, H., Ichikawa, M.: Purification of the brain hormone of the
silkworm Borribyx mori. Blol. Bull. 133, 355-368 (1967).
Khalaf el-Duweini, A.: On the copulation in Alma niZotica Grube, with a
description of the spermathecae in the species. Zool. Soc. Egypt.
Bull. 9 29-39 (1950).
Khalaf el=Duweini, A.: On the internal genital organs of AZma niZotica
Grube, with a note on the septal pouches in the species. Egypt.
Acad. Sci. 9, 102-109 (1953).
Myhrberg, H.E.: Ultrastructural localization of monoarnines in the
central nervous system of Lurribricus terrestris (L) with remarks on
neurosecretory vesicles. Z. Zellforsch. 126, 348-362 (1972).
Oozaki, T.: Observations on the ultrastructure of nerve cells in the
brain of the earthworm Eisenia foetida, with special reference to
neurosecretion. Z. Zellforsch. 72, 534-542 (1966).
Rohlich, P., Aros, B., Vigh, B.: Elektronenmikroskopische Untersuchung
der Neurosekretion im Cerebralganglion des Regenwurrnes (Lurribricus
terrestris). Z. Zellforsch. 58, 524-545 (1962).
Scharrer, E., Brown, S.: Neurosecretion XII. The formation of neuro-
secretory granules in the earthworm Lurribricus terrestris L. Z. Zell-
forsch. 54, 530-540 (1960).
Takeuchi, N:7 On the structure of neurosecretory cells in the cerebral
ganglion of earthworm, Pheretima conmunissima. Sci. Rep. Tohoku Univ.
Ser. IV (Biol.) 33, 429-439 (1967).
Takeuchi, N.: On the-relationship between neurosecretory granules and
intracellular mernbraneous cell organelles in the neurosecretory
cells of cerebral ganglion of Megascolecid earthworm. Sci. Rep.
Tohoku Univ. Ser. IV (Biol.) li, 1-11 (1968).
Study of Sex Control of Gametogenesis by Organ Culture
in the Oligochaete Annelid Eisenia foetida f. typica Say.
M. Lattaud
A. Introduction
Ovarian autodifferentiation was primarily shown in amphipod Crustacea

by Charniaux-Cotton (1957). The same phenomenon was also described in
other Invertebrates, hermaphrodites as well as gonochorists, by Char-
niaux-Cotton (1965) and Guyard (1971), in isopod crustaceans, Hydra,
insects, gastropods, myriapods and nemertines.
An hermaphrodite with separate gonads, Eisenia foetida (Lumbricidae), has

been investigated in this paper. Previous fasting experiments, followed
by refeeding (Lattaud, 1971), confirmed the results obtained by Herlant-
Meewis (1962, 1966) in Eiseniafoetida. At the onset of renewed activity
after feeding, some testis gonia developed into oocytes, whereas ova-
ries did not reveal any spermatogenesis. These results indicated an
ovarian autodifferentiation. At the time of renewed activity, the
testes would be devoid of any androgenic substance, thus permitting
the evolution of gonia into oocytes by autodifferentiation. Neurosecre-
tion is probably necessary for the synthesis of this substance by tis-
sues surrounding the testes.
To check these hypotheses, the sex control of gametogenesis in sexu-

ally active Eisenia foetida in organ cultures has been studied. Testes
and ovaries were cultivated separately, or associated (Lattaud, 1973)
To study the influence of neurosecretion, associations with the CNS
were attempted.
Relexans (1973) cultivated in vitro fragments of the body wall carrying

at the most either the four presumed testes or the two presumed ovaries,
removed from new-born Eisenia foetida. Development of the explants contin-
ued until gonia appeared. Nevertheless, two testes did show premeiotic
oocytes. According to Relexans (1971), gonial sex can be distinguished
by ultrastructural characteristics: "Gonads of both presumed sexes
could undergo inversion". That of the testes is interpreted as a "re-
turn to autodifferentiation of gonia having escaped from the action of
an androgenic factor due to experimental conditions". To explain the
presence of spermatogonia in a few ovaries, Relexans believed that,
in tissue culture, ovarian gonocytes became sensitive to the androgenic
factor - probably also present in female explants (sexual inductors).
B. Material and Methods
The culture mediums used were those of Wolff and Haffen (1953) and
Sengel (1960), the latter used for the culture of regenerating blas-
tema in planarians. Prior to setting up the culture, donors were iso-
lated for 48 hrs in cotton-wool soaked with sterile water to which had
been added a bactericide (furadoine). As anesthetics seemed to have
65
an unfavourable influence on explant survival, earthworms were numbed

by cold at -10°C. Ovaries or testes of Eisenia foetida were removed with
a fraction of the dissepiment. The body wall close to the gonads was
never cultivated. For each cultivated fragment, a symmetrical gonad
was fixed for control purposes. In associations of 4 to 6 testes,
2 earthworms were necessary while a single testis in each animal served
as a control. The CNS, except the posterior part of the nervous cord,
(i.e. the cerebroid ganglia, suboesophageal ganglia, circum-oesophageal
nerve ring and the ventral cord cut at the clitellum level) was culti-
vated. For the sake of simplicity, this shall now be refered to as the
CNS. Association of the explant on the culture medium was achieved by
putting the organs side. by side. In cultures associating the gonads
with the CNS, the latter surrounded the former. Cultures lasted from
6 to 21 days.
C. Results
Ovaries and testes fixed at time of culture setting. Examination of

histological slides showed oogenesis taking place in the ovaries.
From the proximal to the distal region of the gonad, protogonia, sec-
ondary gonia, cells in meiotic prophase and developing oocytes could
be seen. Testes contained primary and secondary gonia, the latter
grouped in spermatogonial follicles. There was neither male meiosis
nor spermiogenesis, though these can occur in Eisenia foetida after a
6 weeks fast (Lattaud, 1971).
Ovaries cultivated separately. Histological study of the ovaries did

not reveal any oogenetic disturbances and development of oocyte took
place normally. The cell diameters were approximately the same as in
the controls.
Testes cultivated separately. The observations are summarized in

Table 1.
Table 1. Culture of testes
Duration of Number of Inversion Without Cytolysis

culture cultivated inversion
(days) testes
6 to 11 17 6 10
12 to 21 29 7 17 5
In 6 testes cultivated from 6 to 11 days, groups of cells in meiotic

prophase with visible chromosomes were observed. This is characteristic
of female development of the gonads (Omodeo, 1952). These cells were
situated between the area of protogonia and spermatogonial follicles.
No developing oocytes were discernable.
Between days 12 to 21 in culture 7 testes were inverted while two of

them showed a complete sexual inversion (Fig. lA). No more spermato-
gonial follicles were seen in these two testes, even though all the
characteristic stages of oogenesis were present - gonia, oocytes in
meiotic prophase with visible chromosomes (Fig. lB) and developing
oocytes (Fig. lC). Distal cells did not appear to have completed their
oogenesis. In the other 5 testes, oocytes in early meiotic prophase
66
Fig. 1. (A,B) Testis cultured separately, inverted into ovary after 12

days of culture. All characteristic stages of oogenesis are clearly
visible in A. The region of oocytes in meiotic prophase can be seen
in B. ee) Section of testis cultivated separately for 13 days exhibit-
ing young developing oocytes. (D) Testis cultured separately, inverted
into ovary after 13 days of culture. Note the area of developing oocytes
between female cells in meiosis prophase (with visible chromosomes)
and spermatogonial follicles.
Sp.F. spermatogonial follicles; GI primary gonia or protogonia; 0 ovary;
00 oocyte; P oocyte at pachytene stage; R I:(. female region; Rd'male region;
eNS central nervous system; T testis; Z oocyte at zygotene stage
67
Fig. 1. (E) Section of an association of two testes and one ovary. After
11 days in culture, there was no sign of gonadal sexual inversion.
(F) Two testes associated with one ovary. After 12 days in culture,
a developing oocyte can be seen on the testis section between proto-
gonia and spermatogonial follicles. (G) Association of two eNS with
one ovary. Histological study does not show any oogenesis disturbance
after 13 days in culture. (H, I) Association of two eNS and one ovary
ana three testes cultured for 21 days. Only the ovary, one testis and
one eNS are visible on H. There is no testicular sexual inversion.
On the other hand, the ovary inverted into testis (I). Note spermato-
gonial follicles situated laterally in the region of developing oocytes
and more mature ones. (For explanation of abbreviations see opposite page)
68
(in the region adjacent to the protogonia and young developing oocytes)
were observed. These oocytes occupied an area between female cells in
.meiotic prophase, with visible chromosomes and spermatogonial follicles
(Fig. 1D). The oocytes could also form a cord spreading between the
spermatogonial follicles (up to 9 developing oocytes were counted) .
In total, 46 testes were cultured separately. Taking into account the
6 cases of cytolysis, the inversion reached 35%.
Testis associated with ovaries. To demonstrate an eventual endocrine

action of testis (or of dissepiment) on ovarian gonocytes, I associated
in the culture one ovary with 1, 2, 3 or 4 testes. The observations are
summarized in Table 2.
Table 2. Culture of testes associated with ovaries

Duration Cultured Cyto lyzed Non- Testes
of culture associations associations inverted inverted Non-
(days) ovaries inverted
{: :: ::
testis 7 2 5 I{Oo.m.p. 4
or
2 testes 8 7 2{00.m.p. ,00 5
c:: :
Oo.m.P.
1 ovary +
r·
3 testes 5
° 5 2{00.m.p.
Oo.m.P.
3
or
4 testes 15 2 13 9
4 Oo.m.P.
op
Oo.m.P.,Oo
Oo.m.P. ,00
00. developing oocytes; Oo.m.P. oocytes in early meiosis prophase with visible
chromosomes.
In these associations, there was never any masculiniz~tion of the

ovaries (Fig. 1E). Oogenesis occurred normally. From the proximal to
the distal part of the gonad protogonia, secondary gonia, female cells
in early meiotic prophase and developing oocytes were observed. There-
fore, the testis had no endocrine action on the ovarian gonocytes.
Among the associated gonads, 9 testes were inverted. A complete sexual
inversion of the gonads was never obtained. In 6 cases, ovotestes ex-
hibited intra-gonadal female pre-meioses (amodeo, 1952) which occurred
between the protogonia and spermatogonial follicle areas. There were
no developing oocytes. In 3 cases, not only female cells in pre-meiosis
but also young oocytes surrounded by a thin cytoplasmic ring were ob-
served. Up to 8 developing oocytes, forming a cellular cord and sur-
rounded by spermatogonial follicles, were counted. The average oocyte
diameter never exceeded 45 ~m (Fig. 1F). These appeared only after 12
to 21 days of culture. In total, 35 associations were made. Taking
into account the 5 cases of cytolysis, 29% of the testes inverted into
ovaries - the inversion being more pronounced after 11 days in culture.
This percentage of sexual inversion was comparable to that obtained
from separate testes-cultures. Therefore, the ovary had no influence
on the testes.
Testes associated with the CNS. To show the possible neurosecretory

intluence on testicular gonocytes, 14 cultures were set up, asso-
ciating 1, 2 or 3 CNS to 1 or 2 testes. After 6 to 21 days of culture,
69
no testis showed any sexual inversion. Gonads consisted of primary

gonia undergoing gonial mitoses and of secondary gonia forming sperma-
togonial follicles. There was neither male meiosis nor spermiogenesis.
By comparison, the number of gonial mitoses did not increase with those
of the control testes. Neurosecretion had no mitogenic effect, while
the CNS helped in maintaining male gametogenesis. It was not possible
to say if this was due to the CNS releasing a masculinizing substance,
or if the neurosecretion induced the synthesis of an androgenic sub-
stance by the testicular tissues. To solve this problem, two cultiva-
tions were made: CNS associated with ovaries and CNS associated with
both ovaries and testes.
Ovaries associated with CNS. In order to study the eventual action of

the neurosecretion on gonocyte differentiation, 1, 2 or 3 CNS with
a single ovary were associated in the culture medium. Twelve associa-
tions were set up. Histological study of the ovaries after 6 to 21 days
in culture, did not show any evidence of interruption in oogenesis.
Oocyte development occurred normally (Fig. 1G). The CNS had no influ-
ence on the developing oocytes and there were no inversions. Therefore,
the CNS had no inhibiting influence on oogenesis and did not release
any masculinizing substance.
CNS associated with ovaries and testes. From previous experiments, it

was hypothesized that the synthesis of an androgenic substance by tes-
ticular tissues would be under CNS control. Thus, in the testes culture
medium one ovary and a few CNS were associated (Fig. 1H). Thirty two
associations were made, with only 4 cases of cytolysis occurring.
There was not a single case of testicular sexual inversion in the
remaining 28 associations. This was a very significant result if it
is considered that 114 testes were cultivated in association. More-
over, the number of gonial mitoses was almost equal to those of the
control testes. However, of 28 ovaries, 5 did invert. In three cases,
the ovotestes contained a few young spermatogonial follicles. These
were located between the gonia and female cells and showed early meio-
sis prophase with the chromosomes visible. In the remaining two, the
inverted gonad contained a male islet comprised of numerous young
spermatogonial follicles. This male area was situated laterally, in
the region of young developing oocytes and older oocytes (Fig. 11).
Ovarian sexual inversion reached 18% but occurred only after 15 days
of culture. At least 3 testes should therefore, be cultured for this
phenomenon to occur. The number of CNS (1, 2 or 3) in the medium did
not seem to have any influence.
D. Summary and Conclusion
In vitro cultures of associated or separate gonads, showed that, in the

absence of any hormonal substance, gonia of all ovaries and of a number
of testes (in about one third of the experiments) spontaneously devel-
oped oogenesis (ovarian autodifferentiation). This autodifferentiation
could explain the transitory oogenesis observed in Eisenia foetida testes
during fasting experiments (Herlant-Meewis, 1962, 1966; Schoumaker,
1969; Relexans, 1970; Lattaud, 1971), or after the removal of cerebroid
ganglia (Herlant-Meewis, 1966). Spermatogenesis requires an androgenic
substance which is not supplied by the neurosecretion. However, when
an ovary was cultivated in association with CNS, no masculinization
occurred. Moreover, the neurosecretion had no inhibitory action on
oogenesis, as it remained unperturbed. Sexual inversion of the ovary
was only observed in CNS-ovary-testes associations.
70
Under the present experimental conditions, the percentage of inversion

was small, but significant (18%). In the presence of the neurosecre-
tion, the testicular tissues produced an androgenic substance - ensur-
ing control of sex and their spermatogenetic activity. Neurosecretion
allowed the synthesis of this substance in 100% of the testes in cul-
ture. The low percentage of sexual inversion of ovaries into testis
(CNS-ovary-testes associations) could be due to a poor diffusion of
the androgenic substance through the culture medium. However, a de-
crease in the synthesis of the substance following an insufficient
neurosecretion (in culture conditions), could also be considered.
Recently, Relexans (1974) experimented on local removal of the nerve

of the genital area, as well as on inhibition of monoaminergic neurons
in newborn Eisenia foetida, with gonads already in situ. He noted sexual
inversion of testes into ovaries. It appears that the results of these
experiments concurred with the present conclusions: synthesis of an
androgenic substance by testicular tissues under the neurosecretory
control.
References
Charniaux-Cotton, H.: Croissance, regeneration et determinisme endo-

crinien des characteres sexuels d' Orchestia g=arella Pallas (Cru-
stace Amphipode). Ann. Sci. Nat. 19, 411-560 (1957).
Charniaux-Cotton, H.: Hormonal control of sex differentiation in In-
vertebrates. In: Organogenesis (eds. De Hann, R., Ursprung, H.),
pp. 701-740. New York: Holt, Rinhart and Winston 1965.
Guyard, A.: Etude de la differenciation de l'ovotestis et des facteurs
controlant l'orientation sexuelle des gonocytes de 1 'Escargot Helix
aspersa MUller. These de Doctorat es Sciences Naturelles, Besan90n
1971 .
Herlant-Meewis, H. Influence de la nutrition sur la reproduction Chez
Eisenia foetida. C.R. Acad. Sci. Paris 255, 2187-2188 (1962).
Herlant-Meewis, W.: Evolution de l' appareil genital d' Eisenia foetida
au cours du jeune, de la regeneration posterieure et a la suite de
l'ablation de ganglions nerveux. Ann. Soc. Roy. Zool. Belg. 96,
189-240 (1966). --
Lattaud, C.: Evolution des testicules et des vesicules seminales chez
les Oligochetes Lumbricidae Eisenia foetida et Lwribricius terrestris, au
cours du jeune, suivi ou non de re-nutrition. C.R. Acad. Sci. Paris
272, 319-322 (1971).
Lattaud, C.: Autodifferenciation ovarienne chez l'Annelide Oligochete
Eisenia foetida f. typica Sav. demon tree au moyen de la cul ture organo-
typique. C.R. Acad. Sci. Paris 276, 1737-1740 (1973).
Omodeo, P.: Cariologia dei Lumbricidae. Caryologia 4, 173-275 (1952)
Relexans, J.-C.: Action du jeune, applique des l'ecfosion, sur le
developpement des gonades d' Eisenia foetida (Annelide Oligochete,
Lombricide). C.R. Acad. Sci. Paris 271, 1530-1533 (1970).
Relexans, J.-C.: Ultrastructure des gonocytes primordiaux et des gonies
chez l' hermaphrodi te Eisenia foetida (Oligochete Lombricide). C. R.
Acad. Sci. Paris 272, 2916-2919 (1971).
Relexans, J.-C.: Mise en evidence, en culture organotypique, de la
bipotentialite des region male et femelle chez les nouveau-nes
Eisenia foetida (Oligochete Lombricide). C.R. Acad. Sci. Paris 277,
1349-1352 (1973). -
Relexans, J.-C.: Recherche experimentale des facteurs de la differen-
cia tion sexuelle chez l' hermaphrodi te simul tane Eisenia foetida Sav.
(Oligochete Lombricide). These de Doctorat es Sciences Naturelles,
Bordeaux 1974.
71
Schoumaker, H.: Quelques anomalies genitales chez Eisenia foetida (Sav.),

(Oligochete terricole Lombricide). Ann. Soc. Roy. Zool. Belg. 21,
169-191 (1969).
Sengel, C.: Culture in vitro de blastemes de regeneration de Planaires.
Wolff, E., Haffen, K., Kieny, M., Wolff, Em.: Essais de cultures in
vitro d'organes embryonnaires en milieux synthetiques. J. Embryol.
expo Morph. 1, 55-84 (1953).
Factors of Primary Sexual Differentiation in the Simultaneous
Hermaphrodite Eisenia foetida (Oligochaeta: Lumbricidae)
J. C. Relexans
A. Introduction
The hypothesis of ovarian self-differentiation expressed by Charniaux-

Cotton (1957, 1959), following. her research on the amphipod Orchestia
gammarella, stipulates that male differentiation of the germ-cells de-
pends on an androgenic factor (hormone), while female differentiation
occurs spontaneously when the androgenic factor is absent or insuffi-
cient. This hypothesis has been checked with some gonochoristic and
hermaphroditic invertebrates.
The determination of primary sexual differentiation among the Oligo-

chaeta (which are hermaphroditic with separated gonads), has been
little studied. It is only known that, as a consequence of an inter-
ruption in gametogenesis resulting from a prolonged fasting period or
from the removal of cephalic ganglia (Herlant-Meewis, 1967), a few
testicular germ-cells can deviate towards oogenesis while the ovaries
remain normal. Some authors (Brien, 1966; Charniaux-Cotton, 1965, 1972)
have interpreted these results as presuming that experimental condi-
tions could weaken an androgenic factor and allow oogenesis through
self-differentiation. Recently, Lattaud (1973) has attested that cul-
tivated adult ovaries of Eisenia kept their normal differentiation in
vitro while explanted testes could turn into ovotestes.
However, the experiments demonstrated here show that in new-born
Eisenia foetida, transplanted or explanted gonads of both presumptive
sexes can be inverted and reveal a sexualizing influence, peculiar to
either genital territory. This paper deals with the factors interven-
ing in the stability of sexual differentiation.
Eisenia foetida has two pairs of testes in segments 10-11 and one pair
of ovaries in segment 13. Identical primary germ-cells (protogonia)
are found at the base of each gonad (Chapron and Relexans, 1971),
while differentiated sexual cells are found at the distal end. Peri-
toneal intra-gonadal cells insert long and thin prolongations between
the sexual cells of gonads of both sexes (Relexans, 1974).
In new-born worms, the gonads are formed by mesodermic swellings con-

taining only a few protogonia and they are sexually undifferentiated.
Gonia appear a few days after hatching. The sexes of these gonia can
be detected with an electron microscope (Chapron and Relexans, 1971),
but not with an optical microscope.
After approximately 10 days, oogonia from pre-meiotic oocytes, and

auxocytes towards the third week. In the testes, the spermatogonia
73
develop into typical spermatic follicles around the second or third

week. Meiosis occurs only after the liberation of spermatic follicles
into the coelomic cavity and seminal vesicles. Therefore, in optical
microscopy, the observation of meiotic figures in ovaries constitutes
the first diagnostic character between ovaries and testes.
The excision of undifferentiated gonads for transplantation or explan-

tation purposes, has to be made before the appearance of the first
gonia, i.e. early after hatching. At this stage, the gonads are too
small to be excised separately. Therefore, in new-born donors, gonads
have been taken together with the adjacent body wall (tegument and
musculature, peritoneum, nervous system). In three-week or older worms,
the gonads can be removed selectively.
C. Results
I. Sexualizing Properties of Genital Territories
1. Demonstration by Transplantation of Undifferentiated Gonads
Gonads of both presumptive sexes, surrounded by the adjacent body wall

and taken from new-born donors, have been transplanted into the ventral
region of the hosts of the same sex and age (reference transplantation),
or into the other sex (heterosexual transplantation), or outside the
ventral genital region (levels 5-6 and 17-18 on the ventral side, lev-
els 5-6, 10-11, 13, 17-18 on the dorsal side). Within 4 to 8 weeks
after the operation, the transplanted gonads were studied from histo-
logical sections and the frequencies and characters of the sexual in-
versions, for each level of transplantation, were noted.
Transplanted gonads can be differentiated according to their presump-

tive sexes (normal gonads) or they can contain a more or less developed
heterodox line (without any conformity with the presumptive sex), the
remaining portion of the gonad being normal (ovotestis); complete in-
version is rare.
In some cases, the heterodox line of these ovotestes contains a few

cells having reached the same stage of development: (1) pre-meiosis
or auxocytes (Fig. 1) in the ovotestis developed from the presumptive
testes; (2) spermatic follicles at the same stage in the ovotestis
developed from presumptive ovaries. In such ovotestes the inversion
phase must have been short and they will be referred to as "short-
lived inversions". In other cases, the well-developed heterodox line
shows various stages of development among its cells: (1) the ovotestes,
originating from presumptive testes, show more or less important female
regions with pre-meiosis and auxocytes; the inversion can be total;
(2) the ovotestes, developed from presumptive ovaries, have male re-
gions with follicles of varying ages (Fig. 2). Complete inversion of
the ovaries has not been observed. This inversion phase must have
taken longer and will be referred to as "lasting inversions".
The frequency of inversions, i.e. the percentage of inverted gonads

in relation to the number of grafted gonads differentiated at the end
of the experiment, varies according to the place of the transplantation
as well as to their short-lived or lasting characters (Relexans, 1973a)
The frequencies and characters of inversion in transplantation exper-

iments are summarized in Table 1.
74
Table 1. Frequencies and characters of inversions observed during

transplantation experiments
Reference Heterosexual Extra-genital

transplantations transplantations transplantations
if Grafts 7% short-lived 14 % short-lived 12 % short-lived

inversions + lasting inversions
inversions
'i? Grafts 30% short-lived 75 % short-lived 30% short-lived
+ lasting + lasting + lasting
inversions inversions inversions
""'=,,---____ P. g . cell.
'o--- P. g. cell.
<F.;--- Y. foil.
~~:;:-_ Aux.
Typ. foil.
<--_ _ Aux.
Fig. l , 100 JJm Fig. 2

Fig. 1. Ovotestis issued from a presumptive testis (short-lived in-
version)
Fig. 2. Ovotestis issued from a presumptive ovary (lasting inversion)
Aux. auxocytes; M.foll. meiotic follicles; p.g.cell. primary germ-
cells; Prem .ooc. pre-meiotic oocytes; Typ.foll. typical follicles;
Y .fo ll . young follicles
In the heterosexual transplantations the frequencies observed were

twice as high as those during the reference transplantations. Further-
more, in male grafts, lasting inversions have been observed. In extra-
genital transplantations, the inversions recall those observed during
the reference transplantations because of their characters (male grafts)
or of their frequencies (female grafts). In general the inversion fre-
quencies in male grafts were greater on the dorsal side than on the
ventral side, while the opposite phenomenon was observed with female
grafts.
75
The inversions observed in reference transplantations can only be at-

tributed to the physiological disturbances that occur in the graft
because of the transplantation itself.
In reference transplantations of male grafts, the inversions (because

of their characters and frequencies) recall those observed after a long
fasting period by Relexans (1970). Therefore, it can be postulated that
the grafts are in a state of temporary trophic deficiency before their
complete healing in the host. However, these inversions do not question
the hypothesis of ovarian self-differentiation, as the trophic distur-
bances could have temporarily weakened an androgenic factor and thus
allowed the female self-differentiation of some germ-cells.
In reference transplantations of female grafts, the inversions can be

attributed to the same causes as those seen previously, but they are
more difficult to explain according to the hypothesis of female self-
differentiation. This would indicate the presence of an androgenic
factor in the transplanted female region, with a sufficient concentra-
tion to cause inversion. The male factor, normally inactive in the
female region of a non-operated worm, would become active after the
transplantation of this same region. This would imply that, under nor-
mal conditions, female differentiation would not be a consequence of
an insufficient concentration of an androgenic hormone in the female
region, but a consequence of a kind of "ovarian insensibility" to the
androgenic action. The trophic disturbances, consecutive to the trans-
plantation, would reduce this insensibility. Therefore, the ovarian
insensibility to the androgenic action (which varies according to the
trophic physiological conditions imposed on the female region) appears
to be a specific activity which inhibits the androgenic effect. This
does not fit with the ovarian self-differentiation hypothesis, which
only explains the sexual differentiation by a variation of the andro-
genic hormone concentration.
Therefore, another explanation is proposed. Within each graft there

could be male and female somatic-dependent potentialities, one of them
normally dominating the other one. The trophic disturbances caused by
the transplantation, would then weaken the dominating potentiality and
momentarily allow the expression of the dominated potentiality. Under
normal conditions, the male potentiality would be dominant in segments
10-11 and the female potentiality in segment 13.
The characters and/or the frequencies of the inversions observed in

heterosexual transplantations show that the genital regions exert a
specific sexualizing influence on grafts of the opposite sex. According
to my hypothesis, this influence in the graft could emphasize the re-
versal of dominance of its potentialities, already started by the trans-
plantation itself. This influence recalls the embryonic induction.
Consequently, under normal conditions, the dominance of one potential-

ity over the other in the male or female region, could determine the
synthesis of a male or female "inductor" acting on the bipotential
germ-cells. However, in the cases of inversion, the reversal of the
normal dominance would determine the synthesis of the other inductor,
without any conformity with the presumptive sex.
Outside the ventral genital regions, the inversions recall those ob-
served in the reference transplantations. They seem to be provoked only
by the trophic disturbances inherent to the transplantation itself, in
relation to the physiological context peculiar to each level. Sexu-
alizing inductors do not seem to be present outside the ventral genital
regions, but variable conditions in the healing of the grafts could
76
intervene in the frequency of inversions. Generally, the worst condi-

tions (e.g. on the dorsal side) would favour the inversions in male
grafts but would have the opposite effect with female grafts.
If the inversions observed in the grafts show a reversal of the normal

dominance of sexual potentialities (the synthesis of the type of in-
ductor being momentarily determined without any conformity with the
presumptive sex), in most cases, the normal dominance is maintained
or restored after a more or less prolonged period, since the trans-
planted gonads keep, or resume, their presumptive differentiations.
Thus, the sexualizing influence is completely restored in the trans-
planted genital territories and can act again on the gonads of the
opposite presumptive sex to which they have been grafted.
2. Restoration of Sexualizing Influence in Transplanted Genital

Territories
In these experiments, a genital territory of one sex' was transplanted.

Then, once healing was achieved, a second undifferentiated genital
territory of the opposite sex was grafted onto the former transplant.
This has been done in order to determine whether the sexualizing in-
fluence of ~he former transplant will exert itself on the latter.
Male and female territories of 3 to 6 week-old immature worms were
transplanted into the pharyngeal dorsal region (level 5-6) of hosts
of the same age. Three weeks later, territories of the opposite pre-
sumptive sex taken from new-born worms were grafted onto the former
transplant.
The use of immature specimens required some preliminary experiments

and the following points have been verified: The genital regions of
immature examples exert a sexualizing influence on grafts of the op-
posite presumptive sex transplanted into these regions (17% short-
lived or lasting inversions in male grafts; 70% inversions in female
grafts). Transplantations of the genital regions of new-born worms
into the dorsal pharyngeal region of immature hosts, bring about ap-
proximately the same results as those obtained at the same level with
new-born hosts (only 18% short-lived inversions in male grafts and
40% short-lived or lasting inversions in female grafts) .
The male presumptive gonads which were grafted onto the female region
transplanted into the pharyngeal region of immature hosts for a 3 week-
period, showed 23% short-lived or lasting inversions. They had under-
gone a feminizing influence from the female territory.
The female presumptive gonads grafted onto a transplanted male terri-

tory showed a high frequency (61%) of short-lived or lasting inversions.
They have undergone the masculinizing influence from the male territory.
Consequently, the sexualizing properties of transplanted genital terri-

tories are wholly restored after the time required for their complete
healing. The dominance of one potentiality over the other one, which
characterizes the differentiation of a genital territory (determining
the synthesis of a specific inductor), does not seem to depend on its
topographic position within the organism. More probably, it has been
established in the organization pattern during the development. The
reversal of this dominance (resulting from the experimental trophic
disturbances) would express a temporary reversibility of the sexual
differentiation of the genital territories having determined the syn-
thesis of the other type of inductor.
77
II. Factors of the Stability of the Sexual Differentiation
1. Importance of Trophic Factors
The influence of trophic factors on gametogenesis in adult oligochaetes

is a well known fact (Herlant-Meewis, 1967; Schoumaker, 1969; Lattaud,
1971), while fasting conditions cause gametogenesis to stop. When de-
velopment occurs (spontaneously during the fasting period or when the
worms feed again) some testes can turn into ovotestes, while the ova-
ries remain· normal. The same phenomenon is found with new-born worms
beginning to feed again after a long fasting period (Relexans, 1970).
Therefore, the factors of male diffe~ntiation are sensitive to trophic
disturbances, both with adult and new-born worms, while the factors
of female differentiation seem to be more stable.
On the other hand, it has already been noted that reference transplan-
tations of male or female territories in new-born worms, could lead to
inversions in the gonads of both presumptive sexes. This can only be
attributed to physiological disturbances (particularly trophic ones),
inherent in the transplantation itself.
Obviously the trophic disturbances are not serious enough to disturb

the female sex during the fasting period, but are strong enough to
disturb both sexes during the transplantations. But as the inversion
frequency of the gonads is higher in female than in male transplants
and since the female potentiality appears to be less sensitive to
trophic conditions than. the male one, it is inferred that the female
potentiality is slower to restore its dominance when the trophic dis-
turbances have been serious enough to depress it. On the contrary,
the male potentiality could easily be restored as the inversions are
always short-lived (except in heterosexual transplantations).
Undifferentiated gonads, excised with the adjacent body wall of new-

born worms, have been cultivated during a period of 13 to 28 days on
the culture medium of Wolff-Haffen (1952) or, alternatively, on this
medium modified by replacing the embryonic chick-extract with lact-
albumine hydrolysate. At the end of the culture the explants were
studied, using histology or electron microscopy (Relexans, 1973b).
Preliminary work with the light microscope showed in some cases, pre-
meiosis in the cultivated presumptive ovaries (normal differentiation)
and in the testes (inversions). Usually, however, the gonad develop-
ment did not proceed beyond the appearance of gonia. Typical spermatic
follicles have not been observed. Consequently, in most cases, the
development of the explanted gonads was insufficient to recognize the
differentiation of the sexual cell-lines, particularly that of the
male cell-line.
However, the sexual orientation of gonia has been detected by electron

microscopy (Chapron and Relexans, 1971). Oogonia were characterized
by the presence of a peripherical system of smooth reticula, while the
spermatogonia (deprived of the reticula) contained numerous free ribo-
somes and often showed "myelinic figures" in the intercellular spaces.
The study of explants by electron microscopy has shown that gonads of
both presumptive sexes, cultivated in vitro, were likely to be inverted
(Relexans, 1973b).
78
These results verified the bi-potentiality of the genital territories

of both sexes: The dominant relationship between the male and female
potentialities can be inverted, due to the trophic disturbances caused
by explantations (as in transplantation). The explantation would seri-
ously disturb the heterocatalytic genic activity and the inductor would
stop being synthesized (de-differentiation). From this stage and with
the restoration of a certain trophic level, most somatic cells would
again synthesize the inductor, conforming with the presumptive sex
(normal differentiation). In other cases, however, some cells could
synthesize the other type of inductor leading to an inversion. This
means, that in these cells, the activity of some genes could be re-
placed under experimental conditions by the activity of other genes.
Therefore, in new-born worms, the differentiation of genital territo-
ries might not be definitively determined.
These in vitro results differ from those of Lattaud (1973), who observed
inversions only in testes, when explanting adult gonads alone. Within
the limits of my hypothesis, the stability of the sexual differentia-
tion would seem to depend on the stage of development.
2. Influence of the Stage of Development: Pharyngeal Transplantations

of Genital Territories of Different Ages
These experiments consisted of comparing the differentiation of the

gonads of both sexes, the latter being excised with the adjacent body
wall, either from new-born or 3 to 6 week-old immature worms, and
transplanted into the same place (pharyngeal dorsal region of the adult
hosts) :
In the pharyngeal transplantation experiments with genital territories

from new-born worms, 18% of the testes produced short-lived inversions
while 40% of the ovaries produced short-lived or lasting inversions.
In the pharyngeal transplantation experiments with genital territories

from immature worms, 17% of the testes were inverted (short-lived or
lasting inversions), while the ovaries did not invert.
Therefore, it is obvious that the differentiation becomes stable in

the female region of adult worms.
3. Influence of the Maintenance of the Integrity in Morphological

Gonad Structures
Pharyngeal implantation of gonads excised from immature worms alone,

leads to different results depending if the gonads have kept their
morphological structures or if germ-cells were accidentally isolated
from the gonads during implantations.
Implanted testes and ovaries showed 90% and 10% short-lived or lasting
inversions, respectively. In relation to the previous results (gonads
+ adjacent body wall), is was observed that the sexual differentiation
was less stable. Therefore, the adjacent body wall plays a part in the
stability of the differentiation, mainly in the male region.
Implantations (and transplantations) of gonads into the pharyngeal re-

gion were sometimes accompanied by the isolation of germ-cells within
the host's pharyngeal musculo-glandular tissue. These cells probably
came from the gonads which had been "spoilt" by the experiment itself.
79
Isolated gonocytes were mostly undifferentiated. When they were differ-

entiated, their development towards oogenesis reached the pre-meiosis,
but rarely the auxocytosis stage; the differentiation towards spermato-
genesis could lead to typical spermatic follicles. A systematic obser-
vation of all gonocytes isolated from their original gonads during
pharyngeal implantations (and transplantations) gave the following
results: (a) In the course of male implantations or transplantations,
both male and female cell-lines have been observed in 85% of the cases,
in among the differentiated extra-gonadic germ-cells; (b) Similarly,
57% of female implantations or transplantations showed both sexual
lines among the differentiated extra-gonadic germ-cells. Therefore,
it is obvious that gonocytes isolated in the pharyngeal tissue, can
differentiate towards oogenesis or spermatogenesis - no matter what
the sex and stage of development of the gonads are from which they
come.
In amphibians (Stark, 1959) and insects (Lauge, 1969), the gonocytes

which did not reach the gonads, could not be differentiated sexually
during development. In my experiments, it is probable that extra-go-
nadal gonocytes could be differentiated, as they have been isolated
from the gonads at a later stage. Presumably, they had undergone an
inductive impulse from some somatic cells of the originating gonad;
their differentiation would not be spontaneous.
The fact that some isolated gonocytes start to differentiate without

any conformity with their presumptive sexes, would imply that before
isolation of the gonocytes, some somatic cells have elaborated the
inductor that caused the inversion. Gonadal morphological structures
were most seriously disturbed by the experimental procedure, in pre-
cisely those cases where the germinal cells had been isolated.
The high frequency of inversions among the extra-gonadal gonocytes

(when experimental circumstances have disturbed the structural integ-
rity of the gonads), could reveal a great affinity for the dominant
reversal of the sexual potentialities in somatic cells. In other words,
the stability of the sexual differentiation (that is characterized by
the synthesis of a specific type of inductor), would be related to the
integrity of the morphological structures in male and female gonads.
Consequently, if in adult worms the ovarian differentiation seems more
stable than the testicular differentiation, one reason for this sta-
bility could be a better capacity for the ovary to maintain its struc-
ture. This feature has actually been observed in vitro.
4. Importance of the Ventral Nervous System
The importance of the cephalic nervous system in the maintenance of

the reproductive activity and, particularly in meiosis, has been dem-
onstrated by Herlant-Meewis (1956, 1957, 1967). The involvement of
neurosecretions in gametogenesis has been verified among oligochaetes
(Rude and Linder, 1964) and leeches (Hagadorn, 1967; Malecha, 1970)
Considering the primary sexual differentiation, the removal of the

anterior nervous ganglions can produce testicular inversions when the
development begins (Herlant-Meewis, 1967). Nevertheless, this operation
is accompanied by a more or less prolonged fasting and it is difficult
to know whether the inversions must be attributed to the consequences
of the fasting period or to the removal of the nervous ganglions, or
to both these effects.
If new-born Eisenia have been fed for one week, further recurrent de-
capitations of these one-week-old individuals will not disturb their
80
gonadal differentiations (Relexans, 1974). Similarly, the substitution

of the brain by a nervous cord fragment in adults (leading to the final
suppression of cephalic ganglia; Cazaux, 1973) does not cause any in-
version but only growth disturbances. Therefore, the cephalic nervous
system does not seem to play any specific part in the primary sexual
differentiation, but could have an influence on trophic factors to
which the male differentiation is particularly sensitive.
Besides the large neurons and reserve cells, the ventral nerve cord
contains neurosecretory cells (Herlant-Meewis, 1966; Herlant-Meewis
et al., 1967) and monoaminergic neurons (Rude, 1966; Ehinger and Myhr-
berg, 1971; Myhrberg, 1972). Its importance in sexual differentiation
has not yet been studied.
a) Demonstration of the Influence of the Nervous Cord in Sexual

Differentiation
Removal of the nervous system in the genital region did not prevent
the worms from feeding, but led to testicular inversions (short-lived
inversions in young worms, short-lived or more or less lasting ones in
older individuals). The low frequency of these inversions recalls the
results obtained during the fasting experiments: removal of the ner-
vous cord in the genital region might cause trophic disturbances which
would be responsible for the testicular inversions.
Only one injection (5 mg/ml) of 6-hydroxydopamine (6-0HDA), an inhibi-
tor of catecholarnines (Ungerstedt, 1971), into the genital coelomic
cavity of new-born worms, caused development to be delayed after a
few weeks. Nine percent of the testes were transformed into ovotestis
(still short-lived inversions), while ovarian differentiation remained
normal. The frequency and character of the testicular inversions were
similar to those of the inversions observed during the fasting experi-
ments with new-born worms.
Therefore, the removal of the nervous cord as well as the inhibition

of the catecholamines, has the same effect as that observed after a
fasting period. These results suggest that the ventral nervous system
could interfere (through its catecholamines), in sexual differentia-
tion, by maintaining a certain trophic level (to which the male poten-
tiality is particularly sensitive) in the innervated region.
b) Importance of the Nervous Cord in Grafts during Transplantation

Experiments with New born Worms
Male grafts deprived of their nervous systems are always more frequent-
ly feminized than complete ones, whatever the transplantation level.
The increase of the inversion frequency remains rather slight in refer-
ence transplantations and outside the ventral genital region. This
increase is very high in heterosexual transplantations: male grafts
deprived of their nervous system and placed into the female region,
showed 52% short-lived or lasting inversions in their gonads, compared
to only 14% with whole grafts.
Female grafts deprived of their nervous systems are always less mascu-
linized than whole grafts. The decrease of inversion frequency is rath-
er slight in reference transplantations and outside the ventral gen-
ital region, but it is much higher in heterosexual transplantations.
In the male region, the frequency of ovary masculinizations varies
from 75% (in whole grafts) to 40% (in grafts deprived of their nervous
systems) .
81
In order to test the action of 6 OHDA in transplantation experiments,

male grafts were transplanted into the female region - the excision
of the nervous system in such grafts resulting in a very high increase
of inversion frequency in relation to whole grafts. The donors of male
grafts were new-born worms which had received an intracoelomic injec-
tion of 6-0HDA. Their male territories (with their nervous cord) were
then transplanted into the ventral female region of non-treated new-
born worms. Under such conditions, the feminization-frequency in trans-
planted presumptive testes reached 40% (short-lived or lasting inver-
sions). This phenomenon recalls the result obtained in this very same
region, with grafts deprived of their nervous system (52% short-lived
or lasting inversions). Therefore, 6-0HDA has realized effects in
grafts possessing their nervous systems similar to those observed af-
ter the nervous cord deprivation.
The removal of the nervous cord or the treatment with 6-0HDA in the
genital region, both lead to inversions - recalling those results ob-
tained with fasting young worms (testicular short-lived inversions).
During transplantations of genital territories, the removal of the
nervous cord from the grafts (or again the preliminary treatment of
male grafts with 6-0HDA in heterosexual transplantations) always leads
to a frequency-increase of feminizations in transplanted presumptive
testes and to a frequency-decrease of masculinizations in ovaries.
In comparison to whole grafts, such variations are particularly sensi-

tive in heterosexual transplantations. My results seem to show that
the ventral nervous system could play a trophic part (determined by
catecholarnines) in the primary sexual differentiation: within the
sexual balance, it would make the expression of the male potentiality
easier.
D. Summary and Conclusions
The transplantations of sexual territories in new-born Eisenia lead to

inversions in the transplanted gonads. The character and frequency of
these inversions depend on the transplantation site. They show that
gonads of both presumptive sexes can be inverted and reveal a male
inductive sexualizing influence in the ventral region of segments
10-11 and a female induction in the ventral region of segment 13. The
interpretation of such results pre-supposes the presence of male and
female potentialities in either sexual territory, one normally domi-
nating the other and thus determining the synthesis of a specific type
of inductor. Inversions would express a reversal of the normal domi-
nance and the synthesis of the other type of inductor.
The differentiation of the genital territories, characterized by the

elaboration of a specific type of inductor, seems to be connected with
ontogenesis. Its instability under experimental conditions can be tem-
porary only: sexual territories transplanted outside their natural
positions are able to restore their original sexualizing influences
after the time required for their complete healing.
The stability of the sexual differentiation can be temporarily affected

by several factors: a slight but lasting trophic insufficiency (fast)
only suppresses the dominating male potentiality in the male region.
On the contrary, serious trophic disturbances (transplantations and
explantations) can supress both the dominating male potentiality in
the male region and the dominating female one in the female region,
thus bringing about inversions of both sexes. Nevertheless, the male
82
potentiality which is more sensitive to trophic disturbances, would be

sooner restored (with the normalization of the trophic conditions)
than the female potentiality, when the latter has been depressed.
The stability of the sexual differentiation asserts itself during the

post-embryonic development in the female territory. The differentia-
tion of the adult ovary seems to be stable, provided that its morpho-
logical structures are not affected by the experimentation. However,
in both sexes and at any developmental stage, the stability of the
differentiation is connected with the integrity of the gonadal morpho-
logical structures.
The ventral nervous cord seems to play (through its catecholamines)

a trophic role in the male potentiality expression.
My experiments lead me to question the hypothesis of ovarian self-

differentiation in Eisenia. The results indicate a balance of sexual
potentialities, determining the synthesis of both specific inductors;
a male one and a female one, in the respective corresponding genital
territories.
References
Brien, P.: Biologie de la Reproduction animale. Paris: Masson 1966.

Cazaux, M.: Contribution a l'etude experimentale du role des ganglions
cerebroides dans la croissance et le developpement sexuel chez le
Lombricien Eisenia foetida Sav. These d' universi te. Bordeaux 1973.
Chapron, C., Relexans, J.C.: Ultrastructure des gonocytes primordiaux
et des gonies chez l' hermaphrodi te Eisenia foetida (Oligochete Lom-
bricide). C.R. hebd. Seanc. Acad. Sci. Paris 272, 2916-2919 (1971).
Charniaux-Cotton, H.: Croissance, regeneration et determinisme endo-
crinien des caracteres sexuels d' Orchestia gammarella Pallas, Crustace
Amphipode. Ann. Sci. nat. 19, 411-560 (1957).
Charniaux-Cotton, H.: Etude comparee du developpement postembryonnaire
de l' appareil genital et de la glande androgene chez Orchestia gamma-
rella et Orchestia mediterranea (Crustaces Amphipodes). Autodifferen-
ciation ovarienne. Bull. Soc. Zool. France 84, 105-115 (1959).
Charniaux-Cotton, H.: Hormonal control of sex differentiation in Inver-
tebrates. In: Organogenesis (eds. R.L. de Haan, H. Ursprung), pp.
701-740. New York-Chic ago-San Francisco-Toronto-London: Holt, Rine-
hart and Winston 1965.
Charniaux-Cotton, H.: Recherches recentes sur la differenciation sexu-
elle et l'activite genitale chez divers Crustaces superieurs. In:
Hormones et differenciation sexuelle chez les Invertebres (Cours
et Documents de Biologie. Dir. E. Wolff), vol. 3, pp. 127-178.
Paris-London-New York: Gordon and Breach 1972.
Ehinger, B., Myhrberg, H.: Neuronal localization of dopamine, noradre-
naline and 5-hydroxytryptamine in the central and peripheral nervous
system of Lwnbricus terrestris L. Histochemie~, 265-275 (1971).
Hagadorn, I. R.: Hormonal Control of Spermatogenesis in Hirudo medicinalis.
In: Neurosecretion (ed. F. Stutinsky), pp. 219-228. Berlin-Heidel-
berg-New York: Springer 1967.
Herlant-Meewis, H.: Reproduction et neurosecretion chez Eisenia foetida
(Sav.). Ann. Soc. Roy. Zool. Belg. 87, 151-183 (1956/57).
Herlant-Meewis, H.: Les cellules neurosecretrices de la chaine nerveuse
d' Eisenia foetida. Z. Zellforsch. 69, 319-325 (1966). .
Herlant-Meewis, H.: Evolution de 1" appareil genital d' Eisenia foetida au
cours du jeune, de la regeneration posterieure et a la suite d'abla-
tion de ganglions nerveux. Ann. Soc. Roy. Zool. Belg. 96, 189-240
(1967). -
83
Herlant-Meewis, H., Naisse, J., Mouton, J.: Ph§nom~nes neuros§cr§toires

au niveau de la chaine nerveuse chez les Invert§br§s. In: Neurose-
cretion (ed. F. Stutinsky), pp. 203-218. Berlin-Heidelberg-New York:
Springer 1967.
Lattaud, C.: Evolution des testicules et des v§sicules s§minales chez
les Oligoch~tes Eisenia foetida et LumlJricus terrestris, au cours du
jeGne suivi ou non de re-nutrition. C.r. hebd. S§anc. Acad. Sci.
Paris 272, 319-322 (1971).
Lattaud, C.: Autodiff§renciation ovarienne chez l'Annelide Oligoch~te
Eisenia foetida f. typica demontree au moyen de la culture organoty-
pique. C.r. hebd. Acad. Sci. Paris 276, 1737-1740 (1973).
Laug§, G.: Etude des gonades des Intersexu§s triploides de Drosophila
melanogaster. Description morphologique, ontogen~se des structures
morphologiques. Ann. Soc. ent. Fr. 5 (2), 253-314 (1969).
Malecha, J.: Etude experimentale de la-maturation sexuelle chez les
Hirudin§es. Bull. Soc. Zool. Fr. 95 (3), 517-528 (1970).
Myhrberg, H.E.: Ultrastructural localization of monoamines in the cen-
tral nervous sy.stem of Lumbricus terrestris L., with remarks on neuro-
secretory vesicles. Z. Zellforsch. 162, 348-362 (1972).
Relexans, J.C.: Action du jeGne applique d~s l'§closion, sur Ie d§ve-
loppement des gonades d' Eisenia foetida (Ann§lide Oligoch~te, Lombri-
cid§). C.r. hebd. S§anc. Acad. Sci. 271, 1530-1533 (1970).
Relexans, J.e.: Transplantations de gonades indiff§renciees chez
1 'hermaphrodite simultane Eisenia foetida (Oligoch~te, Lumbricidae)
Mise en evidence de facteurs locaux (inducteurs?) de la differen-
ciation sexuelle. J. Embryol. expo Morph. 30, 143-161 (1973a).
Relexans, J.C.: Mise en §vidence, en culture organotypique, de la bi-
potentialite des r§gions male et femelle chez les nouveau-nes
d' Eisenia foetida (Oligoch~te Lombricide). C.r. hebd. S§anc. Acad.
Sci. Paris 277, 1349-1352 (1973b).
Relexans, J.C.~echerche exp§rimentale des facteurs de la differen-
cia tion sexuelle chez l' hermaphrodi te simul tane Eisenia foetida Sav.
(Oligoch~te, Lombricide). Th~se Doctorat ~s Sciences, Bordeaux 1974.
Rude, S.: Monoamine-containing neurons in the nerve cord and body wall
of Lumbricus terrestris. J. Compo Neur. 128, 397-412 (1966).
Rude, S.S., Linder, H.J.: The effect of the brain on spermatogenesis
in the oligochaeta Eisenia foetida. Amer. Zoologist. 4, 327 (1964).
Schoumacker, H.: Quelques anomalies gEmi tales chez Eisenia foetid.a Sav.
(Oligochete Lombricid§). Ann. Soc. Roy. Zool. Belg. 99; 169-191
(1969). -
Stark, O.J.: Die streuenden Geschlechtszellen von Triton alpestris. Z.
Zellforsch. 50, 693-748 (1959).
Ungerstedt, U.:~istochemical studies on the effect of intracerebral
injections of 6-Hydroxydopamine on monoamine neurons in the rat
brain. In: 6-Hydroxydopamine and Catecholamine Neurons (ed. T.
Malfors, H. Thoenen), pp. 101-127. Amsterdam-London: North-Holland
Publishing Company 1971.
Wolff, E., Haffen, K.: Sur une methode de culture d'organes embryon-
naires in vitro. Texas Rep. Biol. Med. lQ., 463-472 (1952).
Sex-Determination in Bonellia
R. Leutert
A. Introduction
BonelZia viridis (Echiurida) is well known for its phenotypical sex-deter-

mination. This phenomenon has been investigated by Baltzer (1912, 1914,
1923,1925,1926,1928,1931,1932,1934,1937). Herbst (1928, 1929,
1935, 1940) and other authors. Wilczynski (1960, 1968) has questioned
these results and proposed that the sex-determination was progamic and
genotypical. After Wilczynsk~ (1968) the oocytes would be sexually
determined during oogenesis by different types of nurse cells. His
results were based upon cytological investigations and some experi-
mental data. Therefore, the problem has been reinvestigated, using a
greater number of individuals than those in the experiments of Baltzer,
and applying electron microscopic methods to describe the oogenesis.
This paper reviews the early literature of Baltzer, Herbst and Wil-
czynski and deals with the mechanism of sex-determination in Bonettia,
showing that the differentiation into male and female individuals is
a consequence of a complex interaction between environmental and gene-
tical factors. Furthermore, the question of how this unstable system of
sex-determination was established during evolution will be discussed
shortly.
B. Survey of Earlier Studies
The mature female body of BonelZia is about 8 cm long with a proboscis

reaching a length of more than 1 m. The animals live in small holes
and usually only their proboscis is visible seeking food particles on
the ground. However, the male individuals are of a very small size
(1 to 3 rom) and live parasitically in the uteri of adult females. As
a consequence, the male organs became very rudimentary. Baltzer sug-
gested that the determination and differentiation into the female or
male animals depended on the environment. According to this theory,
only a single type of sexually indifferent larva existed. If the tro-
chophore-like larvae come into close contact with the proboscis of an
adult female, they become sexually rudimentary and develop within
3 weeks into planaria-like males, which penetrate the uterus of their
host and live in it as parasites. Unless the larvae come into contact
with a proboscis of a female, they differentiate into females, reach-
ing sexual maturity within 2 years ("Glaszuchten").
Baltzer succeeded in breeding male individuals in the presence of a

female proboscis ("Russelzuchten") and with extracts of a female in-
testine. This proved the existence of some external factor responsible
for male sex-determination. However, this system of phenotypical sex-
determination does not seem to be a clear cut case. Besides the dif-
ferentiation into mature males and females there are a great number
of larvae developing into intersexes, showing a mosaic of both male
85
and female characteristics (Baltzer, 1925; Zurbuchen, 1937). Moreover,

in breedings without a female proboscis, a certain percentage of lar-
vae develop into males, while in breedings with a proboscis there are
always some female individuals developing. This phenomenon can hardly
be explained by assuming a purely phenotypical sex-determination. These
exceptions caused several speculations about other unknown (probably
genetical) factors influencing the sex-determination.
Herbst's (1928, 1929, 1935, 1940) contributions to the problem were

experiments using a series of other "unphysiological" external factors
that stimulated the indifferent larvae to differentiate into male in-
dividuals. This was achieved by raising the concentration of Cu 2 + and
K+ ions in the sea-water; by diminishing the Mg 2+ and S042- concentra-
tion in the artificial sea-water; or by lowering the pH when dissolving
CO 2 gas in the sea-water. These experiments showed that the differen-
tiation into male individuals was probably an unspecific, general in-
hibition of a female program. In this case, the larvae can only realize
the rudimentary male program.
The investigations by Wilczynski (1960, 1968) were mainly cytological

and led to an entirely new hypothesis for sex-determination in Bonellia.
Wilczynski described two types of nurse cells in oogenesis, that would
determine two types of eggs. Therefore, it seemed that two types of
sexually predetermined larvae existed. The female predetermined larvae
would develop freely into females, the male predetermined larvae how-
ever would try to come into contact with a female proboscis, then at-
tach themselves onto it and develop into dwarf males. The cytological
data of Wilczynski (1968) are in contrast to our investigations, as we
will demonstrate below. In one breeding, Wilczynski (1968) described
a 1:1 relation of males and females. But as the mortality after 3 weeks
was more than 90% in his aquarium, these data cannot be taken as rele-
vant.
C. Results
The classical experiments of Baltzer have been repeated, using a much

greater number of larvae (Fig. 1). The mortality after 30 days was less
than 10%. This low mortality factor was important, otherwise it could
be argued that eventual male or female predetermined larvae died in
the environment favoring the female or male development, respectively.
In breedings without a female proboscis ("Glaszuchten"), more than 50%
of the larvae became female. On the other hand, in breedings with a
female proboscis ("Russelzuchten") invariably more than 50% of the
larvae became male (Fig. 1). This means that at least more than 50%
of the larvae are able to develop in an alternative way (to become
either female or male individuals). The sex-determination in these
larvae is, therefore, purely phenotypic. The amount of indifferent
larvae (larvae that had stopped their development) and intersexes
(showing both male and female characteristics) was about the same in
both cases and did not depend upon the time when the external factors
were applied. Sex-determination is extremely unstable. As the larvae
become female individuals in the "normal environment" (without addi-
tional specific external factors), it might be suggested that the sub-
stance in a female proboscis suppresses the realization of female struc-
tures, but may induce the development of male structures. Furthermore,
as the males showed a lower degree of differentiation than the females
and even the larvae, it can be assumed that the effect of the proboscis
substance was inhibitory, and not inductive to the genetical programs.
This assumption is supported by the experiments of Herbst (1929) who
86
demonstrated that even very unspecific substances might inhibit female

differentiation.
c:::J Glaszuchten (n = 1770 1
~ Rl.isselzuchten ( n=440 1
80
Fig. 1. Breedings of
BoneUia-larvae in dif-
ferent environsments.
"Glaszuchten": without
a female proboscis,
"Rlisselzuchten": the
larvae grew up under
influence of a female
proboscis, indifferent
larvae: larvae that
show intersexual char-
acteristics (mosaic of
male and female struc-
tures) and larvae that
have stopped developing
These experiments have demonstrated that at least some of the larvae

in BoneZZia are capable of developing phenotypically in an alternative
way. However, a small percentage of the larvae differentiated into the
sex other than expected (Fig. 1). This fact suggests the existence of
genetical sex-determination.
Wilczynski's (1968) cytological investigations led him to propose an

early sexual dimorphism of the eggs due to the phagocytosis of two
different types of nurse cells. However, I could not confirm these
results. Electron microscopic investigations showed that the nurse
cells were in close contact with the egg cell through large cytoplas-
mic bridges, but the nuclei of the nurse cells were not incorporated
in the egg cell (Fig. 2) as Wilczynski (1968) had assumed. For this
reason, it seems impossible that two different types of nurse cells
could determine the sex of the ripening egg.
D. Discussion
About 10% of BoneUia larvae do not develop in the expected direction.

This fact can hardly be explained if it is assumed that there is only
phenotypical sex-determination. The exceptions can arise from several
causes: unstable laboratory conditions; unknown environmental influ-
ences; genetical background, etc. To explain all the aspects of this
extremely unstable system of sex-determination, it is therefore sug-
gested that there is a genetical background not allowing all of the
genotypes to develop in an alternative way (Fig. 3). Only larvae show-
ing an equilibrium of male and female determining factors are able to
develop alternatively. When the genetical constitution favors either
the male or female factors, the larvae are genotypically pre-determined
and realize anyway their programmed sexual differentiation: the pro-
grammed females realize their program in spite of the presence of ex-
87
ternal factors that may usually inhibit the development of female fea-
tures. On the other hand, genetically programmed males realize their
program even if the "male determining factors" of a female proboscis
are absent.
Fig. 2. The nurse-cells (upper half of the picture) are connected with
the oocyte by large cytoplasmic bridges (arrows). The nuclei are not
incorporated in the growing oocyte
However, it appears that in nearly 90% of all BoneZlia larvae, sex-deter-

mination is mainly phenotypical. Yet, depending upon the genetical
constitution, not all the larvae can be masculinized to the same extent
by external factors. Some larvae (about 10%) do not have the ability
to develop in an alternative way and are, therefore, genetically pre-
determined to become male or female individuals. Under normal condi-
tions ("Glaszuchten") and without the presence of an adult female,
most genotypes (except genetic males) realize the female program. Under
"unphysiological" artificial conditions ("Russelzuchten"), most geno-
types (except genetical females) realize · their male program, the female
one being suppressed.
The phylogenetic problem of explaining how this unstable system of

sex-determination arose in evolution is uncertain. Most of the fami-
lies related to the Bone LLidae do not show any sex dimorphism. As far as
is known, sex-determination is, in this case, genotypical (Fig. 4).
A solution may lie in the ecological study of BoneZlia viridis, comparing
it with other related families. Most echiurid worms move their whole
body along the sea bed to search for food. For this reason, they can
easily meet each other for reproduction and release their gametes into
the environment. It is likely that in these species the males are al-
88
Degree of
differentiation
IT]
I
t--------:::l---,~ ~
[j]
larvae with
cf
bisex.ual 6
predetermination '
geno - phenotypical geno -

t~pical 4 .5 typIcal
. ~ " .. . ~. : ". :.. .' ", ' • • '.If . ~ • ..lr
..
.. ... ," • .,If • • "r '" • • ~ ..If
~ ~ . ,~ • .If.. . . ~ . ~.
.
• "'. .If' ": Jr •
."
" ..
.." .... .
.
.." . ..
Number of
male and female
determining
genetfc factors ..:;._ __ _ _ _ _ __ _ _ _ _ _ _ _---'''''
".
Fig. 3. Mechanism of sex-determination in Bonellia. 1 Genetical program

for both male and female differentiation in the same individual (sexual
indifferent larvae). 2 "Normal" destiny: in case of absence of the ex-
ternal factors, the female program dominates the male program. Further-
more, a certain percentage of larvae develops into female individuals
in spite of the presence of male-determining factors: genotypical fe-
males. 3 MOre than 50% of the larvae are able to be sexually determined
by the influence of external factors. Unless these factors are applied
in a sufficient dose and at the appropriate stage of larval life, the
larvae become intersexes. 4 Male-'determining external factors (e.g. the
secretory substance of the female proboscis) suppress the female gene-
tical program: the larvae become phenotypical males. 5 If the male de-
termining factors are absent, the larvae may develop into phenotypical
females. 6 Even if the male determining factors are absent, a certain
percentage of larvae develops into male individuals: genotypical males
Female determining genetical factors in a
I.'. ': 1 certain individual
+ + +
I~ ++ +
+ +
.+ +
+
+ +1
Male determining genetical factors in a
certain individual
ready to a certain degree under control of the females, the latter in-
ducing the release of spermatozoa. Bonellia viridis, on the other hand,
lives in small holes and only the proboscis is visible, searching for
food particles. Therefore, it is more complicated for them to meet each
other for reproduction. Consequently, it is possible that the female
89
probiscis influences the coordination of gamete ripening. This influ-

ence has probably become more and more important and controls not only
the release of gametes, but also their ripening, the differentiation
of the testes and even the whole male development. As the dwarf males
live in the female uterus, fertilization is assured.
This evolutionary interpretation is purely speculative. To test this

assumption it would be necessary to investigate the developmental biol-
ogy of a closely related genus (e.g. Acanthobonellia ) , in which the
males are not endoparasitic in the female but already show an extreme
sex-dimorphism.
Echiurus Bonellia
)\
no sex-dimorphism extreme sex-dimorphism

s.ex- determination sex-determination
stable genotypical unstable genotyplcal
Fig. 4. Sex-dimorphism and sex-determination are highly correlated in

Echiurida. This suggests that in phylogenesis the influence of the
adult female became more and more important f or the sexual determina-
tion and differentiation of developing males. Bonellia viridis seems to
represent the last step in the evolutionary scale of Echiurid worms
References
Bacci, G.: Sex Determination. Oxford-New Yo rk-Londo n-Paris: Pergamon

Press 1965.
Bal tzer, F.: liber die Entwicklungsgeschichte v on Bonellia . Verh. Dtsch.
Zool. Ges. 22, 252-261 (1912).
Baltzer, F.: Die Bestirnmung und der Dimorphismus des Geschlechtes bei
Bonelli a . Wlirzburg: Sitz.-Ber. Physikal.-med. Ges . ..11., 1-4 (1914).
Baltzer, F.: Zur Entwicklungsgeschichte und Auffassung des M~nnchens
der BoneUia . Verh. Dtsch. Zool. Ges. 22, 65-68 (1923).
Baltzer, F.: untersuchungen liber die Entwicklung und Geschlechtsbe-
stimrnung der BoneUia. Pubbl. Staz. Zool. Napoli .§.' 223-287 (1925).
90
Bal tzer, F.: tiber die Vermannlichung von Larven durch Bonellia- Extrakte.
Rev. Suisse Zool. 33, 359-374 (1926).
Baltzer, F.: tiber metagame Geschlechtsbestimmung und ihre Beziehungen
zu einigen Problemen der Entwicklungsmechanik und Vererbung. Verh.
Dtsch. Zool. Ges. 18, 273-325 (1928).
Baltzer, F.: Echiurid~ Handbuch der Zoologie (Ktikenthal) 2, pp. 62-
168 (1931).
Baltzer, F.: tiber die ohne Rtisselparasitismus entstandenen Spatmann-
chen (genetische Mannchen) der Bonellia viridis. Rev. Suisse Zool. ~,
281-306 (1932).
Baltzer, F.: tiber die Zuchtmaglichkeit der Bonellia viridis vom Ei bis
zum erwachsenen Zustand. Rev. Suisse Zool. 41, 407-409 (1934).
Baltzer, ~.: Analyse des Goldschmidtschen Zeitgesetzes der Intersexua-
Ii tat auf Grund eines Vergleichs der Entwicklung der BoneUia- und
Lymantria-Intersexe. Roux' Arch. 136, 1-43 (1937).
Fisher, W.K.: A review of the Bonellidae. Ann. Mag. Nat. Hist. ser. 11,
14, 832-860 (1948).
Glaus, H.: Erzeugung, Organisation und Entwicklungsmechanik der Rtissel-
zuchtintersexe von BoneUia viridis. Pubbl. Staz. Zool. Napoli li,
40-113 (1933).
Goldschmidt, R.: Die sexu.ellen Zwischenstufen. Monogr. Physiol. Berlin
23 (1931).
Herbst, C.: untersuchungen zur Bestimmung des Geschlechts. Ein neuer
Weg zur Lasung des Geschlechtsbestimmungsproblems bei BoneUia viridis.
Sitz.-Ber. Heidelberg. Akad. Wiss., Math.-nat. Kl., 3-19 (1928).
Herbst, C.: Weitere Experimente tiber die Vermannlichung indifferenter
Bonellia-Larven durch ktinstliche Mittel. Sitz.-Ber. Heidelberg. Akad.
Wiss., Math.-nat. Kl., 2-43 (1929).
Herbst, C.: Die Abhangigkeit des, Geschlechtes vom Kaliumgehalt des
umgebenden Mediums bei Bonellia viridis. Roux' Arch. 132, 576-599
(1935). -
Herbst, C.: tiber BoneUia-Weibchen mit spaltfarmiger Leibeshahle und
ihre Bedeutung ftir meine Hydratationstheorie der Geschlechtsbestim-
mung. Roux' Arch. 140, 252-284 (1940).
Leutert, R.: Zur Geschlechtsbestimmung und Gametogenese von Bonellia
viridis Rolando. J. expo Embryol. Morphol. 32, 169-193 (1974).
Wilczynski, J.: On the egg dimorphism and sex~etermination in Bonellia
viridis. J. expo Zool. 143, 61-76 (1960).
Wilczynski, J.: On the sex in BoneUia viridis. Acta Biotheoretica 18,
338-370 (1968). -
Zurbuchen, K.: Entwicklungsmechanische untersuchungen an BoneUia viridis
II (Abgektirzter Rtisselparasitismus). Pubbl. Staz. Zool. Napoli l.§.,
28-79 (1937).
Hermaphroditism and Gynandromorphism In Malacostracan
Crustacea
H. Charniaux-Cotton
The malacostracan Crustacea are generally gonochoristic with geneti-

cally determined sex. However, several decapod species and two isopod
groups, Cymothoidae and Cryptoniscidae, are functional protandric hermaph-
rodi tes with ovotestes. Protogyny is known in the isopod Antheridae:
Cyathura carinata and in the Tanaidacae: Heterotanais oerstidi (Juchault,
1966; Noel, 1973).
Malacostraca also exhibit non-functional hermaphroditism: oogenesis
in the male gonads which proceeds to the end of previtellogenesis.
The reverse situation, spermatogenesis in the gonads of females, is
never observed. Oogenesis in testes is known in all species of Tali-
tridae except Orchestia gammareUus, and in several genera of decapods:
Gebia, Astacus, Eupagurus, Inachus, Eriocher, etc. Intersexuality, often found
in Malacostraca, is discussed in this volume by Ginsburger-Vogel.
A. Functional and Non-functional Hermaphroditism: Results Obtained

in Amphipods Talitridae
The first evidence of the presence of sex hormones in invertebrates

was the discovery, in the amphipod Orchestia gammareUus, of the endocrine
functions of the ovaries and of new glands, the androgenic glands (AG) ,
(Charniaux-Cotton, 1952, 1954). Experimentally-induced sex inversions
have demonstrated that the AG control male morphogenesis and spermato-
genetic activity. In the absence of androgenic hormone AH, female mor-
phogenesis occurs, with self-differentiation of the ovary followed by
induction of the female secondary characters by ovarian hormones
(Fig. 1).
The graft of an AG in an immature or mature female brings about mascu-

linization of gonads and secondary sex characters. Conversely, after
the removal of the AG in Orchestia montagui, the testes are completely
transformed into ovaries and such ovaries induce the external female
characters. More details are given in Charniaux-Cotton (1965, 1970,
1972). Similar results have been obtained in terrestrial isopods (Kata-
kura, 1961; Juchault and Legrand, 1964; Legrand and Juchault, 1972).
These experimental sex inversions are possible because primary gonia

persist during the entire life of the Talitridae. These gonia are en-
closed in a strand of mesodermic cells which runs from one end of the
gonad to the other, against the basal lamina. They undergo spermato-
genesis or oogenesis, according to whether or not they receive AH.
Experimental inversions in Talitridae and terrestrial isopods prove
that each sex possesses complete genetic information for the morpho-
genesis of both sexes. In the presence of AH, the genes for male mor-
phogenesis act; in absence of AH, genes for female morphogenesis act
(Fig. 2). It must be assumed that the genetic difference between male
92
g.C. m.C.
p e.<f ch. + ov.

GENETIC \> . ~ te . ~ch. +
spontoneously ~ a.g.
~~~.&tl"'!'??~
.,~a.~~
~~~~matO<.lenesis+
vitellogenesis -
GENE TICO
e dch. + .-
e . ~c h.- .-
omahng behav ior t---'-~
Fig. 1. Diagram of endocrine control of sex differentiation in Talitridae .
The arrows indicate the stimulating (+) or inhibiting (-) effects of
ovarian and androgenic hormones; one of two ovarian hormones is se-
creted only during vitellogenesis (dashed arrows). a . g . androgenic
gland (the androgenic gland does not develop in the female); e . ~ ch.
external female characters; p . e . ~ ch. permanent external female char-
acters; t . e . ~ ch . temporary external female characters; e. if ch .
external male characters; g. c . germ cell; m. c .mesodermal cells; ov .
oviduct; s p. sperm duct. (From Charniaux-Cotton, pp. 135-198. In: "The
Hormones" (eds. G. Pincus and K.V. Thimann). Academic Press 1963)
and female is responsible for the dev elopment or non-development of

the AG itself. The genetic system of sex-determination is not known
in Crustacea Malacostraca; it is possible that some species possess
heterochromosomes (Charniaux-Cotton, 1960a). In O. gammarenus, the female
is homogametic (see Ginsburger-Vogel, this volume). It is recalled that
such a genetic system could consist of a single pair of alleles, as in
the mosquito, Culex mo lestus (Gilchrist and Haldane, 1947) in which the
male's are!::! and the females !!!. The duplication of genes M and m and
m m
the suppression of crossing over in the male result in shielding the
M chromosome from natural selection; deficiencies accumulate and the
"M chromosome will degenerate into a nearly functionless Y. It is a
striking fact that this has not happened in many large and successful
groups" (Gilchrist and Haldane, 1947). No endocrine link between the
sex-determining genes and the AG Anlagen has ever been demonstrated
(Charniaux-Cotton, 1972).
The results obtained in gonochoristic peracarid Malacostraca easily

account for secondary hermaphroditism: it is sufficient to note that
in genetic males, the AG disappear (protandry) or appear late (proto-
gyny). But where genetic males function both as males and as females,
the females disappear. Indeed, in the hermaphroditic prawn P . bo r ealis ,
there is a very small percentage of females, probably declining.
Oogenesis in testes is easily understood. Certain gonia simply escape

the action of AH and evolve spontaneously to oogenesis (vitellogenesis
is inhibited by AH) .
93
I
inactivates
Male-determining Genes aiting in Genes acting in

male morphogenesis ovarian external female
and functions self-differentiation morphogenesis
!t
Genetic male
Female-determining Genes acting in Genes acting in

genes male morphogenesis ovarian 11:: external female
"""-"".roo".,,oo ~
~
•
and functions morphogenesis
Non-development
~
ovar~es
of A G
Genetic female
Fig. 2. Control of sex differentiation in Talitridae. The male and the

female have genes for both sexes. Their genomes differ only in the
genetic control of androgenic glands, AG; AH androgenic hormone;
HO ovarian hormones
Non-functional hermaphroditism shall be considered first and then func-

tional hermaphroditism, in two species of prawns on which experimental
studies have been performed: Pandalus borealis and Lysmata seticaudata.
B. Non-functional Hermaphroditism
Non-functional hermaphroditism is characterized by oogenesis occurring

in testes. It exists in all species of Talitridae (except O. gammareUus)
(Charniaux-Cotton, 1962; Berreur-Bonnenfant, 1971; Carre-Lecuyer, 1970;
Denay, 1970; Fried-Montaufier, 1967), and in several decapods (Baffoni,
1947). In Talitridae, oogenesis occurs to a variable degree in the an-
te~ior region of the testis. Seasonal change in this activity is very
evident in Talitrus saUator (Fig. 3) where in winter, the testes are in
fact completely transformed into ovaries. The AG are very small; how-
ever the titer of AH is sufficient to inhibit vitellogenesis.
In testis of Talitridae there is a gradient of response to AH along the

germinative zone, the anterior region being less sensitive than the
posterior one. This gradient is probably related to the initial action
of AH, which progresses by diffusion along the genital tractus from
the young AG to the anterior end. During periods of genital inactivity,
in the crayfish Pontastacus leptodactylus the AG are very small, spermato-
genesis has stopped, and oocytes appear in testis (Payen, 1974a)(Fig. 4)
94
Fig. 3. Schematic
view of the genital
apparatus of male
Tal itrus saltat or.
(1) Testis without
oocyte; (2) and (3)
testis with oocytes;
(4) testis in win-
ter, the germina-
tive zone gives
rise only to oogen-
esis, the andro-
genic gland is
small. ag genital
apophysis; c d vas
deferens; ga andro-
genic gland; v s
seminal vesicle;
ZG germinative
zone. (From Fried-
Montaufier, 1967)
og
og
C. Functional Protandric Hermaphroditism in the Prawns Pandalus boreali s

and Lysmata se t icaudata
I. Summary of the Life Cycle
The sexual biology is very similar in the two species (Berreur-Bonnen-

fant and Charniaux-Cotton, 1965; Hoffman, 1972). The young prawns
function as males during the first reproductive season (summer), about
one year after they hatch. The copulatory organs (distal end of the
endopodite of the first pleopods, appendix masculina located on the
endopodite of the second pleopods), and other male sex characters have
reached maximum development. The AG are well-developed. During the
succeeding winter, the male characters and the AG begin to regress.
Spermatogenesis is reduced. With the following reproductive season
the next summer, spermatogenesis is again active; the AG are larger
than during the winter (Touir, 1973). The duration of this second pe-
riod of male genital activity varies with the individual; it terminates
as the AG becomes progressively smaller and ultimately, all its cells
degenerate. The total disappearance of the AG, first observed in Ly s-
mata (Charniaux-Cotton, 1958a), just as the experimental removal of AG
in Tali t ridae , induces the reversion of males to females. The external
male characters, already much reduced during the preceeding winter,
disappear. The germinative zone of the gonad gives rise only to oogen-
esis, and vitellogenesis starts. The individual displays the external
female characters. The animal will function a second time as a female
the next summer, and then dies at the age of about 3 1/2 years.
95
Fig. 4. Sections of a testis of Puatastacus ~eptodacty~us during the geni-

tal inactivity. Two acini show an oocyte in previtellogenesis. CF fol-
licular cell; G gonia; LA acinus lumina (empty); NM mesodermic cell
of the acinus; OVC PV oocyte in previtellogenesis; SPZ spermatozoa.
(From Payen, 1974a)
During the male phase, oogenesis proceeds in the gonads until the end
of first growth or previtellogenesis. There is only a single germina-
tive zone in each gonad. The gonia of this zone give rise to spermato-
genesis and ovogenesis, as in the testes of Talitridae (except o. gamna-
reUus), but the modalities differ from P. borea~is to L. seticaudata.
II. Gametogenesis in Panda~us borea~is
During the male phase, oocytes are always present in the central re-
gion of the gonadal tube. The authors consider that there are two dis-
tinct regions of the gonadal tract, female and male; the male region
being functional only during the male phase (Allen, 1959).
Histological study (Charniaux-Cotton, 1965) has demonstrated that

transverse sections are identical along the gonad and that only a sin-
gle germinative zone runs along the tube, attached to the basal lamina
(Fig. 5). During the reproductive season some gonads exhibit spermato-
genesis: the gonia which have left the germinative zone multiply
(Fig. 6A), and the secondary spermatogonia move out from both sides
of this zone and surround the ovarian core (Fig. 5); an older genera-
tion of male cells occupies the cortex. It is remarkable that in other
males (in their first or second season of genital activity) the germi-
native zone gives rise to oogenesis (Fig. 6B). The young oocytes remain
near the germinative zone and are enclosed by follicle cells. The diam-
eter of the ovarian core remains about the same, because the oldest
oocytes degenerate (at the end of previtellogenesis). Thus, the germi-
native zone exhibits alternations of spermatogenetic and oogenetic
activity. The number of such alternations, or what Gontrols them, is
96
not known. (It is possible that a periodic exhaustion of AH occurs in

the germinative zone.)
Fig. 5. Diagrams of the gametogenesis

in Pandalus borealis during the male
phase. The diagram on the right shows
that one germinative zone (+++) runs
along each gonad. The diagram on the
left is a transverse section which
shows the movement of secondary sper-
matogonia around the oocytes. CD vas
deferens; ovc oocyte; Ovd oviduct;
Spg secondary spermatogonia; ZG ger-
minative zone. (From Berreur-Bonnen-
fant and Charniaux-Cotton, 1965)
In the young prawns, soon after metamorphosis, oogenesis appears be-

fore spermatogenesis (Aoto, 1952; Hoffman, 1972). The author is not
in agreement with the interpretation by Hoffman of the development of
the ovotestis in P. platyceros (probably similar to that in P. borealis).
This interpretation is restated as follows: Each young gonadal tube
"has a cortical layer of epithelial cells from which all other gonadal
cells originate". From this gonadal epithelium are derived the "primary
gonial cells ... which differentiate into the follicle cells and oocytes
to form the ovarian portion of the gonad" ... "Oocytes remain attached
to the basement membrane of the gonadal epithelium by stalks". In older
prawns, spermatogonia differentiate in the gonadal epithelium around
the ovarian portion and form the testicular region. The female and male
regions would be separated by large primary oocytes that "remain at-
tached to the basement membrane".
Fig. 6 and our microphotographs (Berreur-Bonnenfant and Charniaux-

Cotton, 1965) fail to confirm the presence of such attachments. As in
Lysmata seticaudata , there is only one germinative zone. Clear proof is
provided by Fig. 6B, of the ovotestis during oogenesis: there are no
spermatogonia around the ovarian region because they have become sper-
matocytes. Moreover, in Crustacea, I have never observed the differen-
tiation of gonia from other cells. The germ cell line is precociously
separate (Payen, 1974a,b).
After degeneration of the AG, spermatogenesis stops definitively; fe-
male self-differentiation occurs and vitellogenesis starts in the
97
largest oocytes. The evolution of the external sex characters has

been described in previous publications (Berreur-Bonnenfant and Char-
niaux-Cotton, 1965).
Fig. 6 A and B. Transverse

sections of the gonad of
Pandalus borealis during the
male phase. (A) The germi-
native zone gives rise to
spermatogenesis. (B) The
germinative zone gives
rise to oogenesis.
c.m. mucus cell; G gonium
in the germinative zone;
spc. spermatocyte; spg .
secondary spermatogonium
(From Charniaux-Cotton,
1965)
98
III. Gametogenesis in Lysmata seticaudata
In L. seticaudata as in P. borealis there is only one germinative zone in

each of the two gonads. This zone is the site of spermatogenesis in
the young male (Touir, unpubl.). Later, in the anterior region of the
testis, oogenesis (pre-vitellogenesis) replace spermatogenesis. The
border between the two processes of gametogenesis lies generally behind
the oviducts (Fig. 7), but is not fixed. Oogenesis continues in the
anterior region of the germinative zone throughout the male phase. It
must be emphasized that, whereas in normal oogenesis the secondary
oogonia do not multiply after leaving the germinative zone, in certain
males in their first season of reproduction the secondary oogonia do
multiply before beginning oogenesis. Thus, in these males game"togenesis
may be said to have an intersex character. The region of transition
between the ovarian and testicular parts of the gonad exhibits a very
significant distribution of the cells: the oocytes in the pachytene
stage and the spermatocytes are never intermingled, and form two dis-
tinct groups. This separation proves that the fate of the germ cells
is determined in the germinative zone.
Fig. 7. Genital apparatus of Lysmata seticaudata in male phase (0') and

in female phase (~). Cd vas deferens; Ga androgenic gland; L limit
between oogenesis and spermatogenesis; Og male genital aperture;
Ovd oviduct; Ovg oogenesis (in the anterior part of the gonad) ;
Spg spermatogenesis; Zg germinative zone
99
As the AG degenerates, oogenesis spreads posteriorly, approximately

as far as the branching-off of the spermiducts (Fig. 7). Nevertheless,
the most posterior gonia continue to develop into spermatozoa even in
old females, i.e. even in the complete absence of AH. Lysmata differs
from Pandalus in this respect. The disappearance of the appendix mascu-
lina has been observed (Charniaux-Cotton, 1960b): the nuclei of its
cells suddenly become pycnotic at stage Do of the last intermolt.
Therefore, the maintenance of these cells requires AH when the moult-
ing hormone is acting.
The study of gametogenesis in L. seticaudata shows that there is a gra-

dient of response to the AH along the germinative zone. The anterior
region gives rise to oogenesis even in the presence of AH; the region
between oviducts and spermiducts gives rise to oogenesis after the
disappearance of AH; the posterior region continues spermatogenesis
in the absence of AH. This gradient is similar in Talitridae, except
for the posterior region.
IV. Studies on Control of ~he Disappearance of AG
1. Role of the Eyestalks
In decapods, the X organ - sinus gland (neurosecretory complex) is

located in the eyestalks. It is well known that this structure secretes
a hormone that inhibits the moulting glands and another hormone that
inhibits vitellogenesis.
The bilateral removal of the eyestalks from male Pandalus platyceros was
performed when the AG were in the process of atrophy and the gonads
were changing into ovaries. Six to eight weeks afterwards, the AG were
hypertrophied and the male stage had been maintained. The gonads show
evidence of spermatogenesis (Hoffman, 1968). Similar experiments have
been performed in L. seticaudata (Touir, 1973); afterwards the animaJs
were kept for a period of eight months. The external male characters,
which already regressed at the time of the operation, redeveloped.
The AG had hypertrophied and spermatogenesis was very active. Thus,
the bilateral ablation of the eyestalks inhibited sex reversal for a
substantial time, perhaps permanently.
It is possible that increased synthesis of crustecdysone mediates, at

least in part, the effect of des talking on AG activity. It was recalled
that after destalking", the molting glands become active). Indeed, in-
j ection of crustecdysone (B-ecdysone) in L. seticaudata during the last
intermolt of the male phase produces the same effect as destalking
(Touir, unpubl.): the animal molts three times before dying, and the
appendix masculina (previously quite reduced) gain substantially in
size. In P. platyceros, the administration of B-ecdysone and destalking
have the same effect: increase in RNA synthesis in the AG (Foulks and
Hoffman, 1974). However, it should be emphasized that sex reversion
occurs in summer when the molt cycles are short, i.e. when high titers
of crustecdysone are frequent (stage D). Therefore, a lack of molting
hormone could not be responsable for the degeneration of AG.
There are two possible explanations for the inhibition of sex reversal
after eyestalk ablation:
1. The AG disappear in the grown male, when cell degeneration is no

longer counterbalanced by mitosis. Ablation of the eyestalk eliminates
its permanently moderating effect on the AG; the mitoses remain numer-
ous and the glands may persist.
100
2. The eyestalks release a specific hormone during- the second season

of genital activity which causes cell degeneration in the AG.
Each hypothesis has been examined.
1. Cellular degeneration is a constant phenomenon in AG, because se-

cretory activity leads to formation of cytolysosomes and pycnotic nu-
clei (Charniaux-Cotton et al., 1966; Payen, 1972). During the male
phase of PandaZus and Lysmata, the AG are apparently too small to sup-
press oogenesis completely. The proof is that after a graft of supple-
mentary AG in L. setiaaudata, spermatogenesis completely displaces oogen-
esis in the ovarian region of the gonad (Touir, unpubl.). During the
second season of active spermatogenesis, the demand for AH increases
because the animal is larger and, accordingly, degeneration of actively
secreting cells becomes all the more rapid.
The neurosecretory complex, X organ - sinus gland, exerts a moderating

effect on growth and activity of AG. When crabs are destalked, the AG
generally hypertrophy (Payen et al., 1971). In young crabs, this hyper-
trophy correlates with precocious genital apparatus development and
spermatogenesis (Demeusy, 1960; Payen et al., 1967, 1969). In males
of L. setiaaudata, destalked during the winter, the AG become very hyper-
trophied and spermatogenesis is more vigorous than in summer (Touir,
1973). The effect of des talking during summer is less marked - the
moderating effect of neurosecretion is weaker in summer than in winter.
Males of P. boreaZis, destalked during their first reproductive season,
also show AG hypertrophy but spermatogenesis halts about one month
after the operation (Charniaux-Cotton, 1970). The interpretation of
this finding is difficult, and more experiments are necessary.
In summary, eyes talk ablation at the time of sex reversal de-inhibits

mitosis and secretory activity in AG, and averts the regression of
these glands.
2. As for the second hypothesis, the results of AG-graft experiments

show that AG degeneration is not due to the release of a specific
factor by the eyestalk, at the time of sex reversal. These experiments
are now discussed.
2. AG Grafts at the End of the Male Phase
Young male AG were introduced in males of L. setiaaudata at the end of

the male phase. Three months later the grafts were not degenerate,
although some of the autochthonous glands were. Sex reversal was inhib-
ited (Charniaux-Cotton, 1960b). When three or more glands were intro-
duced in an individual, spermatogenesis (which had receded posteriorly
to the spermiducts) advanced to the oviducts (Touir, unpubl.).
These experiments show that there is no factor causing degeneration of

AG at the time of sex reversal.
3. AG Grafts during the Female Phase
AG grafted in females of Lysmata and PandaZus do not degenerate, and the

host is masculinized: external male characters develop and vitello-
genesis is inhibited (Charniaux-Cotton, 1959, 1965). If several glands
are implanted in Lysmata, the secondary gonia between the oviducts and
spermiducts multiply (Touir, unpubl.). Thus, there appears to be no
factor causing AG degeneration in females.
101
It is worth noting that after destalking, the females of Lysmata are

masculinized more rapidly than are the normal females (Charniaux-
Cotton, 1970). This is explained by the absence of the moderating ef-
fect of neurosecretory complex on mitosis and secretion in AG.
V. Conclusions
The gonads of hermaphroditic prawns possess but a single germinative

zone. Oogenesis occurs during the male phase, as it does in the testes
of Talitridae (except O. gammareUus) , perhaps because the moderating ef-
fect of the eyes talks on AG activity is overstrong. The disappearance
of the AG occurs at the second season of male genital activity, prob-
ably because cell degeneration is proceeding faster than mitosis.
In absence of AR, external male characters do not persist, and the
gonads self-differentiate into ovaries. However, in Lysmata, spermato-
genesis persists in the absence of AR. As has been shown below, this
situation is general in decapods.
D. Gynandromorphism
I. Description
Gynandromorphism appears to be limited to the decapods, and is a rare

phenomenon. It mayor may not be bilateral. It has been observed in
Macroura: Homarus, Palinurus, Jassus, Cambarus; in Anomoura: Eup~urus; and
in crabs: Uca, Carcinus, Potamon (Juchault, 1966; Farmer, 1972).
Farmer (1972) has recently described in detail, a bilateral gynandro-

morph specimen of Nephrops norvegicus, the sole example found in examin-
ing approximately 40,000 individuals. "The left side of the animal is
completely female and the right side completely male with respect to
both the internal and external anatomy". An AG must have been present;
unfortunately Farmer did not examine the extremity of the spermiduct.
If the endocrine control of sex differentiation was similar in amphi-

pods and decapods, the single AG of this gynandromorph specimen of
Nephrops would have been responsible for the transformation of ovary
into testis. In fact, recent results obtained in our laboratory show
that decapods are different from amphipods in this respect.
II. Control of Sex Differentiation in Decapods
An ovary or a testis implanted in an animal of the other sex does not

undergo transformation, even if it is immature and contains only gonia.
Instead, gametogenesis occurs normally. This is the case in Palaemon
serratus, Carcinus maenas, Pontastacus leptodactylus (Payen, 1974a). This is
also the case in the posterior extremity of the Lysmata gonad, which
continues spermatogenesis during the female phase.
In general, the bilateral graft of AG does not masculinize females.

A single result has been obtained in Carcinus maenas (Charniaux-Cotton,
1958b), and another in Rhitropanopeus harrisii (Payen, 1969): masculiniza-
tion of pleopods occurred on only one side, i.e. the operated females
were transformed into bilateral gynandromorphs.
A careful study of sexual morphogenesis has been made by Payen in crabs

(1974b) and crayfish (1973). The first post-embryonic stage in crayfish
102
and the first post-larval stage in crabs are both sexually undifferen-
tiated. Sexual differentiation first appears as external sex characters
in crabs, and as gonads in crayfish; oogenesis is more precocious than
spermatogenesis. While the male gonad retains its embryonic structure
(the gonia are encompassed by somatic cells), the oocytes begin their
previtellogenic growth in the ovary. The AG have not been demonstrated
until after the differentiation of the male gonads has begun, whereas
the sexual dimorphism of the pleopods is (notably in crabs) already
well-developed at this stage.
Experimental and descriptive studies in decapods suggest that the AG

of decapods (unlike those of o. gammareZZus) do not initiate sex differ-
entiation. However, they probably induce puberty. The simplest expla-
nation of sex differentiation in genetic males is that it results from
the purely local action of an androgen, synthesized in the pleopod An-
lagen and in somatic cells of the genitals. It is emphasized that local
synthesis of androgen in the Anlagen is just as plausible in genetic
males, as synthesis would be in a discrete gland. Gonia and the pleopod
territories in decapods could not be acted upon by a circulating hor-
mone, as AG grafts in young females have shown.
III. Conclusion
Bilateral gynandromorphism is probably of genetic origin: loss of the

M-chromosome in a blastomere in early embryonic cleavage; or perhaps
a double fertilization. As has been shown, the female half of the ani-
mal cannot respond to the single AG. The local androgen inducer hypo-
thesis would explain gynandromorphism in decapods, and also tends to
be supported by the fact that gynandromorphism is a phenomenon unique
to decapods.
E. Discussion and Conclusion
Results obtained in TaUtridae explain both the presence of oogenesis in

testes of many species, and functional hermaphroditism (not uncommon)
in Malacostraca. The AG completely control male differentiation and
are capable of transforming even mature females into males. Conversely,
the ablation of AG produces the inversion of males to females. Func-
tional hermaphroditism itself is the consequence of a total absence
of AG (protandry - more frequent) or of the delayed appearance (proto-
gyny) of AG. In this respect, endocrine control of sex characteristics
in hermaphroditic decapods i$ similar to that in TaUtridae. This herm-
aphroditism is probably secondary (see p. 92).
In gonochoristic crabs, a different mechanism (local inducers) appears
to bring about male sex differentiation. In such decapods, spermatogonia
are definitively determined and spermatogenesis takes place without AH.
Functional hermaphroditism evidently cannot occur in such groups. It
is noted that Lysmata is an intermediate case, because the posterior
extremity of the gonad never undergoes sex reversal. Malacostracans,
whose sex differentiation is governed by local inducers (androgens),
can be gynandromorphic, but only in the presence of genetic abnormali-
ties.
103
References
Allen, J.A.: On the biology of Pandalus borealis Kroyer with reference

to a population of the Northumberland coast. J. mar. biol. Ass.
U.K. 38, 189-220 (1959).
Aoto, T. -;-Sexual phases in the prawn, Pa:ndalus Kessleri Czerniavski, with
special reference to reversal of sex. J. Fac. Sci. Hokkaido Univ.
6 (Zoo1.) 2, 1-20 (1952).
Baffoni, G.M.: Osservazione sulla transformazione del sesso nei Crosta-
cei Decapodi. PUbbl. Stne zool. Napoli 21, 132-147 (1947).
Berreur-Bonnenfant, J.: Etude comparee du testicule de quelques Crusta-
ces gonochoriques ou hermaphrodites. Arch. Zool. exper. gen. Fr.
ill, 375-395 (1971).
Berreur-Bonnenfant, J., Charniaux-Cotton, H.: Hermaphrodisme proteran-
drique et fonctionnement de la zone germinative chez la Crevette
Pa:ndalus borealis Kroyer. Bull. Soc. Zoo1. Fr. 90, 243-259 (1965).
Carre-Lecuyer, M.-C.: Etude histologique du testicule chez le Crustace
Amphipode Talorchestia deshayesei Audouin. C. R. Soc. Biol. Fr. 164,
512-516 (1970). -
Charniaux, H.: Castration chirurgicale chez un Crustace Arnphipode
(Orchestia gaJrrmarella) et determinisme des caracteres sexuels secondaires.
Premiers resultats. C.R. Acad. Sci. 234, 2570-2572 (1952).
Charniaux-Cotton, H.: Decouverte chez un Crustqce Arnphipode (Orchestia
gammarella) d' une glande endocrine responsable de la differenciation
des caracteres sexuels primaires et secondaires males. C.R. Acad.
Sci. 239, 780-782 (1954).
Charniaux-Cotton, H.: La glande androgene de quelques Crustaces Deca-
podes et particulierement de Lysmata seticaudata, espece a hermaphro-
disme proterandrique fonctionnel. C.R. Acad. Sci. 246, 2814-2817
(1958a). -
Charniaux-Cotton, H.: Controle hormonal de la differenciation du sexe
et de la reproduction chez les Crustaces superieurs. Bull. Soc.
Zoo1. 83, 314-336 (1958b).
Charniaux-Cotton, H.: Masculinisation des femelles de la Crevette a
hermaphrodisme proterandrique Lysmata seticaudata, par greffe de
glandes androgenes. Interpretation de l'hermaphrodisme chez les
Decapodes. Note preliminaire. C.R. Acad. Sci. 249, 1580-1582 (1959)
Charniaux-Cotton, H.: Sex determination. In: Physiology of Crustacea
(ed. T.H. Waterman), vol. 1, pp. 411-447. New York: Academic Press
1960a.
Charniaux-Cotton, H.: Physiologie de l'inversion sexuelle chez la
Crevette a hermaphrodisme proterandrique fonctionnel, Lysmata seti-
caudata. C.R. Acad. Sci. 250, 4046-4048 (1960b).
Charniaux-Cotton, H.: Determinisme de l'intersexualite chez les Crust-
aces superieurs et particulierement chez les Arnphipodes Talitridae.
Bull. Soc. Zoo1. 87, 338-366 (1962).
Charniaux-Cotton, H.:-Hormonal control of sex differentiation in Inver-
tebrates. In: Organogenesis (ed. R. de Haan, H. Ursprung), pp. 701
-740. New York: Holt, Rinehart and Winston 1965.
Charniaux-Cotton, H.: Sexualite et activite genitale male chez les
Crustaces superieurs. Bull. Soc. Zool. 95, 565-594 (1970).
Charniaux-Cotton, H.: Recherches recentes sur la differenciation sexu-
elle et l'activite genitale chez divers Crustaces suerieurs. In:
Hormones et differenciation sexuelle chez les Invertebres (ed. E.
Wolff), vol. 3, pp. 127-178. Paris-London-New York: Gordon & Breach
1972.
Charniaux-Cotton, H., Zerbib, C., Meusy, J.-J.: Monographie de la
glande androgene des Crustaces superieurs. Crustaceana ~, 113-136
(1966) .
104
Demeusy, N.: Differenciation des voies genitales males du Crabe Carci-

nus maenas L. Role des pedoncules oculaires. Cah. Biol. mar. l,
259-278 (1960).
De Nay, D.: Etude chez Orchestia mediterranea A. Costa (Crustace Amphipode)
des testicules presentant un hermaphrodisme rudimentaire et des
ovaires de femelles masculinisees. C.R. Soc. Biol. 165, 989-993
(1971). -
Farmer, A. S.: A bilateral gynandromorphism of Nephrops norvegicus (Decapoda:
Nephropidae). Marine Biol. 15, 344-349 (1972).
Foulks, N.B., Hoffman, D.L.: The effects of eyestalk ablation and B-
ecdysone on RNA synthesis in the androgenic glands of the protandric
shrimp, Pandalus platyceros Brandt. Gen. Compo Endocrinol. 22, 439-
447 (1974). -
Fried-Montaufier, M.-C.: Etude de l'hermaprhodisme de la gonade du
male chez le Crustace Amphipode, Talitrus saltator (Montagu). C.R.
Soc. Biol. 161,2104-2107 (1967).
Gilchrist, B.M~Haldane, J.B.S.: Sex linkage and sex determination
in a Mosquito, Culex molestus. Hereditas Lund 11., 175-190 (1947).
Hoffman, D.L.: Seasonal eyestalk inhibition on the androgenic glands
of a protandric Shrimp. Nature 218, 170-172 (1968).
Hoffman, D.L.: The development of the ovotestis and copulatory organs
in a population of protandric shrimp, Pandalus platyceros Brandt from
Lopez sOund, Washington. Biol. Bull. 142,251-270 (1972).
Juchault, P.: Contribution a l'etude de la differenciation sexuelle
male chez les Crustaces Isopodes. Thesis No. 57, University of
Poitiers 1966. --
Juchault, P., Legrand, J.-J.: Transformation de femelles genetiques en
males physiologiques chez les Iniscoides PorceUio dilatatus et HeUeria
brevicornis. C.R. Acad. Sci. 258, 2197-2199 (1964).
Katakura, Y.: Hormonal control of development of sexual characters in
the Isopod Curstacean ArmadiUidium vulgare. Annotationes zool. Japon
34,60-71 (1961).
Legrand, J.-J., Juchault, P.: Le contrale humoral de la sexualite chez
les Crustaces Isopodes gonochoriques. In: Hormones et differen8iation
sexuelle chez les Invertebres (ed. E. Wolff), vol. 3, pp. 179-218.
Paris-London-New York: Gordon & Breach 1972.
Noel, P.: Cycle biologique et inversion sexuelle du Crustace Decapode
Natantia Processaedulis. Cah. Biol. mar. 14, 217-227 (1973).
Payen, G.: Experiences de greffes de glandeS-androgenes chez la femelle
pubere du Crabe Rhithropanopeus harrisii (Gould) (Crustace Decapode).
C.R. Acad. Sci. 268, 393-396 (1969).
Payen, G.: Etude ultrastructurale de la degenerescence cellulaire dans
la glande androgene du Crabe Ocypode quadrata (Fabricius). Z. Zell-
forsch. 129, 370-385 (1972).
Payen, G.: Etude descriptive des principales etapes de la morphogenese
sexuelle chez un Crustace Decapode a developpement condense, l'Ecre-
visse Pontastacus leptodyct1Jlus leptodactylus (Eschscholtz 1823). Ann.
Embryol. Morph. 6, 179-206 (1973).
Payen, G.: Recherches sur la realisation et le contrale de la diffe-
renciation sexuelle chez les Crustaces Decapodes Reptantia. Thesis
CNRS No. AO 9930, University of Paris VI, 1974a.
Payen, G.: Morphogenese sexuelle de quelques Brachyoures (Cyclometopes)
au cours du developpement embryonnaire, larvaire et postlarvaire.
Bull. Mus. Hist. nat., Paris 209 (3), Zool. 139, 201-262 (1974b).
Payen, G., Costlow, J.D., Jr., Charniaux-Cotton~.: Recherches sur
le role de la neurosecretion dans la differenciation sexuelle du
Crabe CaUinectes sapidus Rathbun. C.R. Acad. Sci. 264, 2148-2151
(1967). -
105
Payen, G., Costlow, J.D., Jr., Charniaux-Cotton, H.: Mise en evidence

experimentale de l'independance de la realisation du sexe chez le
Crabe Rhithropanopeus harrisii (Gould) a l' egard du complexe neuro-
secreteur organe de Hanstrom-glande du sinus. C.R. Acad. Sci. 169,
1878-1881 (1969).
Payen, G., Costlow, J.D., Jr., Charniaux-Cotton, H.: Etude comparative
de l'ultrastucture des glandes androgenes de Crabes norrnaux et pe-
donculectomises pendant la vie larvaire ou apres la puberte chez
les especes Rhithropanopeus harrisii (Gould) et CaHinectes sapidus Rath-
bun. Gen. Compo Endocrinol. 11, 526-542 (1971).
Touir, A.: Influence de l'ablation des pedoncules oculaires sur les
glandes androgenes, les gonades, les caracteres externes males et
l'inversion sexuelle chez la Crevette hermaphrodite Lysmata seticau-
data Risso. C.R. Acad. Sci. 277, 2541-2544 (1973).
Temperature-Sensitive Intersexuality and Its Determinism In
Orchestia gammarella Pallas
T. Ginsburger-Vogel
A. Introduction
In comparison with other Tali tridae, Orchestia gamrnareUa can be qualified

as a strictly gonochoristic species. Unlike what is observed in many
other species (cf. Berreur-Bonnenfant, 1971), oocytes are practically
never found in the gonads of males. The regular occurrence of oocytes
in the testes of normal males has been known for a long time (Nebeski,
1880; Della Valle, 1893; Boulenger, 1908). This is quite different from
cases of intersexuality which have been described in several species:
O. gamrnareUa (Dahl, 1946; Charniaux-Cotton, 1953), Talitrus saltator
(Fried-Montaufier, 1968), Orchestia mediterranea (Wildish, 1970). Such
intersexuality consists in the occurrence of primary or secondary
sexual characters of one sex in individuals of the other. In O. gamrna-
rella, a study of the intersexuality transmission (Charniaux-Cotton,
1962) demonstrated the intervention of hereditary factors, it also
showed the appearance of particular phenomena, e.g. the occurrence of
offspring consisting of males exclusively.
B. Relations between Intersexuality and Sex-ratio Abnormalities
I. Description of Intersexes
Before investigating the mechanisms of transmission of intersexuality

in the reproductive process it is necessary to describe more accurately
the morphology and anatomy of intersexes. Studies on the morphology
and biometry of intersexes were carried out by Charniaux-Cotton (1953,
1957, 1962). In O. gamrnareUa, two categories of intersexes can be dis-
tinguished:
1. Functional males which bear oostegits and whose gnathopods have an

intermediary form between the female (indifferentiated) and the normal
male form.
2. Females the intersexuality of which manifests itself externally by

the occurrence of one (sometimes two) male genital apophyses.
Under natural conditions intersex females are far less common than
intersex males. From a strictly morphological point of view, intersexes
can be characterized as "oostegites-bearer males" or "females with male
genital apophysis".
Further studies (Ginsburger-Vogel, 1972a,b) have shown that not only

external characters are involved. In young intersex males, a transitory
oviduct persists during 2 or 3 intermolts after the postembryonic period
of sexual differentiation. At hatching, the male genital apparatus is
indifferentiated and traces of male and female gonoducts are present
107
in both sexes. In normal males, oviduct structures degenerate quickly,

the degeneration being slower in intersex males. Very rarely can oocytes
be observed in the testes.
Immediately after the molt which leads to the first appearance of ex-
ternally visible sex differences, intersex females possess a transitory
spermiduct, the end of which shows a pad of adipose cells, correspond-
ing to the site of the androgenic gland in males. The large majority
of these females is fertile, some. are sterile and among the latter some
undergo a spontaneous inversion of their sexual phenotype and become
sterile intersex males. The androgenic glands of such individuals are
functional and can masculinize females. Oogenesis disappears completely;
the gonads are transformed into testes.
The variability among intersexes which can be subdivided in two cate-

gories only is considerably lower when compared with Gammarus duebeni.
In this latter species, a continuous gradation of intersexual forms is
observed (Bulnheim, 1965). All kinds of intermediates between strongly
female intersexes (corresponding to females with oostegites apophysis)
and strongly male inters exes (equivalent to oostegites bearer males)
exist. Dissimilarities between species are caused in part by differ-
ences in their sexual physiology.
II. Distribution of Intersexes in Natural Populations
The distribution of intersexes in natural populations can be correlated

with some abnormalities of the sex-ratio (Ginsburger-Vogel, 1973a).
Mainly two categories of populations exist:
1. Populations of pebble beaches in which the "masculinity rate" (per-

centage of males in the whole sample) is normal (~ 50%) and where
intersexes are rare « 1%).
2. Populations of estuaries, characterized by masculinity rates much

lower than 50% (mean 25 - 30%) and a relatively high number of inter-
sex males in these (about 15% of males).
III. Occurrence of Abnormalities in the Sex-ratio
In order to study the inheritance of intersexuality, several types of

cross-breedings were carried out (Ginsburger-Vogel, 1973, 1974a,b).
In particular, crosses between females of a population with numerous
intersexes (population of Penze, near Roscoff) and males from lineages
with normal sex-ratio (amphogenous males) are mentioned here. Two se-
ries of crosses were conducted, either at 17° ± 1°C, or at laboratory
temperature (mean of 22°C). In both cases breedin~ was performed in
continuous photophase.
The comparison of the results shows that in both cases the distribution
deviates clearly from the binomial type (Fig. 1). However, the mode of
repartition is very different according to the breeding temperature.
Crosses at 22°C are distributed in two classes: 63 were amphogenous

and 37 arrhenogenous (Fig. 1A), characterized by high masculinity
rates. Among the latter 13 were perfect arrhenogenous (masculinity
rate = 100%). The distribution of masculinity rates is apparently tri-
modal with highest frequencies at 50%, 75% and 100%. This coincides
with the existence of three categories of females, amphogenous, imper-
fect arrhenogenous and perfect arrhenogenous respectively. No inter-
sexes developed from these crosses.
108
Nc.
15
10
15
B:1rc .
10
, i'i IIII1
o 25 50 75 100
Mr .
Fig. 1 A and B. Repartition of mas~ulinity rates of cro3ses: Penze ~

x amphogenous d'. CA) Crosses at 22 C. (B) Crosses at 17 C. No number
of couples; Mr masculinity rate; stippled: pairs producing intersexes
Masculinity rates of crosses conducted at 17°C are distributed in a

very different way: 43 pairs were thelygenous (weak masculinity rate) ,
11 of which were perfect thelygenous, 40 amphogenous and 'only 8 arrhe-
nogenous (2 perfect arrhenogenous). Moreover, an important proportion
of pairs (40/91) gave rise to intersexes. They are found among the
descendants of all three types of mating pairs. But the most frequent
intersex-bearer pairs are the imperfect thelygenous ones (31/32). Only
9/37 were amphogenous and 2/8 arrhenogenous; about 70% of intersex-
bearer pairs are imperfect thelyganous. The proportion of intersexes
is strong among males which derive from imperfect thelygenous pairs,
since 74% of them (113/153) were intersexes; it is weak among males
born from amphogenous (84/626 - 13%) and arrhenogenous (63/252 - 25%)
pairs. Nevertheless, there are great variations in the proportions of
intersexual males; it may occur that all male descendants or a small
number only is intersexual.
These results show the presence of several abnormalities in the sex-

ratio of populations with intersexes, that may be compared with mono-
geny phenomena described in Oniscoidea (Arcangeli, 1931; Vandel, 1935)
In particular, they draw our attention to a similar correlation be-
tween thelygeny and intersexuality. Furthermore, the observed differ-
ences according to breeding temperature invite the study of the in-
fluence of environmental factors.
109
C. Inheritance of Intersexuality and Monogeny
I. Existence of Thelygenous Intersex-bearer Females
Successive pairing with various amphogenous males allows a precise

estimation of the respective role of male and female for the thelygeny
and intersexuality of their progeny. First information is obtained
by an appropriate choice of animals for experimentation: females from
a natural population with many intersexes, males of amphogenous, labo-
ratory-reared lines (their amphogeny was confirmed in most cases and
they do not produce intersexes in particular in brother-sister cros-
ses). Females of thelygenous or amphogenous broods with intersex males
were paired with other amphogenous males under identical experimental
conditions. Our results are given in Table 1.
Table 1. Results of crosses between thelygenous intersex-bearer females

and amphogenous males
No. Progeny of the

thelyg,enous pair cross ~ (thelygen. ) cross d' (thelygen. )
xd' (amphogen. ) x ~ (amphogen.)
S5 1 d'i + 33 ~ 61 ~ 10 d' + 4~
S10 36 ~ 24 ~ 6d' + 16 ~
S11 26 ~ d'i + 16 ~ + h 11 d' + 14 ~
S14 4 d'i + 21 ~ 17 d'i + 1H + ~i 22 d' + 16 ~
S15 44 ~ 50"i + 21 ~ 13 d' + 8 ~
S16 13 d' + 8 d'i + 4~ 5d' + 13d'i + 10 ~

S19 1 d' + 4 d'i + 12 ~ 5H 10 d' + 7~
S225 4 d'i + 41 ~ 40<j> 4d' + 8<j>
S257 d'i + 52 ~ 4d' + 60 <j> 10 d' + 8<j>
S259 1d' + d'i + 22 <j> 47 <j>

S264 18 d' + 2 o"i + 4<j> 120" + 7d'i + 21 <j> 150" + 1H
33 d' + 25 d' i + 21 d' + 43 d'i + 101 d' + 92<j>

295 <j> + 2 <j> i 365 <j> + 2 <j> i
Mr. = 16,3% Mr. = 14,5% Mr. = 52,3%
Whichever amphogenous male was paired with the female, thelygeny and
intersexuality are maintained. We are therefore led to assign an es-
sential role to the female in the occurrence of thelygeny and inter-
sexuality. These females will be called "thelygenous intersex-bearer
females".
II. Matroclinous Transmission of Thelygeny
The modalities of thelygeny and intersexuality transmission were stud-

ied under the same experimental conditions by crosses between normal
or intersex males or females born from thelygenous females with ampho-
genous individuals. Fig. 2 represents the distribution of characteris-
tic masculinity rates for crosses between these females and amphogenous
110
males. Thelygeny, with a great variability, characterizes these cros-

ses, since 70 pairs from a total of 82 are thelygenous; they produced
80 males, 401 intersex males, 4203 females and 26 intersex females
(Mr. of 9,8%). The majority of males (84%) are intersex, independent
of the masculinity rate of the brood.
Nc.
Mr. , 9,79%
20
15
10
o 25 50 75 100
Mr.
Fig. 2. Repartition of masculinity rates of the offspring of 82 crosses

between daughters of thelygenous females and amphogenous males
Thelygeny and intersexuality are maintained in crosses between females

born from thelygenous females and amphogenous males. This confirms the
existence of thelygenous intersex-bearer females. The persistency of
these characters in all offspring - at present verified for three con-
secutive generations - shows that inheritance of thelygeny and inter-
sexuality is essentially matroclinous.
Normal or intersex males born from thelygenous females transmit to

their progeny neither thelygeny nor intersexuality, when they are mated
with arnphogenous females. Offspring from these crosses is either ampho-
genous or arrhenogenous. On the other hand, these males seem to be able
to modify the expressiveness of thelygeny. Indeed, an increase of mas-
culinity rate is observed when in matings with thelygenous females the
arnphogenous male is replaced by an intersex. E. g-: females born from
the same thelygenous female gave rise to 20 males, 29 intersex males
and 256 females (Mr. = 16%) when crossed with amphogenous males, but
78 males, 51 intersex males and 176 females (Mr. = 42,3%) with inter-
sex males. When, however, intersex males exert a modifying influence
on the masculinity rate, they do not suppress the transmission of thely-
geny to the females.
III. Influence of Temperature on Intersexuality and Thelygeny
In order to study the influence of environmental factors, breeding

experiments were conducted under various conditions. The influence of
temperature on the sex-ratio has been investigated by Kinne (1953) in
111
Gammarus duebeni: temperatures below 5°C produce a high percentage of

males and temperatures above 6°C a high prevalence of females. This
result was later disproved by Traut (1962). Bulnheim (1967) showed the
influence of the duration of photoperiod on sex proportion in broods
born from amphogenous females; this influence is not found in O. gamma-
rella.
The lack of thelygenous females in crosses carried out at laboratory

temperature was soon noticed in the first broods which were obtained
from a population with intersexes. Thelygenous, intersex-bearer females
at 17°C were transferred to 22°C and mated with amphogenous males.
Results of both series of crosses are ~iven in row 1 and 2 respectively
of Table 2. In crosses performed at 22 C, thelygeny and intersexuality
are almost entirely missing, which shows that these phenomena are sen-
sitive to an elevation of temperature. The females used at 22°C belong
to two categories: 14 of them produced amphogenous broods and 9 proved
to be arrhenogenous, two of them being perfect arrhenogenous.
Table 2. Temperature-sensitivity of thelygeny and intersexuality.

Comparison between crosses successively reared at 17°C and 22°C.
No.
if if Mr.
85 33 2,8 12 16 41 ,8
8 10 24 o 25 9 73,5
8 13 27 15 65,1 18 6 75
8 225 4 41 8,9 15 34 30,8
8 243 3 2 39 11 , 4 16 11 58,3
8 255 2 29 9,4 25 17 59,5
8 294 4 3 45 13,5 11 11 50
8 297 1 3 35 10,3 12 14 46,2
8 298 4 8 44 21 ,4 8 4 66,6
8 316 5 8 44 21 ,9 6 7 46,2
8 317 4 4 22 26,2 76 2 37 68
8 319 11 2 16 44,8 24 6 80
8 325 16 7 90 20,2 42 13 76,4
8 340 34 o 14 14 50
8 342 4 2 15 28,6 19 20 48,7
8 343 16 33 32,1 31 10 75,6
8 346 20 o 20 16 55,6
8 348 42 2,3 6 9 40
8 353 60 1 ,6 15 14 51 ,6
8 333 3 31 8,7 20 17 54,1
I317 6 2 24 25 16 100
I 319 6 24 24 55,6 45 100
I353 6 18 25 59 10 85,6
Total 68 126 778 20 535 3 295 64,6

112
This temperature sensitivity was found again in the study of the off-
spring of thelygenous females. Results obtained from one couple (1 319 )
and its progeny are particularly interesting for the interpretation
of arrhenogeny. At 17°C this pair - composed of an intersex male and
a female of the same population - produced 6 males, 24 intersex males,
23 females and 1 intersex female. At 22°C mating of the same female
with an amphogenous male produced 46 males and no females. The intersex
male mated with an amphogenous female gave 25 males. Females born from
this pair were mated with amphogenous males at 17°C and produced 24
intersex males and 184 females (Mr. = 11,5%). At 22°C the offspring
consisted of 137 males (Mr. = 100%). Males were coupled with amphoge-
nous females; they produced 288 males. These results are a supplemen-
tary proof of the correlation between the thelygeny of females at 17°C
and arrhenogeny of females or males at 22°C.
IV. Interpretation of Monogeny and Intersexuality Phenomena
Two kinds of interpretations were proposed to explain the existence

of thelygenous females: a peculiar genetic structure (De Lattin, 1952;
Johnson, 1961) or the influence of an epigenetic feminizing factor
(Bulnhei~, 1966; Legrand and Juchault, 1969b) independent of the genome.
The temperature-sensitivity of intersexes and thelygenous females re-

presents an original character in comparison to other Crustacean species
showing monogeny phenomena linked with intersexuality. But exceptions
are known, since two thermostable lines could be isolated (Ginsburger-
Vogel, 1974a). In a related species, Orchestia montagui, thelygeny and
intersexuality are always thermostable (Ginsburger-Vogel, 1974b).
The essentially matroclinous transmission of thelygeny suggests the

existence of a "cytoplasmic factor". The transformation of thelygenous,
intersex-bearer females into normal of arrhenogenous females can be
easily interpreted if one assumes an inversion (at 17°C) of the sexual
phenotype of genetically male individuals. Intersexuality, strictly
linked with thelygeny, appears then as indication for an incomplete
feminization of the affected males. This hypothesis implies the exis-
tence of females the male genotype of which is masked at 17°C, but
expressed at 22°C.
With regard to the mechanism of determination, the question arises

whether sexual digamety exists or a polygenic mechanism as supposed
for Gammarus duebeni by Bulnheim (1969). On the basis of experiments con-
ducted with animals from a population where intersexuality is uncommon,
but where other monogeny phenomena exist, the assumption of polygenic
sex-determination was considered (Ginsburger-Vogel, 1967) for O. g@~a
rella. Further experiments (Ginsburger-Vogel, 1974c), however, disprove
the hypothesis: crosses between amphogenous females and neo-males
(amphogenous females transformed into functionarmales) give rise to
females only; three crosses produced 102 females. Analogous results
were obtained in O. cavimana, a species in which monogeny is unknown
(Ginsburger-Vogel, 1972b). It is concluded that amphogenous females
are homogametic (2AXX) and normal males heterogametic (2AXY).
Possible consequences of an inversion of the sexual phenotype of gene-

tic males are briefly considered.
2AXY-females should arise from crosses between thelygenous females and

amphogenous males at 17°C:
Thelygenous ~ (2AXX) x if (2AXY) 7 thelygenous ~ (2AXX or 2AXY) + if and
ifi (2AXY).
113
Crosses at 17°C between thelygenous 2AXY-females and normal males

should yield 2AYY-individuals:
Thelygenous Q (2AXY) x if (2AXY) 7 ~ (2AXX) + 2 ~ (2AXY) + ~ (2AYY)

+ if and if i (2AXY or 2AYY).
The existence of females with 2AXY or 2AYY genotype and males with
2AYY ~enotype explains the presence of arrhenogenous females and males
at 22 C (Ginsburger-Vogel, 1973), since thelygeny of the female is tem-
perature-sensitive:
Q (2AXY) x arnphogenous if (2AXY) 7 ~ (2AXX) + 2 if (2AXY) + if (2AYY)
(Mr. = 75%).
Q (2AYY) x arnphogenous if (2AXY) 7 if (2AXY) + if (2AYY) (Mr. = 100%) .
if (2AYY) x amphogenous ~ (2AXX) 7 if (2AXY) (Mr. = 100%) .
An attempt was made to verify this hypothesis, in particular by the

study of the offspring of females issued of pair I319, which were mated
with their brothers or with arnphogenous males. 33 females born of a
brother-sister-cross (which gave rise to 69 females) were paired sepa-
rately (at 22°C) with amphogenous males. They produced 843 males, .
1 intersex male and 1 female. From 36 females born of a cross (at 17°C)
between 3 females (derived from pair I319) and an arnphogenous male
(giving 2 intersex males and 58 females), 20 individuals proved to be
imperfect arrhenogenous at 22°C (534 males, 3 intersex males and 176
females; Mr. = 75,6%) and 16 perfect arrhenogenous (464 males, 10 inter-
sex males; Mr. = 100%). These results are in agreement with the fol-
lowing relation:
thelygenous ~ (2AYY) x if (2AXY) 7 ~ (2AXY) + Q (2AYY).
Consequently half the daughters must be imperfect arrhenogenous, the

others perfect arrhenogenous.
The proposed interpretation for the monogeny and intersexuality pheno-

mena is summarized in Fig. 3. It neglects for the moment the influence
of some males. It seems that they confer to their offspring a resis-
tance to sexual inve=3ion. Further experiments must clarify its nature
more precisely.
at 17"C. at22" C.
THELVGENOUS ~
AMPHOGENOUS ~ XX
XX, xv , VV
IMPERFECT XV
ARRHENOGENOUS ~
PERFECT 0 V
ARRHENOGENOUS.f V
r! andintersex r!
XV,VV
Fig. 3. Interpretation of monogeny linked to intersexuality in Orchestia

gammareUa
114
D. Sexual Physiology of Intersex Males and Thelygenous Females
An analysis of the cross-breeding experiments leads to an interpreta-

tion of monogeny and intersexuality by both genetic and epigenetic
factors. It does not inform us as to the mechanisms responsible for
the sexual inversion of genetically male individuals. Investigations
related to the physiology of intersex males and thelygenous females
are still fragmentary. We focus our attention on three series of prob-
lems:
a) Is there a direct action of temperature on sex-determination or

differentiation, or an epigenetic, feminizing factor? In this case,
what is the nature of this factor? We tried to answer the first ques-
tion by experimental induction of intersexuality and thelygeny.
b) Is temperature influence reversible and at which period of the bio-

logical cycle does it act? In particular, is there a critical period?
c) Have the androgenic glands of intersex males a normal function and

can thelygenous females be masculinized by implantation of androgenic
glands of normal males?
I. Experimental Transmission of Thelygeny and Intersexuality
1. Induction of Intersexuality
By grafts of organs of intersex males and thelygenous females to normal

males, it is possible to induce the development of oostegites if these
males are reared at 17°C after grafting. Testes, androgenic glands,
blood (4 to 8 ~l/male) or fragments of muscle of intersex males and
ovaries of thelygenous females were implanted or injected into normal
males. Grafting of homologous organs of normal females served for con-
trol purposes; they were entirely negative. The implantation of organs
of intersex males or thelygenous females, however, induced the appear-
ance of oostegites in some operated individuals (Table 3). The number
of successful experiments was relatively small. This is undoubtedly
related to the age of the recipient: oostegites develop more rapidly
in younger than in older males. The induction of oostegites by implan-
tation of androgenic glands from intersex males confirms earlier ob-
Table 3. Induction of intersexuality by grafts or injections of various

organs of intersex males or thelygenous females in normal males
Implanted Total number Number of Control

organ positive cases experiments
Testis 24 9 0/13
Androgenic
gland 24 8 0/15
Muscle of
intersex d' 21 7 0/15
Blood of
intersex d' 21 8 0/7
Ovary of
thel. <;> 31 15 0/9
115
servations by Charniaux-Cotton (1957, 1962). Induction by other organs

from intersex males or thelygenous females are similar to those obtained
in AY'madillidium vulgare by Legrand and Juchaul t (1 969a). In particular,
the inducing factor is not limited to sexual organs. Moreover, it is
possible to induce intersexuality by grafting organs of normal males
which were transformed experimentally into int:ersexes. 8 positive cases
after graft of testis (4), seminal vesicle (1), androgenic gland (2)
or muscle (1) were observed.
2. Induction of Thelygeny
By the same method 6 we tried to transmit thelygeny to amphogenous fe-

males reared at 17 C. They received ovaries, muscles or blood of thely-
genous females, or testes of intersex males. Whichever the grafted or-
gan, thelygeny was induced (Table 4). This thely~eny is temperature-
sensitive. Some females were crossed again at 22 C and became ampho-
genous.
Table 4. Induction of thelygeny and intersexuality in the progeny of

amphogenous females after implantation of organs belonging to thely-
genous females or intersex males. Mr. = masculinity rate; %i: percent-
age intersex males.
Implanted Number Results Mr. %i.

organ of cases
Ovary 17 50 d' + 9d'i + 519 Cjl + 1 Cjl i 10,2 1 5,3

Muscle 13 69 d' + 41 d' i + 299 Cjl 26,9 37,2
Blood 5 6d' + 17 d' i + 242 Cjl + Cjl i 8,7 74
Testis 22 150 d' + 41 d'i + 518 Cjl + Cjl i 26,9 21 ,4
Both series of experiments show the possibility of an experimental

induction for intersexuality and thelygeny. They confirm the hypothe-
sis of the existence of an epigenetic feminizing factor. The nature
of this factor is known in Gammarus duebeni (Bulnheim, 1967; Bulnheim
and Vavra, 1968) as the microsporidia, Octosporea effeminans or Thelohania
hereditaria. In Armadillidiwn vulgare a "bacteroid" was described in ovaries
of thelygenous females (Martin et al., 1973). In Orchestiagammarella,
the exact nature of the factor is still unknown, although microsporidia
were observed in ovaries of thelygenous females.
II. Action of Temperature
Grafting experiments show that temperature is not the primordial fac-

tor determining thelygeny and intersexuality. Nevertheless, is should
be possible to determine what are the mechanisms of action of temper-
ature, if a critical period exists and if temperature action is re-
versible. A priori, three periods seem possible for a temperature action:
oogenesis, embryonic development or post-embryonic development of
sexual organs. In O. gammarella sex is indifferentiated at birth (Char-
niaux-Cotton, 1959; Hort-Legrand et al., 1974; in particular testes of
normal males can be inverted into ovaries only when they are grafted
into a female during one of the first two intermolts after hatching
(Hort-Legrand et al., 1973).
116
1. Determination of the Period of Temperature Sensitivity
The determination of the period of temperature sensitivity was estab-

lished by three experiments:
a) Isolated thelygenous females stayed at 25°C for 40 days; they were

then transferred to 17°C and mated with arnphogenous males. They pro-
duced amphogenous or arrhenogenous broods without intersexes.
b) Ovigerous thelygenous females were placed at 25°C from laying until

hatching of eggs and the young were reared at 17°C. In this case,
broods have a normal masculinity rate.
c) Ovigerous thelygenous females were reared at 17°C and the progeny

was transferred to 25°C immediately after hatching. In this case,
broods are thelygenous.
These three experiments demonstrate that temperature can exert an in-

fluence during oogenesis or embryonic development, but that there is
no direct action on sexual organogenesis.
More precise experiments were carried out by varying 5he duration of

exposure of females or embryos to a temperature of 25 C.
2. Action on Females
The duration of rearing thelygenous females at 25~C and subsequent

mating with arnphogenous males at 17°C exerts a clear influence on mas-
culinity and percentage of intersex males of broods born of such cros-
ses (Fig. 4). The masculinity rate is low for expositions at 25°C be-
tween 5 - 10 days; it increases rapidly for exposures of 10 - 25 days,
then it becomes stable at a value between 65 and 75% for long exposures
(these high values are explained by the presence of several arrheno-
genous females). On the other hand, percentages of intersex males,
which remain constant for exposures of 5 - 10 days decrease rapidly
for exposures between 10 and 25 days and finally disappear.
This experiment explains the great variability of masculinity rate in

the broods obtained from thelygenous, intersex-bearer females of the
natural population (cf. Fig. 1). The maintenance of thelygeny during
the year is probably caused by the revertibility of short actions of
temperature.
3. Action during Embryonic Development
Magniette (unpubl.) studied effects of various durations of exposure

time at 25°C during the embryonic development of-eggs from thelygenous
females that were mated at 17°C. At present the results obtained do
not indicate a particular sensitivity to temperature during a definite
stage of embryonic development.
The mean masculinity rate of broods depends upon the duration of ex-
posure at 25°C. This does not seem to have any influence on male inter-
sexuality. On the other hand, expositions of any duration at 25°C after
hatching have no effect on masculinity rate. The hi~hest values of
masculinity rate are attained for expositions at 25 C when they cover
the entire period of embryonic development. These results prevent at
present all attempts to determine the inversion stage more precisely.
117
TI I.
•
,," 1
I.
\
T" / \.
•1 \
\
'-J
"-.
0~----r-----.-----.------r-----4--~~r-----.-----'--
o 5 10 15 20 25 30 35 40
days at 25°C.
Fig. 4 A and B. Evolution of masculinity rate and percentage of inter-

sex males in the progeny of thelygenous temperature - sensitive females
in relation to the duration of exposition to 25°C. (A) Percentage of
intersex males; (B) Masculinity rate
III. Temperature and Sexual Physiology of Intersex Males and Thely-

genous Females
The sexual physiology of thelygenous females and intersex males was

compared with that of normal individuals. Our studies were based on
former investigations by Charniaux-Cotton and her coworkers (cf. Char-
niaux-Cotton, 1970).
1. Androgenic Glands of Intersex Males
We tried to find out whether androgenic glands of intersex males are

able to masculinize normal females as do those of normal males. Grafts
of androgenic glands of normal or intersex males into females were per-
formed simultaneously at 17°C and 25°C.
At 25°C, there is no difference between the two sorts of glands: in

both cases females are masculinized.
118
At 17°C, the masculinization is slower than at 25°C with androgenic

glands from normal males. Masculinization is less frequent and always
very slow with androgenic glands of intersex males.
The masculinizing action of androgenic glands of intersex males seems

to be more or less inhibited by the temperature-sensitive feminizing
factor.
2. Masculinization of Thelygenous Females
Normal females are quickly transformed into males since the first signs
of masculinization can be noted after the first molt which follows the
graft of androgenic gland, if it is carried out at the beginning of
the intermolt (Charniaux-Cotton, 1954). Thelygenous females are nor-
mally masculinized if they are reared at 25°C after the graft. But
they show resistance to masculinization at 17°C after a period of
6 months at 17°C (continuation of vitellogenesis). Two out of 25 0 graft-
ed animals showed a beginning of masculinization.
It can be shown that the duration of exposure to 25°C influences the

masculinization of thelygenous females. 5 groups of 10 thelygenous
females received androgenic glands of normal males (2/female), and
were then exposed to 25°C during geriods of 5, 10, 15, 20 or 25 days.
Thereafter they were reared at 17 C. For short periods (5 and 10 days),
no case of masculinization is observed, but for periods longer than
15 da~s an important proportion of females was masculinized. Exposure
at 25 C for more than 10 days seems to suppress an inhibition of mas-
culinization.
From our studies on Orehestia gamnareUa it must be concluded that inter-

sexuality of secondary sex characters and even a total inversion of
the sexual phenotype of genetical males is not the expression of an
eventual disequilibrium at the level of sex-determining genes. It de-
pends rather on the presence and the action of an epigenetic feminizing
factor which is temperature-sensitive in O. gamnareUa. This factor pro-
vokes important changes in the sexual physiology of affected males.
The nature and mode of action of this factor requires further inves-
tigation.
References
Arcangeli, A.: SuI rapporto nurnerico dei sessi negli Isopodi terrestri.
Boll. Musei Zool. Anat. Compo Univ. Torino !1_(13), 1-33 (1931).
Berreur-Bonnenfant, J.: Etude comparee du testicule de quelques Crusta-
ces gonochoriques ou hermaphrodites. Arch. Zool. expo gen. l1£,
375-395 (1971).
Boulenger, C.: On the hermaphroditism of the Arophipod Orehestia deshayesei
AUqouin, pp. 42-47. Proc. Zool. Soc. London (1908).
Bulnheim, H.P.: Untersuchungen liber Intersexualitat bei Gammarus duebeni
Lil. Helgol. wiss. Meeresunters. 1£, 349-394 (1965).
Bulnheim, H.P.: Mikrosporidien-Infektion und Geschlechtsbestirnrnung bei
Gammarus duebeni. Zool. Anz. (Verh. Deutsch. Zool. Ges.) Suppl. Bd.
30, 432-442 (1966).
Bulnheim, H.P.: tiber den EinfluB der Photoperiode auf die Geschlechts-
realisation bei Gammarus duebeni. Helgol. wiss. Meeresunters. 12.,
69-83 (1967).
119
Bulnheim, H.P.: Zur Analyse geschlechtsbestirnrnender Faktoren bei Gmrnarus

duebeni (Crustacea, Arnphipoda). Zool. Anz. Suppl. Dtsch. ~, 244-260
(1969) .
BulnheDm, H.P., Vavra, J.: Infection by the microsporiden Octoporea
effeminans Sp. N., and its sex-determining influence in the Arnphipod
Gammarus duebeni. J. Paras. 54 (2), 241-248 (1968).
Charniaux-Cotton, H.: Phenomenes d'intersexualite chez un Crustace
Arnphipode (Orchestia gammareUa Pallas), pp. 447-449, 72. Congr. A.F.
A.S., Luxembourg (1953).
Charniaux-Cotton, H.: Decouverte chez un Crustace Arnphipode, Orchestia
gammareUa d 'une glande endocrine responsable de la differenciation
des caracteres sexuels primaires et secondaires males. C.R. Acad.
Sci. 239, 780-782 (1954).
Charniaux-Cotton, H.: Determinisme des phenomenes d'intersexualite chez
Orchestia gammareUa. Premiers resultats. C.R. Acad. Sci. Fr. 245,
1665-1668 (1957).
Charniaux-Cotton, H.: Etude comparee du developpement post-ernbryonnaire
de I' apparei I genital et de la g lande androgene chez Orchestia gamma-
reUa et Orchestia mediterranea (Crustaces Arnphipodes). Autodifferen-
ciation ovarienne. Bull. Soc. Zool. 84, 105-115 (1959).
Charniaux-Cotton, H.: Detebmiuisme de l'1ntersexualite chez les Crusta-
ces superieurs et particulierement chez les Arnphipodes Talitridae.
Bull. SoC. Zool. Fr. 86, 484-499 (1962).
Charniaux-Cotton, H.: Sexualite et activite genitale male chez les
Crustaces superieurs. Bull. Soc. Zool. Fr. 95, 565-594 (1970).
Dahl, E.: The Arnphipoda of the Sound. Part 1. Terrestrial Arnphipoda.
Acta Univ. Lund N.F. Ard. 2, 42-6, 1-53 (1946).
Della Valle, A.: Garnrnarini del Golfo di Napoli. Berlin 1893.
Fried-Montaufier, M.C.: Description d'un cas d'intersexualite chez
un Crustace Arnphipode Talitridae Talitrus saltator (Montagu). C.R.
Soc. BioI. 162,362-365 (1968).
Ginsburger-VogeT; T.: Determination du sexe et monogenie chez Orchestia
gammareUa Pallas. C.R. Soc. BioI. ill, 23-29 (1967).
Ginsburger-Vogel, T.: Intersexualite des voies deferentes et inversion
naturelle du sexe chez Orchestia gammareUa Pallas (Crustaces Arnphi-
podes Talitridae). C.R. Acad. Sci. Paris 274, 3449-3452 (1972a).
Ginsburger-Vogel, T.: Demonstration experimentale de l'homogarnetie
femelle chez Orchestia cavimana (Heller) (Crustaces Arnphipodes Tali-
tridae). C.R. Acad. Sci. 275, 3606-3609 (1972b).
intersexualite chez Orchestia gammareUa Pallas (Crustace Arnphipode
Talitridae) • 1. PhEmomenes de monogenie dans la population de Penze.
Arch. Zool. expo gen. 114, 397-438 (1973).
Talitridae ). II. Etude des relation entre la monogenie et l' inter-
sexualite. Influence de la temperature. Arch. Zool. expo gen. 115,
93-127 (1974a).
Ginsburger-Vogel, T.: Les phenomenes de monogenie-lies a l'intersexua-
li.te et leur thermosensibilitei etude comparee chez Orchestia gammareUa
et Orchestia montagui (Crustaces Arnphipodes Talitridae). 8e Eur. Syrnp.
Mar. BioI. (1974b).
Talitridae). III. Etude des phenomenes de monogenie independants de
l'intersexualite dans la population de Bloscon. Arch. Zool. expo
gen. (1974c).
Hort-Legrand, C., Berreur-Bonnenfant, J., Ginsburger-Vogel, T.: Inver-
sion experimentale du testicule en ovaire apres implantation dans
la cavite pericardiale de femelle adulte chez Orchestia gammareUa P.
(Crustaces Arnphipodes). C.R. Acad. Sci. 276, 1891-1894 (1973).
120
Hort-Legrand, C., Berreur-Bonnenfant, J., Ginsburger-Vogel, T.: Etude

anatomique et histologique de la differenciation des gonades males
et femelles chez Orahestia gamma:roeUa Pallas (Crustaces Amphipodes)
pendant la periode post-embryonnaire de la differenciation du sexe.
Bull. Soc. Zool. Fr. 99, 521-524 (1974).
Johnson, G.: ContributioU-a l'etude de la determination du sexe chez
les Oniscoldes: phenomenes d'hermaphrodisme et de monogenie. Bull.
Biol. Fr. et Belg. 95, 177-272 (1961).
Juchault, P., Legrand, J.J.: Croisements de neo-males experimentaux
chez ArmadiUidiwn vulga:roe Latr. (Crustace Isopode Onisco'ide). Mise
en evidence d'une heterogametie femelle. C.R. Acad. Sci. Paris 274,
1387-1389 (1972). -
Kinne, 0.: Zur Biologie und Physiologie von GammCll'US duebeni. VII. tiber
die Temperatur-Abhangigkeit der Geschlechtsbestimmung. Kieler Mee-
resforsch. 9, 126-133 (1953).
Lattin, G. de:-tiber die Bestimmung und Vererbung des Geschlechts eini-
ger Oniscoideen. II. Mitteilung. Zur Vererbung der Monogenie von
Cylistiaus aonvexus, (de G). Z. indo Abst. Vererbungsl. 84, 536-567
(1952) .
Legrand, J.J., Juchault, P.: Le determinisme de l'intersexualite chez
l,es Crustaces Isopodes' terrestres; correlation entre intersexuali te
et monogenie. C.R. Acad. Sci. Paris 268, 1647-1649 (1969a).
Legrand, J.'J., Juchault, P.: Le determinisme de la monogenie chez les
Oniscoldes. C.R. Acad. Sci. Paris 268, 1774-1777 (1969b).
Lueken, W.: tiber Geschlechtsbestimmung bei Landisopoden. Z. wiss. Zool.
166, 252, 352 (1962).
Martin, G., Juchault, P., Legrand, J.J.: Mise en evidence d'un micro-
organisme intracytoplasmique symbiote de l' Oniscolde ArmadiUidiwn
vu~are Latr dont la presence accompagne l'intersexualite ou la fe-
minisation totale des males genetiques de la lignee thelygene.
C.R. Acad. Sci. Paris 276, 2313-2316 (1973).
Nebeski, 0.: Beitrage zur Kenntnis der Amphipoden der Adria. Arb. Zool.
Dist. Wien 3 (1880).
Traut, W.: Zur-Geschlechtsbestimmung bei Gammarus duebeni und verwandten
Arten. Z. wiss. Zool. 167, 1-72 (1962).
Vandel, A.: Croisements deraces geographiques de T. provisorius Rac,
donnant une descendance exclusivement male (holarrhemie). C.R. Acad.
Sci. Paris 200, 263-265 (1935).
Vandel, A.: Recherches sur la genetique et la sexualite des Isopodes.
VI. Les phenomenes de monogenie chez les Oniscoldes. Bull. biol.
Fr. et Belg. 75, 316-363 (1941).
Wildish, D.J.: Polymorphism in Orahestia mediterranea A. costa (Amphipoda,
Talitridae). Crustaceana li, 113-118 (1970).
Parasite-Induced Castration and Intersexuality in Insects
w. WOlker
A. Introduction
The most important epigenetic factors which are able to influence sex-
ual differentiation in the animal kingdom are hormones, temperature
and parasites. Whereas in the first of these factors we are concerned
with specific chemical substances, and in the second' with a single
abiotic environmental component, the third factor is a living organism
with a multitude of individual metabolic demands and metabolic achieve-
ments.
Therefore, the most important theoretical aspect of the parasite-in-

duced changes of sex characters is that these changes must be under-
stood as the consequence of a battle between two animal species, a
mutual action in the parasite-host-system. The most important practical
aspect of these parasitic changes in insects is that the host individ-
uals concerned are then incapable of reproduction, which is a source
of hope for the biological control of pest-insects by means of para-
sites.
The first investigations into parasite-induced changes of sex charac-

ters in insects began more than a hundred years ago (Wulker, 1964).
Amongst the investigations in this field since then, descriptive papers
are predominant. Very little interest has been taken in the physiolo-
gical mechanisms which lead to the depicted final morphological results,
while experimental concepts are almost non existent.
The morphological facts have been summarized and discussed in detail

in a previous paper (Wulker, 1964). In this paper I will therefore
concentrate on the damage mechanism of parasitic intersexuality and
castration and attempt to explain the functional and experimental
aspect of this theme. I will outline new concepts for future investi-
gation as well as present proved facts. This paper confines itself to
two parasite-host-systems, the physiological relationships of which
have been investigated in the last years namely, Gastromermis-Chironomus
and Mermis-Schistocerca.
B. Morphological Considerations
The system, on which I have worked (Gastromermis-Chironomus) , is presented

in Fig. 1. The figure depicts the nematode boring out from the adult
midge.
This process occurs generally along lake shores. The adult worms are
then able to reach the water, copulate and spawn. From the eggs, pre-
parasitic larvae emerge which are only 1 mm long, and which can only
survive if they penetrate into a midge larva. All larval ins tars of
122
the midge may b e infested, but the morphological change of sex charac-
ters due to the parasite can only be detected from the fourth larv al
instar onward.
lmm
~ Fig. 1. The system GastY'o-
meoos- Chironomus . (From
Wlilker, 1 970)
It can already be seen (Fig. 1) that the parasitized midge has male
genital appendages, in the form of forceps (hypopygium) and short
female antennae (arrows). Fig. 2 gives an impression of the essential
morphological changes induced by the parasite.
The front legs of the male midge have long tarsal bristles ("beard").
However, the bristles of the parasitized males are short - as in the
females.
The antennae of the male are characterized by a large spherical pedi-

cellus and an 11-segmented flagellum. The female has a smaller pedi-
cellus and only 5 segments in the flagellum while the bristles are
shorter. With regard to these characters, parasitized males are totally
feminized in most cases.
The 8th abdominal sternite is disc-like in the normal male and its
bristles are equally distributed over the surface. In the same segment
of the female, the vagina and female accessory organs (spermathecae)
are situated, culminating in a genital chamber which has characteristic
chitinous structures. On both sides of this chamber almost spherical
areas of the sternite are to be seen. In parasitized males, all inter-
mediate grades between typically male and clearly female structures
can be realized. However, spermathecae nev er occgr. In contrast to
this, in parasitized females the intermediate scale extends from typi-
cally female structures to a disc-like sternite. In this case, also
spermathecae are never developed.
The external genital appendages of parasitized males and females never

change in the direction of the other sex.
However, the ductus ejaculatorius of the parasitiz e d male may be check-

ed in its development in all graduations between nearly normal appear-
ance and total absence. In the parasitized female, the gluten gland,
which produces the gelatinous substance of the egg-clutch, is found
(if at all) as a small rudiment.
123
gluten gland
f,onl legs antennae 8th sternite genital appendoges duct .ejac.
Fig. 2. Comparison of sexual characters in normal (n) and parasitL;;ed

(p) Chironomus. (Partly from Gotz, 1964)
Whereas in the testes of parasitized males all grades between almost

normal growth ("fully developed sperms) and nearly total suppression
are to be found, the ovary of parasitized females is always a tiny,
scarcely traceable organ, about 1/300th of the normal size (Wlilker,
1961 ) .
The following statements are important for functional considerations:
Firstly, there is no indication for any change of the gonads in the

direction of the other sex, neither in the given case nor in parasi-
tized insects in general. The only exception in print (Rempel, 1940)
rests upon the false assumption, that parasitized Chironomus adults with
male genital appendages and testes, are genetic females. This error
has been rectified by Wlilker (1961) and by the author himself (Rempel
et al., 1962) on the basis of chromosome investigations. It is, in
fact, a question of genetic males with feminization of several external
characters.
Secondly, the two sexes of the host species are damaged in different
degrees. With regard to external sex dimorphic characters the males
are affected to a greater degree (up to complete sex reversal of the
antennae), while the females have, at most, an aberrant 8th sternite.
With regard to the internal organs however, the converse is true: in
the males, in some cases the testes and ductus ejaculatorius are almost
124
normal: while in the females, the ovary, gluten gland and spermathecae
are either absent, or only present as a rudiment.
Thirdly, the degree of intersexuality can be influenced in experiments.

When certain Chironomus species (where the parasite does not occur in
the field) are exposed to the parasite in the laboratory (Wtilker, 1961:
Gotz, 1964:), intermediate forms between male and female antennae corne
into being in the parasitized males - in place of total feminization.
In these cases, the distal part of the antenna is always male, while
the proximal part is female. In other genera of Chironomid midges,
such intermediate forms can also be found in the field (Thienemann,
1950: Wtilker, 1961).
Fourthly, the period of infection can be limited experimentally to

specific larval instars, if Chironomus larvae are exposed to attack
by infective larvae only for a short time (Wtilker, 1961). Early infec-
tion can lead to weak intersexuality, late infection to strong inter-
sexuality. The time-law, which Goldschmidt (1931) also applied to par-
asitic intersexuality, has just as little validity in this field as
in the case of zygotic intersexuality in insects. Moreover, neither
a correlation in the degree of damage to different characters, nor a
dependence of damage on number or sex of the parasites can be proved
(Gotz, 1964~.
C. Mechanism of Parasitic Influence and Sexual Change
The mechanism by which the parasite works can be investigated on two

levels: either on the level of "primary" influences of the parasite on
the host (e.g. changes in the fat body, hemolymph or incretory system
of the host): or on the level of sex dimorphic structures (gonads and
imaginal discs) .
I. Primary Influences of the Parasite
1. Fat Body and Hemolymph
The most important process during the maturation of the insect ovary
is the intake of proteins (so-called vitellogenines) into the growing
oocyte. The parasite also needs a large amount of proteins for its
rapid growth, and it is therefore possible to assume that the parasite
competes with the ovary for the available protein pool. In this case,
it should be possible to detect changes in the protein level either
in the fat body where (insofar as is known: Price, 1973), the vitello-
genic proteins are synthesized, or in the hemolymph, through which the
vitellogenines are transported from the fat body-to the ovary.
Gordon et al. (1973) have recently demonstrated (with the aid of poly-
crylamide disc electrophoresis) a depletion of certain vitellogenic
and non-vitellogenic proteins after infection of the desert locust,
Schistocerca, with 50 eggs of Mermis nigrescens. Thi s depletions occurred
in the fat body two weeks after infection, and in the hemolymph three
weeks after infestation. It is known from investigations by the same
authors (Gordon and Webster, 1972) that the parasite, whose gut is
reduced, has an intake in vitro of considerable amounts of 14C_ or 3H-
marked amino acids, but practically no intake of marked dipeptides.
Therefore, the authors have conceived the idea that the parasite stim-
ulates the catabolism of proteins in the fat body (alternatively that
125
it inhibits the anabolism of proteins in the fat body) and then feeds
on the amino acids accumulated in the hemolymph.
In a previous papier, Gordon and Webster (1971) showed that the carbo-
hydrate content in the hemolymph of parasitized locusts declines from
the 6th day onward after infection. They explain this relatively fre-
quent process in parasite-host systems (Brand, 1972) by the fact that
in this way the nematode assures for itself a source of energy for its
rapid development and protein synthesis, while also providing for its
free·-living phase, when it ceases feeding. Such investigations are
more difficult in the Gastromermis-Ghironorrrus- system because, for example,
the very small host larva gives no opportunity for quantitative esti-
mations within the fat body. In addition, investigations of the hemo-
lymph proteins important for the development of the ovary only seem
worthwhile, when the hemoglobins which are quantitatively preponderant
(90 - 95% of the total protein), have been removed. In view of this,
Wlilker's (1963) failure to show, by paper electrophoresis, changes of
the hemolymph protein pattern in parasitized Chironomus larvae has little
argumentative force.
In the meantime, in accordance with the findings on Sehistoeerea, we have

been able to show that in parasitized Chironorrrus larvae (4th larval
instar), some amino acids are always quantitatively multiplied and
only a few are reduced (Klibler, 1973). The carbohydrate level of the
hemolymph in the Chironorrrus was also shown to diminish after parasiti-
zation (Wlilker, Rossler and Grinzinger, unpublished data). Maier (1973)
observed in parasitized Chironorrrus larvae a greatly diminished phenolase
activity and a displacement in the relative percentage of various hemo-
cyte types. Mliller (1972) observed an increase of non-saturated C18 _
fatty acids. However, these individual results are still far removed
from giving a physiological explanation to parasitic castration and
intersexuali ty in Chironorrrus.
2. Incretory Syst.em
Some authors have described changes of the corpora allata in certain

host insects in connection with parasitization. In the SphaeruZaria-
Bombus-system, Palm (1948) noted that in the corpora allata of the para-
sitized queen-bee, growth normal for spring does not take place and
vacuoles and pycnotic nuclei are more frequent. Brandenburg (1956)
also demonstrated that in the system StyZops-Andrena, normal growth of
the corpora allata does not take place in the parasitized sand-bee,
but that the organ in this case has fewer vacuoles. Panov et al. (1972)
have described, by use of electron microscopy, how larvae of the fly
CZytionryia (parasitic in the bug Eurygaster), cause the destruction of
corpora allata cells probably due to repression of RNA-synthesis. The
damaged organ consisted mainly of connective tissue. It is not clear
whether these damages are brought about by protein deprivation, by
toxic substances, or by other means. As is well known, the corpora
allata promote the differentiation of ovaries by means of a gonado-
tropic factor. On this basis, severeal authors (Palm, 1948; Pouvreau,
1962; Stoffolano, 1967; Gordon et al., 1973; Madel, 1973) supported
the hypothesis that the ovary damage in parasitized insects could be
a secondary consequence of the inhibition of the gonadotropic factor
in question, rather than a consequence of a primary influence of the
parasite on the ovary.
The light- and electron-microscope investigations we have so far made

on the corpora allata of normal and infected Chironomus larvae (Klimmel,
1969; Wlilker and Klimmel, 1974) have not confirmed the growth inhibition
and cellular alterations mentioned above. Instead, we observed increased
126
glycogen content (Fig. 3) and more frequent autophageous vacuoles in

the corpora allata of parasitized Chironorrrus larvae and pupae. More fre-
quent, according to statistics, are also large cavities inside the
organ (Fig. 3) which sometimes occupy half of its volume and apparently
come into being through cell decomposition.
Whether these morphological alterations are equivalent to a decline

in activity of the corpora allata, can only be satisfactorily answered
by hormone titer determinations in normal and infected individuals,
during different developmental phases.
Experimental attempts also offer success. Palm, as well as Brandenburg

have tried (without success) to replicate the parasitic effects of
Sphaerularia and Sty lops with extracts of these parasites. Brand (1972)
proposed to modify the consequences of parasitization by treating the
hosts with insect hormones.
Summarizing, the analysis of primary parasite-induced changes in the

fat body, the hemolymph and the incretory system of the host insects,
promises some insight into the problem of gonad alteration. The re-
spective alterations of imaginal disc and secondary sexual characters
may be more difficult to explain in this way.
II. Effects of the Parasites on Sexual Characters
1. Gonads
a) Ovary
The statement of a parasitic "castration" does not clarify whether it

is a matter of destruction of a fully grown organ or an early suppres-
sion of its development. Also uncertain is whether the organ is rudi-
mentarY or totally absent and if degenerative processes are involved
or not.
According to electron microscopic comparisons of the ovaries of normal

and infected adult females of Chironomus, it is a question of a suppres-
sion of the ovarian development with successive lysis or degeneration
of certain ovary cells (Wlilker and Winter, 1970; Wlilker, 1971). The
differentiation of inner and outer cells at the beginning of the fourth
larval instar and the separation of the peritoneal epithelium from the
prospective follicular epithelium takes a normal course in the parasi-
tized larva. Also, typical for normal development, is the appearance
of oocyte-nursecell-complexes which are connected by fusomes (cf.
Wlilker and Winter, 1970).
The decisive influence of the parasite is the fact that it prevents

the formation of invaginations in the follicle epithelium, i.e. the
morphogenetically fundamental process in the formation of the ovarioles.
Accordingly, all subsequent processes (e.g. migration of oocytes and
nurse cells into the invagination and formation of the egg passage)
fail to occur. The follicle epithelium disintegrates, with typical
morphological changes (e.g. occurrence of vesicular smooth endoplasmic
Fig. 3. Corpora allata cells with high glycogen content (top) and cav- ~
ity (bottom) with collapsed membranes (Mb) and axon (Ax) in corpora
allata of parasitized Chironomus larvae (4th larval instar). C cavity,
N nucleus. (Photo G. Klimmel)
127
128
reticulum, swelling of mitochondria). The inner ceils of the ovary

(oocyte-nurse cell-complexes) may continue growing to a certain degree.
However, at least a few of them degenerate accompanied, for example,
by vacuolozation, chromatin condensation, ultimately leading to final
necrotic damage. Instead of its normal straight surface, the ovary
wall exhibits atypical folding.
b) Testes
According to Wlilker's (1961) statements, the testes of the infected male

adults do not achieve full growth, but produce in some cases normal
sperms.
As a result of electron microscope studies on normal and damaged testes

(Elsenhans, unpubl.) we have since learned that smaller dimension and
reduced sperm production may be explained by the fact that the marginal
tissue of the testes (which apparently has the nourishing function for
spermatocytes) is less developed. In testes in which the development
has been earlier arrested, the sperms may not achieve full differen-
tiation .(Fig. 4) and, in certain places, they are integrated with de-
generative 'processes.
Fig. 4. Testes of a parasitized male ChironoTITUs adult. Developing sperms

(arrows), some of them integrated in dark degenerating complexes.
(Photo U. Elsenhans)
129
2. Imaginal Discs
The differentiation of all sexual characters, demonstrated in Fig. 2,

results from the growth of imaginal discs. The formation of these discs
begins with invaginations of the hypodermis in different parts of the
larval body (4th larval instar). Wlilker and G6tz (1968) have demon-
strated in detail the growth of these imaginal disc in the Chironomus.
The front legs of the adult originate from discs of the first thoracic
segment and the antennae from anlagen, which are buried deep in the
head and first thoracic segment of the larva. In the male, there is
only one genital imaginal disc in the 9th abdominal segment, the oral
part of which forms the ductus ejaculatorius and the caudal part the
male genital appendages and adjacent structures. On the other hand,
in the female, three successive anlagen are present: in the 8th abdom-
inal segment, there is an imaginal disc, which forms vaginal struc-
tures, oviduct and spermathecae; in the 9th segment an imaginal disc,
which forms the gluten gland; and at the base of the anal feet a paired
anlage, which forms the female genital appendages (cerci).
Wlilker (1961) published the first descriptions of parasitogenic changes

of imaginal discs, with Chironomus as the object of his investigations.
Bailey and Gordon (1973) have recently depicted the suppression of
imaginal discs in Aedes, after infection with Reesimermis.
In the case of Chironomus, some imaginal discs (anlagen of the female

genital appendages, wings, middle and hind legs) are never affected
by parasites. In others (anlagen of front legs) the effect must be
restricted to limited cell areas, from which the bristles of the tarsi
emerge. In other cases (male and female genital imaginal disc of 9th
segment), only the oral part of the disc is affected. A total change
of the developmental program, from male to female, occurs in the an-
tennal imaginal disc of the infected male larva. In males, the para-
sitic influence may lead to the appearance of an imaginal disc in the
8th abdominal sternite, which is never present in the normal male.
This situation is a parallel to the thermally induced intersexuality
in Aedes (Anderson, 1967).
We have more precise information as to the damage of the genital imag-

inal discs of the 9th abdominal segment. Fig. 5 compares the gluten
gland anlagen of normal and parasitized larvae of the same age. The
anlage in normal development grows tongue-like in an oral direction,
while in the caudal region, the aperture of the gland is surrounded
by a protrusion. In the infected individual, it is only the region of
this aperture which is more or less normally developed. Longitudinal
sections (Fig. 6) make it clear that epithelial cells (typical for
imaginal discs), are present in the caudal and distal parts. The gland
lumen, with its typical microvilli border, is also to be seen here.
Separated, and more proximal from the epithelial cells, is a complex
of loosely arranged cells, surrounded by irregular foldings, which
seem to be connected to the basement lamina. The origin and possible
homology of these cells is not known. Nerve sections may be present
(Fig. 6).
As the front part of the gluten gland disc is never present (whereas
the hind part is more or less normal), this indicates differential
competence of different parts of the disc for the parasitic influence.
Also, in the male imaginal disc of the 9th abdominal segment, the
front part (which forms the ductus ejaculatorius) is competent for
the parasitic influence (Fig. 7), whereas the hind part is undisturbed
and produces genital appendages.
130
Segment vm
Gluten gland in segment ]X
Fig. 5. Female genital imaginal discs of Chi ronomus (8th and 9th abdom-
inal segment), phase 6 of 4th larval instar; normal (left) and para-
sitized (right) larva. Suppression of front part of the gluten gland
after parasitization
This differential competence for the parasitic influence is not sur-

prising. The extensive experimental studies on the imaginal discs of
other dipteran genera, which Ursprung and Nothiger (1972) summarized
in their outstanding book of 1972, indicate that the determination of
individual disc regions can occur separately and at different points
of time.
The findings I have reported here are fragmentary, insofar as the de-
velopmental capacities of the chironomid imaginal discs are totally
unknown. The first attempts to mechanically produce imaginal disc de-
fects, comparable to parasitic damage (Wlilker, umpubl.), indicate a
considerable ability to regenerate parts of the disc. The parasite
seems to be able to suppress these formative properties for periods
of time.
An other experimental possibility would be to transplant imaginal discs

of Chironomus. In this way we could examine for instance, whether an
iJmaginal disc of an uninfected donor will be suppressed in its further
development after transplantation in a parasitized larva, or vice versa;
131
to what extent the suppression of the disc of an infected donor can be

revoked in a healthy larva.
Fig. 6 a-d. Female genital imaginal disc of Chironomus (9th abdominal

segment), sagittal sections, phase 7 of 4th larval instar, parasitized.
(a) General view; (b) hind part; (c) front part, border betwe en compact
and loosely arranged cells; (d) loosely arranged cells with nerve ( N).
f foldings connected with basement lamina, mv microvilli in the gland
lumen
D. Gene tic and Epigenetic Factors
Insofar as concerns the insect groups here in question (Dipte ra, Hymen-
optera, Rhynchota, Orthoptera) we can proceed from the assumption of a
genetic sex determination; in the case of Chirnonomus this has bee n
proved (Beermann, 1955).
Parasites apparently have the ability to interfere in the differentia-

tion of sexual characters (which is a consequence of this genotypical
sex determination), sometimes as far as complete sex-reversal of some
characters.
In the case of Chironomus , this change is even possible when the imag-
inal discs have already begun to grow in the 4th larval instar. Gotz
132
(1964) has shown that experimental infection during the 4th larval
instar may still lead to intersexuality. When, on the other hand, the
host larva is able to encapsulate an adolescent worm, a normal midge
results. The decisive-damaging influence seems to take place in a
short period during the fourth larval instar and can only be achieved
by fully grown parasites.
Fig. 7. Male genital imaginal disc of Chironomus (9th abdominal segment),

phase 7 of 4th larval instar. Normal (left) and parasitized (right)
larva. Suppression of ductus ejaculatorius after parasitization
Rempel (1940) assumed that in the case of Chironomus, the parasite feeds
on the ovary and in this way eliminates the source of a female gonadal
hormone. On the basis of the host's bisexual potency, this would clear
the way for the realization of male characters (testes and genital ap-
pendages). As the result of these assumptions, Rempel developed theo-
retical conceptions of the consequent disturbance of the gene balance.
However, as the presuppositions for these considerations ("every indi-
vidual starts out as a female") were wrong, they are not tenable.
In the findings which have so far been made about parasitic intersexu-
ality and castration in insects, I cannot see any indication for the
existence of gonadal hormones. On the contrary, the occurrence of fe-
male antennae and a female imaginal disc on the 8th segment in a genet-
ically male intersex (which may have nearly normally developed testes)
can be taken as counter-evidence for an hormonal control of the exter-
nal sexual characters.
133
The question as to whether gonadal hormones are present in Chironomus,

can only be answered by gonad transplantation. Such transplantations
have already been successfully completed, but the period of survival
of the insects under treatment was too short for satisfactory obser-
vations of the effects on the donor and acceptor gonads.
The importance of experimental and physiological considerations for

further clarification of parasite-induced castration and intersexuality
in insects, cannot be overstressed.
References
Anderson, J.F.: Histopathology of intersexuality in mosquitoes. J. expo

Zool. 165, 475-496 (1967).
Bailey, D.H., Gordon, R.: Histopathology of Aedes aegypti (Diptera: Culi-
cidae) larvae parasitized by Reesimermis nielseni (Nematoda: Mermithi-
dae). J. Inv. Path. il, 435-441 (1973).
Beermann, W.: Geschlechtsbestirnrnung und Evolution der genetischen
Y-Chromosomen bei Chironomus. Biol. Zbl. 74, 525-544 (1955).
Brand, Th.v.: Parasitenphysiologie. Stuttgart: Fischer 1972.
Brandenburg, J.: Das endokrine System des Kopfes von Andrena vaga Pz.
(Ins., Hymenopt.) und Wirkung der Stylopisation (Stylops, Ins.,
Strepsipt.). Z. Morph. Okol. Tiere 45, 343-364 (1956).
Gotz, P.: Der EinfluB unterschiedlicher-Befallsbedingungen auf die
mermithogene IntersexualiUit von Chironomus (Dipt.). Z. Parasitenk.
24, 484-545 (1964).
Goldschmidt, R.: Die sexuellen Zwischenstufen. Monographien aus dem
Gesarntgebiet der Physiologie der Pflanzen und Tiere ll, 1-528 (1931)
Gordon, R., Webster, J .M.: Mermis nigrescens: Physiological relation-
ships with its host, the adult desert locust Schistocerca gregm'ia.
EXp. Parasitol. 29, 66-79 (1971).
Gordon, R., Webster~J.M.: Nutritional requirements for protein syn-
thesis during parasitic development of the entomophilic nematode
Mermis nigrescens. Parasitology.§..i, 161-172 (1972).
Gordon, R., Webster, J.M., Hislop, T.G.: Merrnithid parasitism, protein
turnover and vitellogenesis in the desert locust, Schistocerca gre-
garia Forsk&l. Compo Biochem. Physiol. 46B, 575-593 (1973).
Kubler, H.: Das Aminosaurespektrum von Chironomus-Larven (Dipt.) und
seine Beeinflussung durch parasitare Mermithiden. Thesis, Univer-
sity Freiburg 1973.
Klirnrnel, G.: Zur Feinstruktur der Corpora allata von Chironomus. zool.
Anz., Suppl. 32, 123-135 (1969).
Madel, G.: Zur Biologie des Hurnrnelparasiten Sphaerularia bombi Dufour
1886 (Nematoda, Tylenchida). Bonner Zool. Beit~. 24, 134-151 (1973)
Maier, W.: Die Phenoloxydase von Chironomus thummi und ihre Beeinf lussung
durch parasitare Mermithiden. J. Ins. Physiol. ~, 85-95 (1973).
Muller, R.: Das Fettsaurespektrum von Larven von Chironomus thummi und
Ch. anthracinus (Dipt.) und seine Beeinflussung durch parasitare
Mermithiden. Thesis, University Freiburg 1972.
Palm, N.B.: Normal and pathological histology of the ovaries of Bombus
Latr. (Hymenopt.). Opuscul. entomol., Suppl. 7, 1-101 (1948).
Panov, A.A., Bassurmanova, O.K., Belyaeva, T.G.:-Ultrastructural changes
in the corpus allatum of the bug, Eurygaster, infected by the larvae
of Clytiomyia heUio. J. Ins. Physiol. 18, 1787-1792 (1972).
Pouvreau, A.: Contribution it l' etude de Sphaerularia bombi (Nematoda,
Tylenchidae), parasite de reines de bourdons. Ann. Abeille 5, 181
-199 (1962). -
134
Price, G.M.: Protein and nucleic acid metabolism in insect fat body.
Biol. Rev. i§., 333-375 (1973).
Rempel, J.G.: Intersexuality in Chironomidae induced by nematode para-
sitism. J. expo Zool. 84, 261-278 (1940).
Rempel, J.G., Naylor, J.M., Rothfels, K., Ottonen, B.: The sex chro-
mosome constitution of Chironomid intersexes parasitized by nema-
todes. Canad. J. Genet. Cytol. 4, 92-96 (1962).
Stoffolano, J.G., Jr.: The synchronization of the life cycle of dia-
pausing face flies, Musca autwrrnalis, and the nematode, Heterotylenchus
autwrrnalis. J. Inv. Path. 9, 395-397 (1967).
Strarnbi, A.: Influence du parasite Xenos vesparwn Rossi (Strepsiptere)
sur la neurosecretion des individus du sexe femelle de Polistes
gallicus L. (Hyrnenoptere, Vespide). C.r. Acad. Sci. Paris 260,
3768-3769 (1965).
Thienemann, A.: Lunzer Chironomiden. Arch. Hydrobiol., Suppl. l§,
1-202 (1950).
Ursprung, H., Nothiger, R., eds.: The biology of imaginal disks.
Berlin-Heidelberg-New York: Springer 1972.
Wtilker, W.: untersuchungen tiber die Intersexualitat der Chironomiden
(Dipt.) nach Paramermis-Infektion. Arch. Hydrobiol. Suppl. 25, 127-
181 (1961).
Wtilker, W.: Parasitologische und biochemische Verwandtschaft in der
GattungChironomus (Dipt.). Naturwissenschaften 50, 49-50 (1963).
Wtilker, W.: Parasite-induced changes of external and internal sex
characters in insects. Exp. Parasitol. 15, 561-597 (1964).
Wtilker, W.: Parasi tismus des Nematoden Gastromermis rosea in Chironomus
anthracinus (Diptera). Film C 1024 des Inst. f. d. Wissensch. Film
Gottingen 1970.
Wtilker, W.: untersuchungen tiber die Ultrastruktur der Gonaden von
Chironomus (Dipt.). 2. Ovarstruktur nach Schadigung durch parasi-
tare Mermithiden. Z. Parasitenk. 36, 73-94 (1971).
Wtilker, W., Gotz, P.: Die Verwendung~er Imaginalscheiben zur Bestim-
mung des Entwicklungszustandes von Chironomus- Larven (Dipt.). Z.
Morphol. Tiere 62, 363-388 (1968).
Wtilker, W., Ktirnrnel~G.: Investigations on the ultrastructure of cor-
pora allata in mermithized Chironomus (Insecta, Diptera). Proc. 3th
Intern. Congress Parasitol. Mtinchen 1974 (in press).
Wtilker, W., Winter, G.: untersuchungen tiber die Ultrastruktur der Go-
naden von Chirononrus (Dipt. ). 1. Normalentwicklung der Ovarien im
4. Larvenstadium. Z. Zellforsch. 106, 348-370 (1970).
Hermaphroditism In Insects. Studies on Icerya purchasi
M. Royer
A. Introduction
Hermaphroditism occurs in many different zoological groups. Insects,

however, which represent by far the largest subphylum in the animal
kingdom, are almost exclusively gonochoristic. Accidental cases of
hermaphroditism have been reported for CUlex (Sicart and Larrouy, 1962)
and Polyergus rufescens (Beck, 1958). These studies, however, are merely
morphological and it is probable that we are dealing here with exam-
ples of gynandromorphism rather than hermaphroditism. Reports on three
other genera - Termitoxenia, Perla, and Icerya - seem to be more inter-
esting although many questions remain as yet unanswered.
I . Termi toxenia
Wasmann (1900, 1901, 1902) indicated that in Termitoxenia, commensal

diptera of fungi-cultivating termites, the animals first pass through
a male phase during which spermatozoa are formed. Later the female
stage develops during which oocytes in two ovarioles start to grow.
This interpretation is supported by Assmuth (1913, 1923). According
to Brues (1908) and Bugnion (1913) Termitoxenia is a gonochoristic in-
sect and these authors believe that the assumption of hermaphroditism
is based on an erroneous interpretation of certain body structures.
II. Perla marginata
Schonemund (1912) claims to have discovered the existence of a her-

maphroditic gonad in male larvae of this species. Junker (1923) made
a histological study and described that the male larva has a testis
(formed by numerous vesicles) with ovarioles on its top. Vitellogene-
sis in the ovarian tubes is blocked as soon as the testis becomes ac-
tive. Therefore, hermaphroditism never becomes functional and the
adults are physiological males or females. The causes for this sexual
anomaly are unknown. It has only been established that the chromosomal
set of the female is 24 and that of the male 22.
III. Hermaphroditism in Icerya
In the genus Icerya, three hermaphroditic species have been described:

Icerya rurchasi (P ierantoni, 1911, 191 4a), Icerya bimaculata (Hughes-
Schrader, 1963), Icerya zeteki (Hughes-Schrader and Monahan, 1966).
In 1914 Pierantoni described the hermaphroditic structure of the juve-

nile gonad of individuals which according to him are of male origin,
alongside of which true females also exist. In 1925 Hughes-Schrader,
however, pointed out that the animals are normally protandric her-
maphrodites with self-fertilization and that pure males occur only
136
rarely. In further studies Hughes-Schrader described the gametogenesis

(1926-1927), some cytological peculiarities (1925) and the chromosomal
determination of sexuality in comparison with other Coccidae (1948).
Since these important studies do not yet allow a full understanding of

the determination of hermaphroditism in Icerya pUY'chasi, we have taken
up this problem for further elucidation.
B. Results
I. Sexual Phases in the Hermaphroditic Individuals of Icerya pUY'chasi
These individuals go through three larval instars before reaching the

imaginal stage.
1. Gonad at the First Larval Instar
The preliminary gonadal anlage shows little change of size during this
instar. The gonad is ovoid with a central constriction which later
disappears. Large cells which form the more dorsal part of the prelim-
inary anlage separate into islets at the periphery of the central mass,
which is formed by small cells. During this period the gonad is formed
by two distinct cellular lots.
2. Gonad at the Second Larval Instar
At first one observes an elongation of the gonad which maintains its

globular shape. All the central cells start to divide and cysts are
formed (Fig. 1). These cells are haploid (the diploid set is 2N=4) and
show a particularly long prophase with two plumose-like chromosomes.
This aspect is similar to the pachytene stage of a usual meiotic di-
vision. Chromosomes resemble "lampbrush chromosomes" at certain moments
of male meiosis in some insects, but since there is a haploid set only,
pairing does not occur. The gonad is in a male phase.
Chromophilic granules are frequently noticed at the center of a sper-

matogenetic lobe (Fig. 2) and can be observed during the whole course
of the testicular development.
We have never observed divisions in these islets of chromophilic gra-

nules. They apparently do not result directly from mitotic or meiotic
phenomena where chromosomal parts might be eliminated in this form.
The picture corresponds rather to a cellular degeneration. Lateral
islets of large cells remain together as groups without noticeable
transformations. This period corresponds to the neginning of spermato-
genesis and we observe simultaneously the presence of parietal and
quiescent oocytes. The gonad is therefore an ovotestis in a male phase
with female potentiality.
3. Gonad at the Third Larval Instar
Spermiogenesis generally ends as this period begins. Then the large

parietal cells advance and form an ovariole lateral to the testicular
part. The oocytes increase (Fig. 3) and the volume of the gonads be-
comes much larger due to the large number of follicles. This period
generally indicates a sexual inversion of individuals which proceed
from a dominating male condition to a female phase.
137
Fig. 1. Section of the gonadal

region during the testicular
stage (second larval ins~ar).
Two chromosomes in the cells
clearly visible ( eh chromo-
somes; et cyst; PM early
metaphase)
Fig. 2. Chromophilic granules (Se) in a testis (em male cells)

Fig. 3. The beginning of ovarian development (Gg genital duct;
Ov ovariole)
4. Gonad at the Fourth Instar
This last instar is characterized by the maturity of the ovarian ele-

ments. The oocytes which cease growing are fertilized by previously
formed spermatozoa (self-fertilization).
138
5. Gametogenesis during the Various Developmental Stages
It is concluded that the male germinal cells from the second instar
are haploid, as all of them contain only two chromosomes. From this
fact (Hughes-Schrader, 1948), it must be assumed that they do not go
through a true meiosis. There is simply a series of normal mitoses
leading to 32 spermatozoa from one spermatognonium.
The occurrence of chromophilic granules is explained by Hughes-Schrader

and Monahan (1966) as an indication for the elimination of an abortive
chromosomal mass (one genome) assuring a chromatic reduction during
the first larval instar. The arguments of these authors are, however,
not convincing, as the pictures may be interpreted perfectly in terms
of mitotic divisions.
We have demonstrated that a long mitosis in the spermatogenetic process

during the second instar shows some kind of a "meiotic anlage". This
observation cannot be reconciled with the interpretation of the pre-
viously authors since it would imply that a chromatic reduction pre-
cedes the usual heterotypic division and that two meiotic divisions
- the second one being incomplete - would follow each other. There is
no reason to assume such an abnormality. Instead of this we advance
the hypothesis that in Icerya spermatogenesis is characterized by the
absence of a true meiosis, a long mitotic prophase before the formation
of binuclear spermatids and by the degeneration of a certain number of
sex cells. In the spermatogenetic process only cells are involved which
are already haploid.
Observations on the quiescent cells along the testicular wall show

that these are diploid female cells. Therefore, the gonad of the second
larval instar has to be considered as an ovotestis with male functions
and female potentiality.
II. Gonad Formation
Following Pierantoni's (1914b) and Shinji's (1919) studies, we have

shown (Delavault et al., 1969) that the posterior blastomeres of the
periblastula become associated with an accumulation of symbiotic bac-
teria which colonize the posterior pole of the egg.
By mitotic divisions, these cells form an envelope around the micro-

organisms, this structure forming the polar mass. In its initial phase
a selection of larger cells can be observed. Following their further
development we note that these cells represent the origin of the sexual
cells. Therefore this phase is characterized by the polar concentration
of initial germinal cells.
It must be emphasized (Royer, 1971) that from now-on among the diploid
cells with four chromosomes, haploid cells with two chromosomes can
also be observed.
By invagination the polar mass is transported passively to the anterior

pole of the egg.
Later, initial germinal cells which until then were associated with
the polar mass, detach themselves from it and settle on the partitions
of a mesodermic duct (Royer, 1973).
In the course of the embryonic development the remaining mass of sym-

bionts is integrated in the anterior part of the proliferating abdomen
and comes close to the gonadal anlagen.
139
The association of gonads and symbionts remains intimate. On each side

gonad and micro-organisms form a dumb-bell shaped organ. At one end
the symbionts are grouped in mycetocytes, while at the other end the
germinal cells are linked to each other by mesodermic cells.
During this developmental stage one clearly observes two lots of dif-
ferent germinal cells (Fig. 4): some are relatively small but their
size is larger than that of somatic cells. They are grouped into a
sphere. The others are large cells, attached to the gonadal anlage in
the posterior part which consists of the smaller cells. A double origin
of gametogenic tissues can be seen in this phenomenon.
Fig. 4. Section of a gonadal anlage. The dualistic structure with big

cells (Geg) and small cells (Peg) can be seen. Symbionts (S) in the
vicinity
Fig. 5. Haploid nuclei among diploid cells. (Cd diploid cell; Ch ha-
ploid cell)
III. Early Segregation of Haploid and Diploid Cells
The preceding conclusion is'well supported by studies of the caryo-

types in young larval stages.
1. Haploid Cells in the Diploid Embryo
At the periblastula stage young embryos show large blastomeres. The

nuclei are large and chromosomes can easily be observed. Mitoses are
sufficiently numerous to allow the satisfactory observation of the end
140
of prophases or the beginning of metaphases. Among the diploid cells,

that is in the hermaphroditic embryos - male embryos are totally ha-
ploid - scattered haploid cells are always visible (Fig. 5).
Such cells are mainly located at the posterior pole, near the symbio-
tic mass. Some can also be seen in the posterior and ventral area which
will supply the blastoderm. They seem to be absent from the anterior
area of the egg.
In the following stages the haploid cells are still discernible, but
then they are smaller and mitoses are less frequent. In the embryo's
invagination phase they can be found as clusters associated with the
polar mass and they migrate together with it.
Except for the size, we can note the homology of the two chromosomes
in these haploid cells with those of spermatocytes. In the same way
these cells are comparable with haploid cells in the rare, purely male
embryos.
2. Haploidy, Diploidy and Sexuality
Since it is possible to dlstinguish haploid cells among diploid cells

from the very early stages of the embryo, one may wonder about the
role of these cells.
They are located in two particular zones. Some are differentiated in

close vicinity to the symbionts and we can see a similarity in the
arrangement of haploid cells and of a part of initial germ cells.
The others are associated with the formation of the ventral plate
which undergoes the constitutive invagination of the blastoderm and
of the amniotic membrane.
They undergo this invagination together with other cells of this layer
and join the cells of the polar mass. Thus, the two main aggregati0ns
of haploid cells integrate into the polar mass. Although we could not
investigate the development of the haploid cells in all details, we
assume that they are sexual cells, because they are located in the
embryo and show migrations similar to the other initial germ cells.
The morphological distinction of the two categories of cells of the
gonadal anlage does not seem to be possible before the onset of organo-
genesis. The sudden difference in size of these cells is remarkable
but we are unable to provide an explanation for this. These difficul-
ties prevent us from concluding with absolute certainty that the ha-
ploid and sexual cells are identical.
On the other hand, we know that in some Coccidae as in Pseudococcus or

Gossyparia, a part of the chromosomes is heterochromatic. During inter-
phase their appearance differs from that of ordinary chromosomes
(Schrader, 1923i Brown and Nur, 1964). Even the expulsion of complete
haploid set of chromosomes is known to occur in some species.
Therefore the question arises whether heterochromatinization might be

the cause of hermaphroditism.
a) In the more simple case, such as in Phanococcus citri, the chromosomes

of paternal origin remain in a permanent metaphasic condition. The
cells have a diploid constitution but are physiologically haploid.
It does not seem (Schrader and Hughes-Schrader, 1926, and Hughes-
Schrader, 1948) that in the Iceryinae in general and in Icerya purchasi
in particular, there are heterochromatic chromosomes.
141
b) In other species, "the paternal lot is completely eliminated from

embryos destined to become male" (Brown and Nur, 1964). In this way
all embryonic cells become haploid.
Royer and Delavault (1974) have indicated that the formation of males
in Icerya purchasi can be explained by a phenomenon which is comparable
to this chromosomal elimination. But in hermaphrodites, where haploid
and diploid cells are placed side by side, such a process does not seem
to be conceivable unless we assume that it would act selectively by an
expulsion of chromosomes which affect some cells only, though it ap-
pears to be quite difficult to support this hypothesis.
c) In some parthenogenetic species, as in Pulvinaria hydrangea, the first

two nuclei fuse with one another after cleavage. 5% of the embryos,
although derived from unfertilized eggs, show heterochromatinization
in some chromosomes. In the individuals showing this peculiarity all
the cells are affected.
Here again, the situation is entirely different from that encountered

in Icerya purchasi.
d) In some other species, the polar bodies sometimes fuse with a nu-
cleus of a blastomere. This occurs in Pseudococcus citri (Hughes-Schrader,
1948). The pentaploid cell thus formed represents the origin of the
first mycetocyte. We then note that the three chromosomal masses de-
rived from the polar bodies become heterochromatic, while the other
embryonic cells are euchromatic, and all of· ,them are diploid. There-
fore we can distinguish cells with different caryotypes within the
same embryo. One again, Icerya purchasi cannot be compared with these
cases.
The phenomenon of heterochromatinization cannot explain the juxtaposi-

tion of haploid and diploid cells in the embryo. The segregation must
start before the first cleavage divisions.
IV. Study of Fertilization
Following Hughes-Schrader (1927) we have observed the meiotic divisions

of the oocyte in order to find an eventual anomaly which could explain
the formation of the two groups of cells (diploids and haploids). These
studies revealed a normal meiosis which is entirely comparable to what
Pierantoni (1912) and Hughes-Schrader (1925) had described. Therefore
a more detailed study of the events at fertilization is indispensable.
1. Polyspermy
The spermatozoa penetrate the posterior pole of the egg, where symbionts
are inoculated into the oocyte. Polyspermy is the rule (Fig. 6). 4 - 10
spermatozoa enter the oocyte and disperse within the whole posterior
area of the egg. Some of them degenerate but others become pronuclei:
a constriction is seen in their central part, the junction between the
two parts disappears, the two elements of the spermatozoon being the
two chromosomes. This corresponds to observations by Hughes-Schrader
(1946). 3 - 6 male pronuclei which are formed in this way can be rec-
ognized in the oocyte.
2. Male Pronuclei and Amphimixis

Amphimixis is achieved by the spermatozoon which has become a male
pronucleus and which is nearest to the female nucleus. At the same
142
time meiosis is coming to its end and the female pronucleus is formed.
The two chromosomal sets constitute the first diploid nucleus. The
spermatozoa in which transformation is not finished degenerate, in
accordance with Hughes-Schrader's observations. On the other hand,
those which have reached the pronucleus stage persist (Fig. 7).
Fig. 6. Polyspermy in an oocyte. The spermatozoa are stretched; note

the two chromosomes (Chr)
Fig. 7. Pronucleus (Pr) in an egg. Two chromosomes (Chr ) can be ob-
served
At the end of this process, the normal diploid nucleus is formed and
a few haploid nuclei, located near the polar mass, still persist in
the area of the future "ventral plate".
There is a parallel between this fact and our previous observations.

The haploid cells which were then detected in the young periblastula
are found in the same place. Therefore it is likely that these cells
derive from divisions of supernumerary pronuclei.
Although we have not been able to follow their development more strict-
ly, it remains worthy of note that, with a difference of 4 or 5 steps
of successive divisions, very similar phenomena can be found.
3. Comparison with Other Special Cases
The existence of haploid embryonic cells which derive from supernumer-

ary spermatozoa after fertilization reminds us o~the androgenic devel-
opment of the male part in hermaphrodites.
But this is not a question of strict androgenesis, which is unknown

in the animal kingdom. When male nuclei develop in a female cytoplasm,
the zygotic nucleus divides simultaneously. The development of the
pronuclei seems to be some kind of a "surplus "-phenomenon , an essential
difference in relation to genuine androgenesis.
Our hypothesis requires several premises, one of which, polyspermy,

is in fact a relatively frequent phenomenon in insects.
Furthermore, it has to be assumed that the supernumerary nucleus must

be able to divide. In Bombyx it has been demonstrated that male pronuclei
143
can also show a sLngle dLvision, but in this case it seems that the
evolution of the zygotic nucleus does not inhibit the ability of ha-
ploid nuclei to divide - at least in a first stage. The action of the
amphimixis-nucleus on the other nuclei seems to be more or less power-
ful to the species. Therefore it is possible that this inhibition ei-
ther does not exist or is very weak in Ieerya.
Moreover, the egg contains symbionts, whose presence may play an im-
portant role in the reactivity of the yolk to polyspermy. It is con-
ceivable that they hamper the already weak inhibitory action of the
zygotic nucleus on the other pronuclei.
The interaction of symbionts with the host's sexuality is a well-known

phenomenon. Bacci (1965) analyzed this phenomenon after Buchner's works
(1954) on Stitoeoeeus sjoestedti, and Lampel's studies (1958) on Pemphigus
spirotheeae (Aphidae).
We have attempted to dLscover what the significance of symbiotic bac-
teria could be for the sexuality of Ieerya purehasi. For that purpose
we have tested the action of various antibiotics. Unfortunately, none
of them has had influence on these bacteria until now.
However, even if it has not been possible to prove the action of sym-
bionts upon the egg's mitotic dynamism, we must keep in mind that the
male pronuclei which seem to develop are in close vicinity to the sym-
biotic mass, and can thus be shielded from an inhibiting influence by
the diploid nucleus. Then they might remain able to divide.
It seems useful to recall that Schrader (1923) and later Walczuch (1932)
have shown that, in some cochineals such as Pseudoeoeeus citri, the polar
bodies are not expelled and that syrnbionts are integrated into these
cells. In this case, cells which do not derive from the zygotic nucleus
develop. Although they are not pronuclei the phenomenon is, neverthe-
less, similar. Cells which do not result from the amphimictic nucleus
can exist, so that our hypothesis that supernumerary haploid male cells
develop is not improbable.
C. Conclusion
Among insects only the Iceryidae, and the genus Ieerya in particular
seem to have a true functional hermaphroditism. Our studies on Ieerya
purehasi reveal that the hermaphroditic condition is established at
the moment of fertilization. The usual polyspermy is followed by a
transformation of the spermatozoa into pronuclei. One of these becomes
involved in the amphimictic process and the ensuing development of the
soma and the initial female germinal cells which are diploid. The other
pronuclei represent the source for the ameiotic male germinal cells.
From the moment of fertilization the male and fem~le gametogenic cells
become definitely segregated and fixed with regard to their further
morphogenesis. Such a strict determination of sex corresponds to what
is known of the general gonochorism in insects.
References
Assmuth, J.: Termitoxenia assmuthi Wasmann. Anatomisch histologische

Untersuchung. Nova acta acado Leop. Carol. 98, 191-316 (1913).
Assmuth, J.: Ametabolie und Hermaphroditismus bei den Terrnitoxeniiden.
BioI. ZentraI. 43, 268-281 (1923).
144
Bacci, G.: Sex determination. Pergamon Press 1965.

Beck, H.: tiber einen weiteren Fall von lateralem Hermaphroditismus bei
der Ameise Polyergus rufesaens Latr. Zool. Anz. ill, 243-249 (1958).
Brown, S.W., Nur, V.: Heterochromatic chromosomes in the coccids.
Science 145, 130-136 (1964).
Brues, C.T.::Some stages in the embryology of certain degenerate Phori-
dae and the supposedly hermaphroditic genus Termitoxenia. Science 11,
942 (1908).
Buchner, P.: Endosymbiosestudien an SchildUiusen Stiatoaoaaus sjoestedti.
Z. Morph. 5kol. Tiere 43, 262-312 (1954).
Bugnion, E.: Termitoxenia. Etude anato-histologique. Ann. Soc. entomol.
Belg. 57, 23-44 (1913).
Delavault, R., Royer, M., Duvigneau, C.: Chronologie de l'embryogenese
d' Iaerya purahasi. Ann. Embryol. Morphog. 3, 71-80 (1969).
Hughes-Schrader, S.: Cytology of hermaphrodItism in Iaerya purahasi
(Coccidae). Z. Zellforsch. 2, 264-292 (1925).
Hughes-Schrader, S.: Spermatogenesis in Iaerya purahasi a correction.
Science 63, 500-501 (1926).
Hughes-Schrader, S.: Origin and differentiation of the male and female
germ cells in the hermaphrodite of Iaerya purahasi (Coccidae). Z.
Zellforsch. ~, 509-540 (1927).
Hughes-Schrader, S.: A new type of spermiogenesis in iceryine coccids,
with linear alignment of chromosomes in the sperm. J. Morphol. 78,
43-71 (1946).
Hughes-Schrader, S.: Cytology of Coccids (Coccoidea, Homoptera). Adv.
Gen. 2, 127-203 (1948).
Hughes-Schrader, S.: Hermaphrodism in an African Coccid, with notes
on other Margarodids (Coccoidea - Homoptera). J. Morphol. 1...!l,
173-184 (1963).
Hughes-Schrader, S., Monahan, D. F.: Hermaphroditism in Iaerya zeteki
Cockerell and the mechanism of gonial reduction in Iceryine Coccids
(Coccoidea: Margarodidae Morrisson). Chromosoma 20, 15-31 (1966).
Junker, H.: Cytologische Untersuchungen an den Geschlechtsorganen der
Halbzwi tter-Steinfliege, Perla maginata. Arch. Zellforsch. 11., 185-
259 (1923).
Lampel, G.: Die symbiontischen Einrichtungen im Rahmen des Generations-
wechsels monozischer und heterozischer Pemphiginen der Schwarz-
und Pyramidenpappel. Z. Morphol. 5kol. Tiere 47, 403-435 (1958).
Pierantoni, U.: Larven-Hermaphrodi tismus von Iaerya purahasi. Z. wiss.
Insektenbiol. 7, 322-323 (1911).
Pierantoni, U.: Studio sullo svilluppo d' Iaerya purahasi Mask. Parte 1.
Origine ed evoluzione degli elementi sessuali femminili. Arch.
Zool. Ital. 5, 321-400 (1912).
Pierantoni, U.: -Studio sullo svilluppo d' Iaerya purahasi Mask. Parte 2.
Origine ed evoluzione degli organi sessuali maschili. Ermafrodismo.
Arch. Zool. Ital. 7, 27-49 (1914a).
Pierantoni, U.: Studio sullo svilluppo d' Iaerya purahasi Mask. Parte 3.
Osservazioni di embriologia. Arch. Zool. Ital. 7, 243-274 (1914b).
Royer, M.: Presence de cellules haploides parmi les blastomeres di-
ploides de l'embryon d' Iaerya purahasi. C.R. Acad. Sci. Paris 272,
3344-3347 (1971).
Royer, M.: Role de la polyspermie dans la formation des lignees cellu-
laires haploide et diploide chez l'insecte Iaerya purahasi. Ann. Embryol.
Morphog. 6, 243-252 (1973).
Royer, M., Delavault, R.: Formation des males chez Iaerya purahasi, in-
secte hermaphrodite. C.R. Acad. Sci. Paris 278, 2205-2208 (1974).
Schonemund, E.: Zur Biologie und Morphologie einiger Perla. Zool. Jahrb.
Abt. Anat. 34 (1912).
Schrader, F.: Astudy of the chromosomes in three species of Pseudoaoaaus.
Arch. Zellforsch. 12, 45-62 (1923).
145
Shinji, G.O.: Embryology of Coccids with especial reference to the

formation of the ovary, origin and differenciation ot the germ
cells, germ layers, rudiments of the midgut and intracellular sym-
biotic organisms. J. Morphol. 33, 73-167 (1919).
Sicart, M., Larrouy, G.: Hermaphrodi tisme partie I chez un Culex hOY'tensis.
Bull. soc. hist. nat. Toulouse 97, 410-412 (1962).
Walczuch, A.: Studien an Coccidensymbionten. Z. Morphol. okol. Tiere
25, 623-739 (1932).
Wasmann, E.: Tenntoxenia, ein neues flugelloses, physogastres Dipteren-
genus aus Termitennestern. 1. AuBere Morphologie und Biologie.
Z. wiss. Zool. 67, 599-617 (1900).
Wasmann, E.: TeY'mitoxenia, ein neues flugelloses, physogastres Dipteren-
genus aus Termitennestern. 2.
Z. wiss. Zool. 70, 289-298 (1901).
Wasmann, E.: Zur naheren Kenntnis der termitophilen Dipterengattung
TeY'lTlitoxenia. Verhdl. 5. intern. Zool. Congo Berlin, p. 852-872 (1902)
Sex Specific Cell Differentiation in Different Types of
Intersexes of Lymantria dispar L.
G. C. Mosbacher
From studies made by Goldschmidt (1934), the gipsy moth, Lymantria

dispar L., is known as a classical example for genetically induced in-
tersexuality. Its exceptional position among other intersex producing
insects is based firstly on the distinct sex dimorphism, which permits
investigation of the differentiation of various sex specific charac-
teristics, and secondly on the existence of different types of inter-
sexuality (Mosbacher, 1968, 1973). Besides the usual diversities in
the internal and external genital apparatus, both sexes of L. dispar
show marked differences particularily in the size, in the wing color
(dark brown in males, white in females) and in the structure of the
antennae (strongly pinnate in males and slightly pinnate in females
(Fig. 1 a,b) .
Fig. 1. (a) Male,

(b) female, (c)
ZZ intersex and
(d) WZ intersex
of Lymantria dispar
As Goldschmidt (1934) has shown, diploid intersexes of Lymantria arise

from crossing different geographical strains differing in the valences
of their sex factors. Depending on the type of crossing, the intersexes
appear either in the homogametic, male sex, the so-called ZZ intersexes
(Fig. 1c), or in the heterogametic, female sex, the WZ intersexes (Fig.
1d). The two types of intersexes differ remarkably in the appearance
of their sex dimorphic characteristics. In ZZ intersexes, the wings
exhibit an irregular mosaic of dark male-like and white female-like
areas and the antennae display long and short branches side by side.
The sexual wing mosaic is found in all degrees of intersexuality. With
the increasing degree of intersexuality the number and the extension
of female-like areas increase as well as the frequency of short bran-
147
ches in the antennae. In WZ intersexes, the wings are uniform in color.

With increasing degree of intersexuality the wing color changes from
light brown in female-like WZ intersexes to medium and finally to dark
brown in male-like WZ intersexes. All branches of the antenna are
uniformly elongated and the grade of elongation is correlated with the
degree of intersexuality. Accordingly, the coarse sexual mosaic, which
was demonstrated to be characteristic for the ZZ intersexes, does not
appear to any degree in WZ intersexuality.
There are two possible explanations for the intermediate differentia-

tion of the sex dimorphic characteristics in WZ intersexes. First, it
may be caused by the existence of a very fine cell mosaic, resembling
the cell mosaic found by Seiler (1972) on the antennae of Solenobia
intersexes. Secondly, the sexual differentiation of the single cells
themselves may be an intermediate one.
To investigate the sexual differentiation of single somatic cells in

ZZ and WZ intersexes, the wing scales and the sensilla of the antennae
are suitable, since they are sex-dimorphic and relatively complicated
structures, each derived from a single cell.
The wing scales of both sexes of L. dispar differ not only in color
but also in size and shape. Analysis of the sex dimorphism in scales
was restricted to an area of about 3 mm 2 in the center of the uppers ide
of the forewing between the veins M3 and CUI (see marked area on the
wing sketch in Fig. 2), because the scales of different wing regions
differ greatly.
In Lymantria, as well as in other species of Lepidoptera (Henke and

Pohley, 1952; Pohley, 1953; Pentz and Krause, 1969) three main types
of scales can be distinguished, designated as cover-, middle- and
lower-scales. They are arranged in irregular transverse rows, whereby
the sequence of the scale types is not definitely fixed. The most pro-
nounced sex-dimorphism is encountered in the cover-scales. Long, slen-
der cover-scales with distal margins showing two, more or less sharp
points, predominate in females (Fig. 2, F 2 - S ), whilst a smaller number
are either relatively short, broad and sharply toothed (F 6 _ 9 ) or very
long and lanceolate (F 1 ). In comparison, the male cover~scales of the
investigated area are usually much shorter and broader than the female
scales. Their distal margins are never sharply pointed, but crenate
(Fig. 2, Ml - 7 ). Furthermore, the scales differ in their fine struc-
ture, in particular, the density of the longitudinal ribbing is higher
in the male than in the female scales. Electron microscopic analysis
of the ultrastructure of the scales has been initiated.
A similar sex-dimorphism is also exhibited by the middle-scales, espe-

cially by the shape of the distal edge, which is dentate in females
(F 10 - 13 )' but crenate in males (MS-10). A less pronounced dimorphism
is observed in the lower-scales (F 14 - 16 , MI 2-14/.-
In the white wing areas of the ZZ intersexes only scales with typical
female differentiation in color, size, form and fine structure are
to be found (ZI-9). Only slight divergences in the frequency of differ-
ent subtypes (e.g. higher numbers of lanceolate scales; Fig. 2, ZI_2)
and in the shape occur, especially amongst short scales (compare Z7-9
and F 6 - 9 , Fig. 2). Such differences are probably due to clonal varia-
tion. In dark brown areas however, the scales show pure male diffe-
rentiation in every respect (ZlO-16). Areas with male scales are
sharply separated from areas with female scales. Even at the border
line between two sexually differing areas no intermediate types of
scales are observed.
148
Fig. 2
149
Fig. 2. Different types of scales from the center of the upperside of

the forewing (marked area in the wing sketch). F I - 9 female cover-scales,
F IO - 13 female middle-scales, F 14 - I6 female lower-scales; MI - 7 male
cover-scales, Me-II male middle-scales, M12 - 14 male lower-scales;
ZI-9 cover-scales from a female-like area of a ZZ intersex, ZIO-16
cover-scales from a male-like area of a ZZ intersex; WI- 7 cover-scales
of a WZ intersex. x 135
150
In WZ intersexes, the medium brown wing color is not caused by a mix-

ture of pure male and pure female scales, as might have been expected
rather, all scales are more or less uniformly medium brown in color.
The cover-scales are considerably longer than those of the male and
broader than those of the female (W l _ 7 ). The distal margins are neither
sharply dentate as in females, nor distinctly crenate as in males -
the margins are usually lobed. The density of ribbing is not as high
as in males, but higher than in females. This indicates that the scales
of the WZ intersexes exhibit a sexually intermediate differentiation.
Apart from the wing color, the antennae of L. dispar exhibit the most
striking sex-dimorphic characteristic. In the male antenna two rows
of long, slender and light brown colored branches are located on the
shaft. In females, the branches are considerably shorter, somewhat
thicker and darker colored than in males. The longest branches of the
female antenna reach only about a fifth of the length of the male bran-
ches (Fig. 3).
Branch
N' e i • i e i
40 ZZ-J
o
2 2 o 2 4 o 2 mm
40 WZ-J ,
o
2 2 o 2 4 o 2 mm
Fig. 3. Antennographs of normal sexes and intersexes of L . di spar . ZZ - I

ZZ intersex, WZ - Il female-like WZ intersex, WZ -I 2 intermediate WZ in-
tersex, WZ-I3 male-like WZ intersex. Vertical line: shaft of the an-
tenna; horizontal lines: length of the branches. Branches are numbered
from the base to the top of the antenna
151
In ZZ intersexes, irregular series of long, pale, male-like branches

and series of short, dark colored, female-like branches appear side
by side on the shaft. (In Fig. 3 the dark, female-like branches of the
ZZ intersex antenna are represented by thicker lines.) This shows that
a coarse sexual mosaic is also present in the antennae of ZZ inter-
sexes, as well as in wings.
The lower row in Fig. 3 exhibit the antennographs of 3 WZ intersexes,

which differ in the degree of their intersexuality. In female-like
intersex (WZ-I1)' all the branches are slightly but uniformly elongated.
With an increase in the degree of intersexuality (WZ-I2, WZ-I3) the
length of the branches increases continuously until they achieve the
length of the male type. Unlike ZZ intersexes, no coarse sexual mosaic
is visible in the antennae of the WZ intersexes.
An analysis of sensilla was made to determine whether a cellular mosaic

exists in the intersexual antenna. Some preliminary results have al-
ready been published by Scheffler (1972) and Mosbacher and Scheffler
(1974). In the male antenna, large numbers of hair-like sensilla are
located on the flanks of the long branches (approx. 26,000/antenna)
(Fig. 4a). The majority of these sensilla consist of long sensilla (s.)
trichodea, which are responsible for the perception of the female sex
attractants (Schneider, 1971). Among the long s. trichodea one can ob-
serve a smaller number of short s. trichodea (Fig. 4b, arrow), nearly
half as long as the long s. trichodea. The ultrastructure of the short
s. trichodea is, on the whole, very similar to that of the long s. tri-
chodea - their function is unknown. In addition, still shorter s. basi-
conica can be observed (Fig. 4b, rings). The fine structure of the
latter is completely different from that of the s. trichodea, but iden-
tical to the s. basiconica of Bombyx mori (Steinbrecht, 1973). Neverthe-
less, the s. basiconica are sometimes rather difficult to distinguish
from the short s. trichodea in the light microscope.
On the short, dark branches of the female antenna only short s. tri-
chodea and s. basiconica can be observed (Fig. 4c). The long s. tri-
chodea are completely absent. (The long bristles on the tips of the
branches are s. chaetica and are present in both sexes.)
The diagrams on the upper left and the lower right of Fig. 5 are ob-
tained from measuring the length of the sensory hairs of female and
male antennae. In the female, two peaks are found, one at 30 - 40 ~m,
representing the basiconic sensilla and the other at 70 ~m, represent-
ing the short s. trichodea. These peaks are also recognizable in the
male, where a third, more pronounced peak is observed at about 200 ~m.
This represents the long s. trichodea, which are clearly distinguish-
able from the short s. trichodea by their length. The large amount of
long s. trichodea causes the pronounced decrease of the columns repre-
senting the s. basiconica and the short s. trichodea, although the
absolute number of short sensilla is approximately equal in male and
female (see Scheffler, 1972, for more information on density and dis-
tribution of the sensilla).
The long male-like branches of the antennae of ZZ intersexes exhibit

the typical male pattern of sensilla, especially due to their fringe
of long s. trichodea (Fig. 4d). Correspondingly, the short female-like
branches show, on the whole, the typical female sensilla pattern, i.e.
only short s. trichodea and s. basiconica, but no long s. trichodea.
Even within individual branches, dark female areas with short sensilla
and light male areas with long sensilla may occur side by side (Fig.
4e). Obviously, a pure sexual mosaic is also present in the antennae
of the ZZ intersexes.
1~
Fig. 4 a-f. Sensilla of normal and intersexual antennae. (a) Part of

a male antenna, one row of branches removed. (b) Isolated branch of
a male antenna, displaying long s. trichodea, short s. trichodea (arrow)
and s. basiconica (rings). (c) Part of a female antenna. (d) Part of
the antenna of a ZZ intersex with long, male-like and short, female-
like branches. (e) Isolated branch of a ZZ intersex antenna with a
male-like and a female-like area. (f) Isolated branches of a male,
a male-like, an intermediate and a female-like WZ intersex (from top
to bottom). Scale = 0,2 mm
In Fig. 4f isolated branches of 3 WZ intersexes with different degrees

of intersexuality are compared with the branch of a male. With the in-
creasing degree of intersexuality the number as well as the length of
the sensilla increase. The diagrams of Fig. 5 obtained by measuring
the sensory hairs of WZ intersexes are a more precise demonstration
of the gradual increase of the sensilla. The WZ intersexes are arranged
in 4 classes according to the degree of their intersexuality (Mosbacher,
153
19(3). The sensilla pattern of the female-like WZ intersexes in class I

is rather similar to that of the female. Dnly a slight increase of the
average number of the sensilla is observed, due to the slight elonga-
tion of the branches in this group of intersexes. In the intersex
classes II - IV however, the number of sensilla increases continuously
almost reaching the male level, which is about 10 x higher than the
female level. Simultaneously, a great number of the sensilla increase
more and more in length. In classes III and IV it is obvious that most
of the hair-like sensilla are intermediate in length. They are longer
than the short s. trichodea of the females and shorter than the long
s. trichodea of the males.
WZ-l rL
.·261
cf
.·351
Fig. 5. Length of the sensory hairs of the sensilla trichodea and the
sensilla basiconica on the antennae of normal sexes and WZ intersexes
with different degrees of intersexuality. Classification of WZ inter-
sexes according to Mosbacher (1913); WZ-I I female-like intersexes,
WZ-III and WZ-IIII intermediate intersexes, WZ-I IV male-like inter-
sexes. n average number of hair-like sensilla on the 15th exterior
branch, calculated from 5 specimens of each group
Thus, it is evident that the sexual differentiation of single somatic

cells, namely the scale forming cells and the trichogenic cells of the
hair sensilla, in WZ intersexes are not always strictly male or female
differentiated - but intermediate. This result contradicts the hypoth-
esis of Seiler (1969, 19(2) (based on research on Solenobia and Droso-
phila intersexes) that in insects the cells always exhibit a pure sex
specific, male or female differentiation.
The analysis of further intersexual characteristics in ZZ and WZ inter-

sexes lead to basically the same result, which is formation of a sexual
mosaic in ZZ intersexes and intermediate differentiation in WZ inter-
sexes. Thus, in the accessory genital apparatus of one ZZ intersex,
154
derivatives of the male genital imaginal disc (the so-called Herold

organ) and derivatives of female genital discs may occur in more or
less normal differentiation next to one another. Accordingly, one or
more of the 4 female genital discs were observed, as well as the He-
rold organ in the larvae. (Mosbacher, 1968, 1973). In the gonads of the
ZZ intersexes, male and female germ cells may be observed side by side.
The two types of germ cells are generally normal in development and
show only slight degeneration, especially when they are situated in
different compartments of the intersexual gonad (Fig. 6a). In the prox-
imal gonoducts mosaic epithelium can be observed, consisting of por-
tions of typical cylindrical vas deferens epithelium and flat oviduct
epithelium (Fig. 6b), similar to the gonoduct mosaic found in Solenobia
inters exes (Seiler, 1969).
In contrast to this, WZ intersexes show no alternative sex specific

cell differentiation in any other respect. The development of the sex-
specific non-homologous organs, like the gonoduct system, is strongly
disturbed. Frequently, only fragments of the female or male gonoduct
system are observed. The gonads of individuals with low intersexual
degrees usually contain strongly degenerated eggs in the ovarial tu-
bules. With higher degrees of intersexuality, the compartments are
often entirely empty or they contain atypically differentiated and
strongly degenerated cell material (Fig. 6c).
Studying the development of the gonads in WZ intersexes, it was shown

that oogenesis and spermatogenesis occur simultaneously and not suc-
cessively in the intersexual gonads, as it might have been expected
according to the hypothesis of the turning point, established by Gold-
schmidt. The development of male and female gonia in the intersexual
compartments starts at the same time as in normal sexes and at first
proceeds at a similar rate. In later larval stages however, when higher
degrees of differentiation should be reached, the development is re-
tarded and gametogenesis is disturbed. A large proportion of male and
female germ cells degenerate and disappear (Mosbacher, 1969).
From the results, it can be established that the ZZ and the WZ inter-
sexes of Lymantria apparently represent two completely different types
of intersexuality, which differ fundamentally in their capacity to
develop sex-specific structures. The ZZ inters exes are capable of auton-
omous sex-specific cell differentiation and resemble the inters exes
of Solenobia (Seiler, 1969) and Drosophila (Lauge, 1969a,b) in morpholo-
gical characteristics, as well as in the developmental physiology. The
fact that in WZ intersexes the somatic and the generative cells are
unable to reach sex specific differentiation, may have two reasons.
Firstly, it is possible that no sexual determination of cells occurs
and that the intermediate differentiation proceeds under the simulta-
neous influence of male and female sex factors. Secondly, one may as-
sume that the cells become sexually determined, either male or female.
The sex specific differentiation, however, might-be disturbed by extra-
cellular factors. This second assumption is supported by the observa-
tion, that the germ cells in the intersexual gonads start but do not
complete sex specific differentiation.
Furthermore, the question arises whether the gonosomal constitution is

responsible for the differences in the intersexual habitus, previously
described. It may be assumed that the W chromosome causes the deficiency
of sex specific differentiation of cells and organs in WZ intersexes.
This assumption is contradicted by the following:
Goldschmidt (1934) has already obtained WZ intersexes, of the so-called

Gifu type, by crossing European females of Lymantria and males from a
Japanese population found in the environment of the Japanese town of
155
Fig. 6 a-c. Gonads of adult ZZ and WZ intersexes~ (a) Section through

part of the gonad of a ZZ intersex. To the right, ovarial compartment
containing groups of oocytes (1) and nurse cells (2), only slightly
deformed and degenerated; in between cells of the follicle epithelium
(3). Below that, the fqlded epithelium of an ovarial tubule (4). To
the left, testicular compartment filled with spermatogonia (5) and
bundles of spermatozoa (6); on the lower left, vas deferens epithelium
(7). (b) Proximal part of a gonoduct of a ZZ intersex, joining at the
"ampulle" (1) of a testicular compartment. Mosaic epithelium consist-
ing of cylindrical vas deferens epithelium (2) and flat oviduct epi-
thelium (3). In the lumen of the duct, bundles of ripe spermatozoa (4)
On the right, epithelium of an ovarial tubule (5). (c) Section through
a gonadal compartment of a male-like WZ intersex with degenerated 00-
cytes and nurse cells (1) and atypically developed spermatocytes (2).
Scale = 0,1 mm
156
Gifu. These WZ intersexes of the Gifu type exhibit a distinct sexual

mosaic in all sex-dimorphic organs such as wings, antennae, accessory
genital apparatus and gonads, very similar to the mosaic shown by the
ZZ intersexes (Mosbacher, 1973).
In addition I found a second special type of WZ intersexes, by crossing
Japanese males with females of the bOY'digalensis race from Portugal.
These bOY'digalensis intersexes show a mosaic in the genital anlagen and
its derivatives, that is, the gonoducts and the genital armature. Con-
trary to this, the wings and antennae are intermediate in structure,
resembling the standard WZ intersexes. It is difficult to determine,
whether a mosaic or an intermediate diffferentiation occurs in the
gonads, because the bOY'digalensis intersexes are very female-like and
only well-developed eggs were found in the gonads.
Finally, a third type resulting from crossing Japanese males and fe-
males of a certain German strain, named D2, exhibit a mosaic only in
the gonads, where normally differentiated male and female germ cells
were observed side by side. All the other intersexual characteristics
are intermediate in differentiation (Mosbacher, 1973).
ZZ Types of WZ Intersexes
Organs Intersexes
Gifu Bordigalensis D2 Standard
Wings ••••
•••• •••••
••• ~ ~ ~
Anten nae •••• •• •••
•••• ••• ~ ~ ~
Gen i ta 1 •••• •••• • •••
anlage n •••• • ••• • ••• ~ ~
•••• • •••
Gonads
•••• • ••• ? •• •••
••• ~
•• Mosa i c Intermediate
••
Fig. 7. Differentiation of sexual mosaics and of sexual intermediate
structures in different types of ZZ and WZ intersexes of LymantY'ia
dispaY' L.
These observations, summarized in Fig. 7, indicate firstly that the

capacity of the cells to develop sex specificly may differ in different
organs and secondly, that a continuous sequence leads from the zz in-
tersexes via the Gifu-, the bOY'digalensis- and the D2 intersexes to the
standard WZ intersexes. Following this sequence, the capacity for auton-
omous sex-specific differentiation of cells decreases gradually. This
appears to justify the consideration that the so-called standard WZ
intersexes are not the normal case of intersexuality, as stated by
Goldschmidt (1934, 1955), but that they are an extreme, although often
observed type of intersexual differentiation.
157
References
Goldschmidt, R.: Lymantria. Bibl. genetica 11, 1-186 (1934).

Goldschmidt, R.: Theoretical genetics. Berkely-Los Angeles: Univ.
California Press 1955.
Henke, K., Pohley, H.J.:Differentielle Zellteilungen una Polyploidie bei
der Schuppenbildung der Mehlmotte Ephestia kiihniella Z. Z. Naturforsch.
7b, 65-79 (1952).
Lauge, G.: Recherches experimentales sur la determination et la dif-
ferenciation des caracteres morphologiques et histologiques des
intersexues triploides Drosophila melanogaster Meig. I. Mise en evidence
de phases de differenciation au cours du developpement. Ann. Embryol.
Morph. 2, 245-270 (1969a).
Lauge, G.:-Etude des gonades des intersexues trip~oides de Drosophila
melanogaster. Description morphologique. Ontogenese des structures
histologiques. Ann. Soc. Ent. Fr. 5, 253-314 (1969b).
Mosbacher, G.C.: Morphologische und entwicklungsphysiologische Unter-
suchungen zum Problem der IntersexualiUit bei Lymantria dispar L.
(Lepidoptera), pp. 1-188. Habil.-Schrift Univ. Saarbrticken 1968.
Mosbacher, G.C.: Die postembryonale Entwicklung der intersexuellen
Gonaden von Lymantria dispar L. und ihre Bedeutung ftir die Interpre-
tation des Intersexualitatsphanomens, pp. 144-152. Verh. Dtsch.
Zool. Ges. K6ln 1969.
Mosbacher, G. C.: Die Intersexuali ta t bei Lymantria dispar L. (Lepidoptera).
Z. Morph. Tiere 76, 1-96 (1973).
Mosbacher, G.C., Scheffler, H.J.: Geschlechtsspezifische Differenzie-
rung somatischer Zellen bei Intersexen von Lymantria dispar L. Verh.
Dtsch. Zool. Ges. Bochum 1974 (in press).
Pentz, S., Krause, G.: Standortbedingte Modifikationen der normalen
Schuppe und der R-induzierten Schuppe PS 1 auf weiblichen und mann-
lichen Hinterfltigeln von Plodia (Lepidoptera). Z. Morph. Tiere 65,
336-373 (1969).
Pohley, H.J.: Untersuchungen tiber differenzielle Zellteilungen und
somatische Mutationen am Schuppenkleid der Mehlmotte Ephestia kiihnieUa
Z. Biol. Zbl. 72, 577-598 (1953).
Scheffler, H.J.: Untersuchungen tiber den Bau und die Entwicklung der
Antennen bei den normalen Geschlechtern und Intersexen von Lymantria
dispar L. (Lepidoptera), pp. 1-122. Diss. Univ. Saarbrticken 1972.
Schneider, D.: Specialized odor receptors of insects. In: Gustation
and Olfactation, pp. 45-60. London-New York: Academic Press 1971.
Seiler, J.: Intersexuality in Solenobia triquetreUa F .R. and Lymantria
dispar L. (Lepid.). Questions of determination. Monitore Zool. Ital.
(N.S.) 3, 185-212 (1969).
Seiler, J.: Vergleich des Ftihlers der reinen Geschlechter und der tri-
ploiden Intersexe von Solenobia triquetrella. Molec. Gen. Genetics 115,
146-194 (1972).
Steinbrecht, R.A.: Der Feinbau olfaktorischer Sensillen des Seiden-
spinners (Insecta, Lepidoptera). Z. Zellforsch~ 112, 533-565 (1973).
Structural and Endocrinological Aspects of Hermaphroditism
in Pulmonate Snails, with Particular Reference to Lymnaea
stagnalis (L.)
J. Joosse
A. Introduction
The subclass Pulmonata of the Gastropoda is subdivided into two orders:

the Basornrnatophora, living mainly in freshwater habitats; and the Sty-
lornrnatophora, having a terrestrial way of life.
All pulmonate snails and slugs are hermaphrodites. The male and the
female cells are produced within each of the acini of the ovotestis.
However, as has been shown in some recent studies on the gametogenesis
of various pulmonates, there is a considerable variation with regard
to the simultaneousness of the production of the male and female cells.
In the freshwater snail, Lyrrmaea stagnaUs, hardly any protandry can be
distinguished in the juveniles. During the reproductive season the pond
snail shows a simultaneous production of male and female sex cells
(Joosse, 1964). In the perennial terrestrial species, HeUx aBpersa,
male cells are first produced in spring at the start of the reproduc-
tive season. This production of sperm is followed by a period of simul-
taneous hermaphroditism (Guyard, 1971). Finally, in the slug, Agriolimax
reticulatuB, the spermatogenic and oogenic periods are successive with
an overlap period (Runham and Laryea, 1968). The term simultaneous
hermaphroditism cannot be applied in a general way, but needs to be
further detailed at the species level.
The increase in knowledge of the reproductive systems and activities

of pulmonate gastropods during the last 15 years coincides with a rapid
increase in the knowledge of the endocrine system of this group. These
two fields are inter-related in the study of the endocrine control of
hermaphroditism in snails. This paper discusses the present state of
our knowledge in this field, as far as the pulmonate gastropods are
concerned.
B. Functional Morphology of the Ovotestis
The gonad of the pulmonate snails is located in the apex of the shell.
There, it is surrounded by the numerous blind sacs of the digestive
gland. In the planorbid snails the gonad lies posterior to the diges-
tive gland.
The gonad or ovotestis consists of a number of sac-like acini. In

Lyrrmaea stagnaliB there are approximately 25. In this species the devel-
oping sex cells show a peculiar arrangement in the acini (Joosse and
Reitz, 1969). The male and female cells, as well as the Sertoli cells
and the follicle cells, most probably originate from the germinal epi-
thelial ring (Fig. 1). The female cells do not multiply after their
differentiation, but remain in contact with the acinar wall and migrate
to a special area of the acinus. Here they become sessile and surrounded
159
by a number of follicle cells, while the yolk is stored in the oocytes .

This area of the acinus is, therefore, called the vitellogenic area.
The male cells are always attached to Sertoli cells, which apparently
migrate from the epithelial ring to the vitellogenic area - since the
spermatids ripen at the rim of this area. Spermiation occurs here.
At spermiation the cytoplasmic remnants of the spermatids are phago-
cytized by the Sertoli cells, which persist after spermiation. They
form a layer of phagocytizing cells that cover the vitellogenic area.
It is yet not understood how, at ovulation, the oocytes pass through
the layer of phagocytizing cells and whether the remaining follicle
cells degenerate or develop in the same layer of cells.
efferens
aCinus lumen
early
oocytes
degenera ling
oocyte
Fig. 1. Scheme of an acinus of the ovotestis of Lynmaea stagnalis. (For

details, see text)
On the basis of this model it can be stated that the acini can be di-
vided into three areas: the vitellogenic area, the spermatogenic zone
around this area, and the remaining part of the acinar wall covered
by the germinal epithelium. Moreover, it is clearly demonstrated that
the male cells are separated from the female cells by the layer of
Sertoli cells. It must be stressed that this model of the organization
160
of the acini in Lyrrmaea stagnalis has to be considered as a highly spe-

cialized one. A comparable arrangement of the developing sex cells i~
found in planorbid snails (Barth and Jansen, 1962). However, in Stylom-
matophora no indications of a special organization hav e been suggested,
e.g. for Arion ater (Ltisis, 1961; Smith, 1966), for Agriolimax re t i eulatus
(Runharn and Laryea, 1968) or for Helix aspersa (Guyard, 1971).
The question arises, what factors determine this organization in the

acini of the ovotestis of Lyrrmaea stagnalis . From a plastic reconstruction
of a gonad, it appeared that the vitellogenic areas are only present
in those parts of the acini that are located at the periphery of the
ovotesticular complex (Fig. 2). These areas are opposite to the lobes
of the digestive gland (Joosse and Reitz, 1969). Between the gonad and
the digestive gland, blood lacunae and vesicular connective tissue
cells are found. These cells are the storage centers for glycogen
(Sminia, 1972). These data suggest that the nutritive requirements of
the oocytes during vitellogenesis maybe the primary factor for the
organization of the acini, while the position of the spermatogenic
zones is secondary to this.
Fig. 2. Schematic representation of the reproductive system of Lymnaea

stagnalis . a acinus lumen; ag albumen gland; be bursa copulatrix; e car-
refour; dg digestive gland lobes; mg muciparous gland; og oothecal
gland; p penis; pc pars contorta; p g prostate gland; so spermoviduct;
sp sperm duct; sz spermatogenic zone; va vitellogenic area; vd vas
deferens; vs vesiculae seminales
C. Storage and Transport of Sex Cells in the RepFOductive Tract
The structure of the reproductive tract of the pulmonates is extremely

complicated (Fig. 2). The basic scheme shows a hermaphrodite duct, or
spermoviduct, arising from the gonad and giving rise to a female and
a male duct. In the Basornmatophora, the female and the male duct are
separate and each have a genital aperture. In the Stylornmatophora the
lumina of the ducts remain connected for some distance from their ori-
gin and after having been separated, they open into a cornmon genital
atrium.
The transport of the gametes through the reproductive tract is illus-

trated here, on the basis of the data for the pond snail (Joosse et al.,
161
1968; Plesch et al., 1971). A scheme of the reproductive tract of this

species is shown in Fig. 2. After spermiation, the sperms are trans-
ported to the spermoviduct where it is stored in great quantities in
the numerous vesiculae seminales. At copulation one snail acts as a
male, the other as a passive female. During a copulation sperm is
transported into the sperm duct. The prostate gland adds its secretory
products to the sperm mass that leaves the "male" via the vas deferens
and the penis. In the "female" copulatory partner, most of the sperm
is transferred to the bursa copulatrix. A minor quantity of sperm is
transported through the female duct (which has a ciliated groove) to
the region of the carrefour (Horstmann, 1955). The sperm cells are
probably passive during their transport. Sperm taken from different
parts of the reproductive tracts of animals during copulation or ovi-
position, never showed active movements.
In the laboratory and at a temperature of 20°C, the pond snail may

produce an egg mass every second or third day, each egg mass contain-
ing up to 200 eggs. The egg masses are formed in the following way.
After ovulation, the oocytes are assembled at the end of the spermovi-
duct. Ovulation can be induced with special external stimuli (see be-
low). It has been shown that the oocytes appear at the end of the
spermoviduct about ten minutes after the start of the ovulation stimuli
(Joosse and'Dijkstra, unpubl.). Therefore, ovulation must be a rapid
process. At the end of the spermoviduct the oocytes are fertilized and
one by one, the fertilized oocytes are transported into the haustra of
the pars contorta of the oviduct. In the haustra the oocytes are sur-
rounded by the perivitelline fluid - a secretion produced by the al-
bumen gland. Two egg membranes are secreted by the haustra around the
perivitelline fluid, while the secretion produced by the muciparous
gland fuses the eggs together. The oothecal gland then secretes a cap-
sule around the entire mass of eggs. The egg mass leaves the body via
the vagina and the female genital aperture, being pressed against the
substrate at oviposition.
D. Resorption of Sex Cells
An extensive system for the resorption of gametes is found in the re-

productive organs of the pond snail. Sperm cells are frequently phago-
cytized by the cells in the wall of the vesiculae seminales (Joosse
et al., 1968). Moreover, the sperm transferred to the bursa copulatrix
during a copulation, is digested in the lumen of this organ. The pro-
ducts of the digestion are then resorbed by the bursa cells (Horstmann,
1955) .
The oocytes are frequently resorbed by their foll~cle cells. In the

ovotestes of specimens of Lymnaea stagnalis collected in the field, the
frequency of the degenerating oocytes appeared to vary from 2 - 10%
whereas, in laboratory bred specimens, their frequency appeared to be
as high as 20 - 60% of the total number of female cells (Joosse, 1964;
Joosse et al., 1968).
A short time after oviposition a varying number of oocytes, and/or

eggs, are found in the reproductive tract. These are all transported
to the bursa copulatrix and digested in this organ, in the same way
as described for the sperm (Joosse and Backhuys, unpubl.).
Resorption of sperm has also been reported for the terrestrial snails
Helix pomatia (Breucker, 1964) and Helix aspersa (Guyard, 1971). In these
162
species the sperm is resorbed in spring, just before the production of

new sperm cells.
On the basis of the data reported above and on data of a starvation
experiment with pond snails, it has been suggested that the resorption
of gametes may play an important role in the metabolism of pulmonate
snails (Joosse et al., 1968).
E. The Endocrine System of the Pulmonata
The structure of the endocrine system of the Pulmonata will be briefly

di'scussed in this paper. It may serve as a basis for the survey of the
endocrine control of reproduction of the pulmonates. For a more de-
tailed review of the literature the papers by Durchon (1967), Joosse
(1972), Martoja (1972) and Boer and Joosse (in press), should be con-
sulted.
I. Neurosecretory Cells
Various types of neurosecretory cells have been described in pulmonate

gastropods. Up to ten types could be distinguished by Wendelaar Bonga
(1971) in the central nervous system of Lymnaea stagnaUs. The cell types
are characterized by their constant position in the various ganglia;
the variations in color of the secretory products after staining with
alcian blue/alcian yellow; the size and structure of the elementary
granules and the location of the neurohemal areas. It is character-
istic of the gastropods that no defined neurohemal organs are found.
The storage and release of the neurohormones occurs in neurohemal areas
at the periphery of nerves, connectives and commissures and in the
connective tissue around the central ganglia. The neurohemal areas
have a great extension in Lymnaea stagnaUs (Fig. 3). Thus, the entire
periphery of the large median lip nerves and the cerebral commissure
is occupied by the neurohemal ends of neurosecretory cells, located
in the cerebral ganglia. In the Stylommatophora, neurohemal areas have
been described in the connective tissue near the dorsal body (Nolte,
1965; Kuhlmann, 1966) and at the periphery of the cerebral arteries
(Van Mol, 1960).
II. Dorsal Bodies
The dorsal bodies are endocrine organs attached to the perineurium of

the cerebral ganglia (Fig. 3). Lever (1957, 1958) described them for
several species of basommatophoran pulmonates and suggested their endo-
crine function. In all species of Basommatophora~ paired medio-dorsal
bodies are found at the dorsal side of the origins of the cerebral
commissure. In some species (e.g. Lymnaea stagnaUs), a second pair of
bodies is present at the latero-dorsal side of the cerebral ganglia.
The medio- and latero-dorsal bodies. contain one type of secretory
cell. These dorsal body cells release their products into the inter-
cellular spaces of the bodies. The dorsal body cells show great sea-
sonal variations; preceding the reproductive period, the cells are
hyperplasmic (March-April) (Joosse, 1964).
Dorsal bodies are also found in theStylommatophora, but in several

species the cells appeared to be dispersed in groups in the connective
tissue sheath around the cerebral ganglia (Nolte, 1965; Cook, 1966;
Kuhlmann, 1966; Van Mol, 1967; Boer and Joosse, in press).
163
Fig. 3. Scheme of the cerebral

ganglia of Lymnaea stagnalis
with the position of the large
groups of neurosecretory cells
and the dorsal bodies. MDB,
LDB medio- and latero-dorsal
body; CDC caudodorsal neuro-
secretory cells; LGC light
green neurosecretory cells
III. Gonad
There is good exper~mental evidence that pulmonate gonads produce one

or probably two sexual hormones. However, the cells involved in the
synthesis of these hormones are as yet not identified. In Lymnaea stag-
nalis and Biorrrphalaria glahrata, the Sertoli cells, as well as the follicle
cells, have been studied histochemically and ultrastructurally with
respect to a supposed production of steroids. However, up to now the
results have been negative (de Jong-Brink, pers. comm.). Using bio-
chemical techniques, Gottfried and Dorfman (1970a,b) have shown that
steroids, related to the vertebrate gonadal hormones, are synthesized
by the ovotestis of Ariolimax californicus. However, whether these steroids
have an endocrine role in pulmonates has not been proved.
IV. Optic Tentacles
Many authors report extremely contradictory effects of extirpation of

the optic tentacles on the reproductive activity of Stylommatophora.
Several authors attributed the changes in reproductive activity to
humoral factors produced by the optic tentacles. Therefore, many stu-
dies have been devoted to the identification of the production centers
of hormones in the optic tentacles (e.g. Bierbauer and Vigh-Teichmann,
1970). From this literature it must be concluded that at the moment,
no cell type in the optic tentacles can be considered as a source of
hormones (Boer and Joosse, in press) .
164
The eyes of the Basornmatophora are located at the base of the tentacles.
The structure of the eyes and the tentacles is less complicated in com-
parison with those of terrestrial pulmonates. There are no indications
of an endocrine role of these structures in the Basornmatophora.
F. Endocrine Control of Reproduction
The interest in the endocrinology of the pulmonate gastropods has been

rapidly increasing during the last 15 years. Some factors which have
influenced this development are the detailed description of the endo-
crine system of various pulmonates (Lever, 1957; Wendelaar Bonga, 1971);
the introduction of specialized experimental techniques, like those
for the narcotization and operation of snails and slugs (Joosse and
Lever, 1959; Bailey, 1969); a.nd the introduction of the organ culture
technique for tissues of pulmonate gastropods (Gomot and Guyard, 1964;
Bailey, 1973).
I. Differentiation of Sex Cells
Gomot and Guyard (1964) and Guyard (1970, 1971) have studied the dif-
ferentiation of the sex cells in juvenile gonads of Helix aspersa in
organ culture conditions. In the anhormonal medium the female cells
show autodifferentiation. The male cells differentiate only in the
presence of a cerebral ganglion in the cultures. Apparently an andro-
genic factor is produced by these ganglia, but the androgenic cells
have not been identified.
The control of the differentiation of sex cells has not been studied
in the Basornmatophora. However, the same results will most probably
be obtained, since an autodifferentiation of the female cells and the
existence of an androgenic factor, have also been demonstrated in
other groups of gastropods (Streiff, 1967; Choquet, 1971).
Pelluet and Lane (1961) and Pelluet (1964) have studied the differen-
tiation of sex cells in vivo. They extirpated the optic tentacles of
Arion ater and Arion subfusaus and the homogenates of the optic tentacles
and cerebral ganglia were then injected. The ovotestes of the experi-
mental animals were histologically studied. From the results, Pelluet
and Lane concluded that the optic tentacles produce a hormone that
inhibits oogenesis; the cerebral ganglia produce a hormone that stimu-
lates oogenesis; .while a hormone from the optic tentacles would have
a stimulating effect on spermatogenesis. These results do not agree
with those obtained with the organ culture technique.
The experimental procedures of Pelluet and Lane nave been applied by

a number of investigators to various species of terrestrial snails and
slugs (Joosse, 1972). The results are highly contradictory. This may
be the result of different experimental conditions, or species specific
reactions to extirpation of the optic tentacles. Recently, Wattez
(1973) repeated the original experiments of Pelluet and Lane, on one
of the species studied by them (Arion subfusaus). Wattez observed a clear
feminization of the gonad in the slugs after removal of the optic ten-
tacles at birth, while spermatogenesis appeared to be blocked in the
operated animals. This indicated that the optic tentacles exerted a
stimulating influence on spermatogenesis and an inhibitory influence
on the differentiation of the female cells. These conclusions, as far
as the optic tentacles are concerned, are in line with those obtained
by Pelluet and Lane. However, the influence of the optic tentacles was
165
identical to those ascribed to the androgenic factor. Wattez is, as

yet, not inclined to suggest an endocrine role for the optic tentacles.
In view of the above information it is suggested that the control of

the differentiation of the sex cells in pulmonates is effected by a
single androgenic factor produced by the cerebral ganglia. This factor
then stimulates the differentiation of the male cells and suppresses
the autodifferentiation of the female cells. The release of the andro-
genic factor may then be controlled by the optic tentacles, either by
nervous or hormonal means.
II. Spermatogenesis
During the winter season, spermatogenesis is blocked in Lymnaea stagnalis

(Joosse, 1964) as well as in Helix aspers a (Guyard, 1971). This block is
caused by the low environmental temperatures: adult pond snails with
active spermatogenesis bred at 20°C, showed a gradual disappearance
of the spermatogenic stages following meiosis, after a transfer to a
temperature below 10°C. After about 80 days, only spermatogonia and
primary spermatocytes persisted in the ovotestis (Joosse and Veld,
1972). This indicates that spermatogenesis is temperature-dependent,
but only at the stage of meiosis. It can not be conluded from these
data whether the influence of the temperature has a direct or an in-
direct effect on the male cells.
In Helix aspersa, the mitotic activity of the various stages of spermato-

genesis is stimulated by the androgenic factor of the cerebral ganglia
(Guyard, 1971). However, all stages following upon the spermatogonial
stage degenerate in an anhormonal culture medium. These results indi-
cate that the effect of low temperature on spermatogenesis is mediated
by the androgenic factor.
III. Oogenesis-vitellogenesis
In the anhormonal organ cultures of juvenile gonads of Helix aspersa,

the female cells showed autodifferentiation. However, the oocytes did
not start the storage activity known as vitellogenesis (Guyard, 1971).
After extirpation of the dorsal bodies of juvenile specimens of Lymnaea

stagnalis, it appeared that vitellogenesis had ceased. Subsequent im-
plantation of dorsal bodies into the operated snails, led to a rapid
production of ripe oocytes and the start of the ovipository activity
(Joosse and Geraerts, 1969). Therefore, it seems that the dorsal body
hormone stimulates vitellogenesis in the oocytes. The cessation of
oviposition after extirpation of the dorsal bodies of adult pond snails
must be interpreted as due to an indirect effect of the absence of
vitellogenic activity (Joosse, 1964).
IV. Growth and Differentiation of the Reproductive Tract
Abeloos (1943) castrated various species of slugs and observed a cessa-

tion in the growth of the reproductive tract and even a regression in
size. Laviolette (1954) further extended these results. He transplanted
juvenile gonads or reproductive tracts in the hemocoel of normal or
castrated acceptor slugs. It appeared that the growth and differentia-
tion of the hermaphrodite duct and the albumen gland was controlled by
a humoral factor produced by the gonad. Recently, Runham et al. (1973)
published the results of experiments on the slug Agriolimax retieulatus.
Dur.tng the life cycle of this slug the gonad and the reproductive tract
166
have first a male phase, then a female phase and a period of overlap
in between. Transplantation of undifferentiated tracts into animals in
the early protrandric phase resulted in the differentiation and growth
of the prostate glands, whereas the female parts of the Anlage remained
undifferentiated and small. Transplantation of reproductive tracts of
juvenile snails into the hemocoel of egg-laying animals was followed
by the development of the female parts only. During the period of over-
lap the male and the female parts of the tracts developed. The authors
concluded that two hormones are involved in the growth and differen-
tiation of the reproductive tract, one controlling the male and a sec-
ond, controlling the female part of this tract.
In view of the experiments of Laviolette (1954) and Runham et al. (1973)

it is most probable that the male and the female hormone are produced
by the gonad. However, Bailey (1973) has shown in preliminary experi-
ments that in organ culture conditions the gonad fa~ls to stimulate
the development of an Anlage of the reproductive tract of Agriolimax
reticulatus, although the organs are in close proximity. The development
of the Anlage starts when the gonad, the juvenile tract and a brain
are present in one culture medium. This led Runham et al. (1973) to
suggest that the gonad will stimulate the brain to produce a male and
a female factor. However, from a comparative endocrinological point of
view, there is more evidence for the reverse pathway: the brain will
stimulate the gonad to produce the sexual hormones, and these in turn
stimulate the development of the tract.
The effects of castration on the reproductive tract of basommatophoran

snails are variable, due most probably to surgical difficulties. Harry
(1965) succeeded in the castration of juvenile specimens of Taphius
(Bionrphalaria) glabratus. After complete castration the whole reproductive
tract failed to develop. In contrast, Brisson (1971) observed no changes
in the reproductive tract of various planorbid snails after castration.
Geraerts and Algera (1972) have studied the growth and differentiation
of juvenile reproductive tracts of Lymnaea stagnalis after implantation
in hosts with or without dorsal bodies (Fig. 3). They report that the
growth and differentiation of the female part of the reproductive tract
and of the female accessory sex glands (Fig. 1) is dependent on the
presence of the dorsal body hormone in the host. Referring to the pres-
ent interpretation of the results of Laviolette (1954), Runham et al.
(1973) and Bailey (1973) one is inclined to suggest that the dorsal
body hormone does not have a direct effect on the female organs, but
an indirect one via the gonad: the dorsal body hormone stimulates the
gonad to produce a female sexual hormone that in turn stimulates the
growth and differentiation of the female part of the reproductive tract.
The problem remains to determine how the seasonal changes in the activ-
ity of the dorsal bodies is controlled. From preliminary experiments
wi th Lymnaea stagnalis, Geraerts (1973) has obtained some evidence that
shows that a humoral factor produced by the lateral lobes of the cere-
bral ganglia (Lever and Joosse, 1961; Brink and Boer, 1967) may be
involved in the control of the dorsal bodies.
V. Ovulation
As has been mentioned, the ovulation and oviposition of Lymnaea stagnalis

can be stimulated by external factors: e.g. a sudden raise in tempera-
ture, food supply, or the addition of oxygen to the water (Woerdeman
and Raven, 1946). These stimuli have been used in experiments dealing
with the synchronization of oviposition in Lymnaea stagnalis (Joosse and
Dijkstra, unpubl.). Every third day the water was changed, and oxygen
167
was supplied. In this way, up to 70% of the animals' could be stimulated

to oviposition. Ovulation occured within minutes, as the oocytes had
reached the end of the spermoviduct within 10 minutes after the stimu-
lation with oxygen. This indicates that a rapid mechanism for ovulation
exists in this species, which could be of a nervous or hormonal charac-
ter.
Geraerts and Bohlken (pers. corom.) have recently shown, that the caudo-
dorsal neurosecretory cells (Fig. 3) in the cerebral ganglia of Lymnaea
stagna Us produce an ovulatory hormone. Cauterization of these cells
results in a cessation of the ovipository activity. Injection of the
homogenates of cerebral coromissures (which are the storage and release
centers of this neurohormone) induces the process of ovulation and ovi-
position. The presence of an ovulatory hormone has also been demonstrat-
ed in the opisthobranch gastropod, Aplysia eaUforniea, by Kupfermann (1972).
VI. Conclusion
On the basis of the results mentioned above, a tentative hypothesis is

suggested for the endocrine control of hermaphroditism in pulmonates.
The male and the female activities are controlled by different hormones.
The androgenic factor is produced by the cerebral ganglia. This factor
is most probably a neurohormone. The androgenic factor stimulates the
differentiation of the male cells and the production of a male sexual
hormone by the gonad. This sexual hormone stimulates the growth and
differentiation of the male part of the reproductive tract. In Stylom-
matophora, the production of the androgenic factor is dependent on the
optic tentacles. The female activity is controlled by a neurohormone
of the brain. This neurohormone stimulates the development of the dor-
sal bodies and the production and release of the dorsal body hormone.
The dorsal body hormone stimulates vitellogenesis in the oocytes and
the production of a female sexual hormone by the ovotestis. The female
sexual hormone controls the growth and differentiation of the female
parts of the reproductive tract. Ovulation and oviposition are und~r
the control of a neurohormone. The release of the ovulatory hormone
can be influenced by environmental factors. The control of the activity
of both sexes in one hermaphrodite organism will not be as separate as
suggested by this hypothesis. As mentioned, the androgenic factor has
an inhibitory influence on the autodifferentiation of the oocytes.
Moreover, it is beyond doubt that intimate relations will appear to
exist between the systems for the control of reproduction and those
for the control of growth and metabolism.
References
Abeloos, M.: Effets de la castration chez un mollusque, Lim= m=imus L.

C.R. Acad. Sci. Paris 216, 90-91 (1943).
Bailey, T.G.: A new anaesthetic technique for slugs. Experientia 12,
1225 (1969).
Bailey, T.G.: The in vitro culture of reproductive organs of the slug
AgrioUm= retieulatus (Mull.). Neth. J. Zool. 23, 72-85 (1973).
Barth, R., Jansen, G.: tiber die Formveranderungen des Golgi-Apparates
wahrend der Spermiogenese von Australorbis glabratus olivaeeus (Mollusca,
Pulmonata, Planorbidae). Mem. Inst. Oswaldo Cruz 59, 83-114 (1961).
Bierbauer, J., Vigh-Teichmann, I.: Histophysiologicar-examination of
the optic tentacle of pulmonates. II. Cytochemistry of the special
and secretory cells. Acta Sci. hung. 11, 11-24 (1970).
168
Boer, H.H., Joosse, J.: Endocrinology. In: The Pulmonates (ed. V. Fret-
ter). Oxford: Academic Press (in press) .
Breucker, H.: Cytologische Untersuchungen des Zwitterganges und des
Spermoviductes von HelixpomatiaL. Protoplasma 58,1-41 (1964).
Brink, M., Boer, H.H.: An electron microscopical investigation of the
follicle gland (cerebral gland) and of some neurosecretory cells
in the lateral lobe of the cerebral ganglion of the pulmonate gastro-
pod Lymnaea stagnalis L. Z. Zellforsch. 79, 230-243 (1967).
Brisson, P.: Castration chirurgicale et regeneration gonadique chez
quelques planorbides (Gasteropodes Pulmones). Annales d'Ernbryol.
et de Morphogen~se 4, 189-210 (1971).
Choquet, M.: Etude du cycle biologique et de l'inversion de sexe chez
PateUa vu'igata L. (Mollusque Gasteropode Prosobranche). Gen. Comp.
Endocr. li, 59-73 (1971).
Cook, H.: Morphology and histology of the central nervous system of
Succinea putris (L.). Arch. Neerl. de Zool. 11, 1-72 (1966).
Durchon, M.: L'endocrinologie des Vers et des Mollusques. Paris: Masson
et Cie. 1967.
Geraerts, W.P.M.: Effects on growth of endocrine centres in the cere-
bral ganglia of Lymnaea stagnalis. In: Seventh conference of european
comparative endocrinologists (ed. J. Szentagothai and F. Hajas).
Budapest: Akademiai Kiada 1973.
Geraerts, W.P.M., Algera, L.H.: On the influence of the dorsal bodies
and the adjacent neurosecretory cells on the differentiation of
the reproductive tract in Lymnaea stagnalis. Gen. Comp. Endocr. .u!.,
592 (1972).
Gomot, L., Guyard, A.: Evolution en culture in vitro de la glande herm-
aphrodite de jeunes de l'esp~ce Helix aspersa (MulL). C.R. hebd.
Seanc. Acad. Sci. Paris 258, 2902-2905 (1964).
Gottfried, H., Dorfman, R.I~Steroids of Invertebrates. IV. On the
optic tentacle-gonadal axis in the control of the male-phase ovo-
testis in the slug (Ariolim= californicus ). Gen. Comp. Endocr. 12,
101-119 (1970a).
Gottfried, H., Dorfman, R.I.: Steroids of Invertebrates. V. The in
vitro biosynthesis of steroids by the male-phase ovotestis of the
slug (Ariolim= californicus). Gen. Compo Endocr. 15,120-138 (1970b)
Guyard, A.: La differenciation gonocytaire des mollusques gasteropodes
en culture in vitro. Ann. Biol. 9, 401-408 (1970).
controlant l'orientation sexuelle des gonocytes de l'escargot He'iix
aspersa (Muller.) Thesis, Besan90n (1971).
Harry, H.W.: Evidence of a gonadal hormone controlling the development
of the accessory reproductive organs in Taphius (Biomphalaria) glahratus
(Say) (Gastropoda, Basornrnatophora). Trans. Amer. micro Soc. ]i, 157
(1965) .
Horstmann, H.J.: Untersuchungen zur Physiologie der Begattung und Be-
fruchtung der Schlarnrnschnecke Lymnaea stagnalis L. Z. Morph. tikol.
Tiere 44, 222-268 (1955).
Joosse, J.: Dorsal bodies and dorsal neurosecretory cells of the cere-
bral ganglia of Lymnaea stagnalis L. Arch. Neerl. de Zool. 12, 1-103
(1964).
Joosse, J.: Endocrinology of reproduction in Molluscs. Gen. compo
Endocrinol., Suppl. 3, 591-601 (1972).
Joosse, J., Boer, M.H.,-Cornelisse, C.J.: Gametogenesis and oviposition
in Lymnaea stagnalis as influenced by y-irradiation and hunger. Symp.
zool. Soc. Lond. 11., 213-235 (1968).
Joosse, J., Geraerts, W.J.: On the influence of the dorsal bodies and
the adjacent neurosecretory cells on the reproduction and metabolism
of Lymnaea stagnalis. Gen. compo Endocrinol. 11, 540 (1969).
Joosse, J., Lever, J.: Techniques of narcotization and operation for
experiments with Lymnaea stagnalis (Gastropoda Pulmonata). Proc. Kon.
Ned. Akad. V. Wetensch., Amsterdam, C~, 145-149 (1959).
169
Joosse, J., Reitz, D.: Functional anatomical aspects of the ovotestis

of Lymnaea stagnaUs. Malacologia 9, 101-109 (1969).
Joosse, J., Veld, C.J.: Endocrinology of reproduction in the hermaphro-
dite gastropod Lymnaea stagnaUs. Gen. Compo Endocrin. 18, 599-600
(1972). -
Kuhlmann, D.: Die Dorsalk6rper der Stylommatophoren. Z. wiss. Zool.
173, 218-231 (1966).
Kupfermann, I.: Studies on the neurosecretory control of egg laying
in Aplysia. Amer. Zool. 12, 513-520 (1972).
Laviolette, P.: Role de la~onade dans le determinisme humoral de la
maturite glandulaire du tractus genital chez quelques gasteropodes
Arionidae et Limacidae. Bull. Biol. Fr. et Belg. 88, 310-332 (1954).
Lever, J.: Some remarks on neurosecretory phenomena in Ferrissia sp.
(Gastropoda, Pulmonata). Proc. Kon. Ned. Akad. v. Wetensch., Amster-
dam, C 60, 510-522 (1957).
Lever, J.:lOn the relation between the medio-dorsal bodies and the
cerebral ganglia in some pulmonates. Arch. Neerl. Zool. 11, 194-201
(1958) •
Lever, J., Joosse, J.: On the influence of the salt content of the
medium on some special neurosecretory cells in the lateral lobes
of the cerebral ganglia of Lymnaea stagnaUs. Proc. Kon. Ned. Akad.
v. Wetensch., Amsterdam, C 64, 630-639 (1961).
Llisis, 0.: P0stembryonic changes in the reproductive system of the slug
Arion ater rufUS L. Proc. Zool. Soc. London ill, 433-468 (1961).
Martoja, M.: Endocrinology of Mollusca. In: Chemical Zoology, Vol. VII
Mollusca (eds. M. Florkin and B.T. Sheer). Ch. II, p. 349-391.
New York-London: Academic Press 1972.
Mol, J.J. van: Phenomenes neurosecretoires dans les ganglions cere-
broide d' Arion rufus. C.R. Acad. Sci. Paris 250, 2280-2281 (1960).
Mol, J.J. van: Etude morphologique et phylogenetique du ganglion cere-
broides des Gasteropodes Pulmones (Mollusques). Acad. Roy. Belg.
Mem. Sci. 37 (5), 1-168 (1967).
Nolte, A.: Neurohamal-"Organe" bei Pulmonaten (Gastropoda). Zool. Jb.
Anat. 82, 365-380 (1965).
Pelluet, ~: On the hormonal control of cell differentiation in th~
ovotestis of slugs (Gastropoda Pulmonata). Canad. J. Zool. ~,
195-199 (1964).
Pelluet, D., Lane, N.J.: The relation between neurosecretion and cell
differentiation in the ovotestis of slugs (Gastropoda: Pulmonata).
Canad. J. Zool. 39, 691-805 (1961).
Runham, N.W., Larye~ A.A.: Studies on the maturation of the reproduc-
tive system of AgrioUm= reticulatus (Pulmonata, Limacidae). Malaco-
logia 7, 193-208 (1968).
Runham, N~W., Bailey, T.G., Laryea, A.A.: Studies on the endocrine
control of the reproductive tract of the grey field slug AgrioUm=
reticulatus. Malacologia li, 135-142 (1973).
Sminia, T.: Structure and function of blood and connective tissue cells
of the freshwater pulmonate Lymnaea stagnaUs studied by electron
microscopy and enzyme histochemistry. Z. Zellforsch. 130, 497-526
(1972) •
Smith, B.J.: Maturation of the reproductive tract of Arion ater (Pulmo-
nata: Arionidae). Malacologia 4, 325-349 (1966).
Streiff, W.: Recherches cytologiques et endocrinologiques sur le cycle
sexuel de Calyptraea sinensis L. (Mollusque Prosobranche hermaphrodite
protandre). Thesis Toulouse, No. 294 (1967).
Wattez, Chr.: Effet de l'ablation des tentacules oculaires sur la go-
nade en croissance et en cours de regeneration chez Arion subfuscuS
Draparnaud (Gasteropode Pulmone). Gen. compo Endocr. ll, 1-8 (1973).
Wendelaar Bonga, S.E.: Ultrastructure and histochemistry of neurosecre-
tory cells and neurohaemal areas in the pond snail Lymnaea stagnaUs (L.)
Z. Zellforsch. 108, 190-224 (1970).
Woerdeman, M.W., Raven, C.P.: Experimental embryology in the Netherlands
1940-1945. New York: Elsevier 1946.
Protandric Hermaphroditism in Prosobranch Gastropods
s. Le Gall and W. Streiff
Very few investigations have been carried out about factors responsible
for sex differentiation in protandric hermaphrodite molluscs. In Crepi-
dula plana, Gould (1947) showed the occurrence of a feminizing influence
when individuals are isolated and a masculinizing effect of the female
at the bottom of the chain on upper individuals of the association.
However, experiments of Coe (1942) on Crepidula onyx and Wilczynsky
(1958) on Crepidula fornicata, partly disprove Gould's work.
More recent studies, using adaptations of the organotypic culture

method developed by Wolff and Haffen (1952) have shown the occurrence
of factors which influence sex differentiation and the functions of
the gonads and the genital tract in Calyptraea sinensis (Streiff,1967),
Crepidula fornicata (Lubet and Streiff, 1969; Le Gall, 1974) and Patella
vulgata (Choquet, 1969).
A. Gonadal Differentiation Factors
Gonad differentiation factors which were only known in Calyptraea sinensis

have also to some extent been found in Crepidula fornicata.
The fundamental phenomenon of sex differentiation is oogonial self-

differentiation of undifferentiated protogonia.
Culture of isolated gonadal primordia on anhormonal medium shows that

protogonia transform themselves directly into oocytes, without any
further oogonial mitosis or vitellogenesis.
Thus, the gonad of Calyptraea is fundamentally female and its male dif-
ferentiation must necessarily depend on the influence of a masculiniz-
ing factor.
Experiments using the organotypic culture proved that male diff.eren-

tiation of protogonia and maintenance of the male phase depend on a
masculinizing factor which is released by cerebral ganglia during the
male phase.
The presence of this factor opposes oogonial self-differentiation of

protogonia.
B. Factors for Gonad Function
Oogonial self-differentiation alone cannot achieve the normal evolution

of oocytes in Calyptraea. .In particular, oogonial mitosis and vi tello-
genesis never occur.
171
Studies by organotypic cultures demonstrated the existence of two fac-

tors which complete oogonial self-differentiation:
a) a cerebral mitogenic factor which is responsible for oogonial mul-
tiplication,
b) a cerebral previtellogenetic factor controlling processes involved
in the preparation of vitellogenesis.
These factors act as an impulse at the moment of sex inversion and at

each new period of oogenesis.
Factors influencing male development could not be established. In this

species spermatogonial multiplication cannot be distinguished from
male differentiation.
In Patella, which was studied by Choquet (1969), oogonia and spermato-

gonia are present in the gonads of very young individuals; sex differ-
entiation has already taken place, but several function-promoting fac-
tors could be demonstrated:
a) a cerebral mitogenic factor, which controls the multiplication of

spermatogonia and oogonia,
b) a cerebral vitellogenic factor,
c) a tentacular factor which inhibits spermatogenesis and determines
the resting periods in males.
In this case, the cerebral mitogenic factor, which acts on spermato-

gonia previously differentiated represents an active factor different
from any sex differentiation factor.
In the species so far investigated, differentiation factors or processes

are not sufficient to accomplish the whole male or female gametogenesis.
For such a complete gametogenesis the intervention of several factors
must be assumed: mitogenic, previtellogenic factors, and an inhibiting
principle.
It is a striking fact that these functional agents are not sexualized:

it is rather the reaction of the differentiated male or female gonia
which characterizes the actual sex.
In Patella, Choquet (1969) clearly established that the cerebral mito-

genic factor released by the cerebral ganglia of female individuals
affects spermatogonia, and, inversely, the factor released by the male
brain acts on oogonia when they have reached the threshold of respon-
siveness.
In the same way, in Calyptraea, the cerebral previtellogenic factor re-

leased at the period of sex change triggers vitellogenesis not only
in oocytes but likewise in atypic spermatozoa and somatic cells of the
ovarian wall. The full development of the gonads of either sex involves
two processes:
a) a sex-differentiating mechanism which results from oogonial self-

differentiation or from the influence of a masculinizing factor,
b) a process depending on the influence of factors that act on pre-
viously differentiated gonia and allow complete gametogenesis.
172
C. Sexual Cycle and Comparison of Sex Inversion in Calyptraea and PateUa
In Calyptraea, two antagonistic processes control the change of sex:

oogonial self-differentiation and the effects of a cerebral masculiniz-
ing factor.
At the beginning of the animal's life, the cerebral masculinizing fac-

tor, released by the male nervous system, prevents oogonial self-dif-
ferentiation and achieves male differentiation.
At the moment of sex inversion, the masculinizing factor disappears,

oogonial self-differentiation starts, and the function-promoting fac-
tors bring about complete oogenesis (Fig. 1).
Immature
Male Female
p v. Pv.
" I "'
.
I \
\
I
, r \
Spermatogenesis ,', Oogenesis \
I ' I \
__ ________~~J
\
,'Ot \ I
r~~~~~~====~======~~===-
- ~~~'~~
......, -" ___________ \~\~-r 0
5 10 15 20mm
Fig. 1. General scheme of endocrine mechanisms controlling the gonadal

cycle in Galyptraea sinensis (from Streiff, 1967) . ....... Cerebral mascu-
linizing factor; ------- cerebral previtellogenic factor; oogo-
nial self-differentiation. Dt Atypic deutoplasmic spermatogenesis;
Pv Previtellogenetic factor
In this general scheme, the change of sex is attributed to active sub-

stances. The presence or absence of the masculinizing factor controls
sex inversion and explainsprotandric hermaphroditism.
In Patella the young gonad already contains spermatogonia and oogonia.

Differentiation has occurred previously. Thereafter, the gonadal cycle
and change of sex are only controlled by function-promoting factors.
During the male phase the mitogenic factor operates on the two types
of gonia, but the oogonia have not yet reached the threshold of respon-
siveness. Only spermatogonia divide themselves and spermatogenesis is
accomplished.
spermatogenesis is arrested by the tentacular inhibiting factor at

each resting period. Change of sex occurs when the oogonia become sen-
sitive to the mitogenic factor during a male resting stage. The oogo-
nia, less sensitive than spermatogonia to the inhibiting factor, mul-
tiply actively whereas spermatogonia are blocked.
In this species, sex inversion only occurs by the interplay of asexual

factors, and thresholds of sensitivity in the gonia to those factors
(Fig. 2).
173
Immature
Male Change of sex Female
I" .... ,
""" ... ,
,
I
I
.... " ......
:. I I
:
..
I I
\ I I
I I
I I
· t. ..... ..
I I
:1·
• I
. I
I
I
,
I
I
I
I
I I, :
.~I .'.
a r 2 a 3 a 4 a 5 yea rs
Fig. 2. General scheme of endocrine mechanisms controlling the gonadal

cycle in Patella vulgata (from Choquet, 1969) . . . . . . . . Cerebral mitogenic
factor; -.-.-.- tentacular inhibiting factor; ------- cerebral previ-
tellogenic factor; evolution of oogonias' thr~shold of sensi-
tivity to mitogenic factor. a Gametogenic activity; r resting period
The comparison of these two types in the group of protandric pro so-
branchs demonstrates a remarkable difference. Whereas in Patella only
factors which regulate functions bring about sex inversion, in Calyptraea
change of sex is directly controlled by differentiation factors. This
gives some ideas of the possibilities of variation occurring in sex
change mechanisms in the prosobranch group.
D. Differentiation Factors of the External Genital Tract
The only experimental data are those on Calyptraea sinensis (Streiff, 1967)
and Crepidula fornicata (Lubet and Streiff, 1969; Le Gall and Streiff,
1974; LeGall, 1974).
Organotypic culture technique demonstrated the neu~rality of the ex-

ternal genital tract in these molluscs. These organs, when completely
isolated or when parts of them were cultured on an anhormonal medium
show no involution or advancing development over a period of more than
200 days.
Such a neutrality implies that any change in the organ, advancement or

retrogression, can only be brought about by factors that operate on
this organ. Organotypic culture associations demonstrated the existence
of three factors which are responsible for the cycles of the genital
tract:
a) a morphogenetic factor for the external male genital tract, released
by the right ocular tentacle. This factor appears in the young male and
disappears progressively in the female,
174
b) a morphogenetic factor for the female genital tract, released by

the central nervous system. It appears at the period of sex inversion
and seems to disappear afterwards,
c) a retrogressive factor for the male genital tract, released by the
pleural ganglia. This factor appears in the male and becomes prepon-
derant when the change of sex begins.
The cycle of the external genital tract during the male phase (when
the male genital tract develops and is maintained) occurs by the action
of the tentacular morphogenetic factor (lately recovered in the pedal
ganglia - Le Gall and Streiff, 1974) in spite of the presence of a
retrogressive factor.
At the moment of sex inversion, the female genital tract is established

by an impulse which originates from a factor released by the central
nervous system.
During the female phase, the pedal morphogenetic factor disappears

progressively while the pleural retrogressive factor becomes more and
more active (Fig. 3) (Stre:i,.ff, 1967).
Male Change Female

1m mature Penis of sex Oviduct
-~----------------~~~-~.--------------------------------~.--
I ,
•.• " " I I
.' ..... I
,.... I
.: I
, ••• .1
I···.•••.
...
I I
.
I I •••.
.. I, '.
I I '"
I I ••..
: I
I
.I
I
,_
'"
............ 0)
5 10 15 20mm
Fig. 3. General scheme of endocrine mechanisms controlling the genital

tract cycle in Calyptraea sinensis (from Streiff, 1967). • ...•.. Male
morphogenetic factor; ------- female morphogenetic factor;
pleural retrogressive factor
These factors have also been found in Crepidula fora.i.cata. By the use of
several convergent techniques (organotypic culture, organ removal or
grafting) an annual activity cycle of the retrogressive factor and a
gradient of effectiveness of this factor related to the position of
the individual in the association could be demonstrated (Le Gall, 1974).
An exhaustive study enabled us to localize the origin of the morpho-

genetic factor for the penis and the role of the right ocular tentacle.
In the right ocular tentacle of the male, the morphogenetic ~actor is

concentrated in the basis of the tentacle. A dif~usion gradient of the
factor occurs around this zone: the distal end of the tentacle, the
basal region of the penis, and the insertion zone of tentacle and penis
in the neck display a weak effectiveness in culture.
175
No other part of the animal ever shows any morphogenetic activity.

The base of the tentacle shows an annual activity cycle (maximum in
January, minimum in April) and a decreasing activity-gradient from
young male to old female (Le Gall, 1974).
This factor is not specific: it is active in hermaphrodites as well as

in gonochorists (Streiff and Le Breton, 1970).
In the pedal ganglia, the only part of the perioesophagial ring which
displays morphogenetic activity, the right pedal ganglion is more ac-
tive than the left one, and maximal activity occurs in the median third.
An analysis of the efficiency of the factor released by this zone dem-
onstrates that the annual cycle and the efficiency gradient related
to the individual's position in the association overlaps with the in-
fluence exerted by the tentacle. The pedal factor is also non-specific
and shows activity in gonochorists as well as in hermaphrodites (Le
Gall and Streiff, 1974).
The identity of action, the parallelism between annual cycle and effi-
ciency gradient of tentacle and pedal ganglia, show the identity of the
substances released and raise the problem of the mutual relationship
between these two organs.
Numerous experiments of organ removal or grafting provided clarifica-

tion of the role of the tentacle (Le Gall, 1974).
From these experiments (Fig. 4) we can conclude that the tentacle as

well as the pedal ganglia are indispensable for penis regenerative
morphogenesis. After repeated removal of the right ocular tentacle the
penis never regenerates. Although absence of the pedal ganglia does not
prevent tentacle regeneration, their presence is essential for penis
regenerative morphogenesis. After removal of the pedal ganglia, re-
grafting allows regeneration of the penis. Finally, the most striking
and significant result is the fact that after simultaneous removal of
pedal ganglia and right ocular tentacle, the regenerated tentacle re-
veals itself as inactive in organotypic culture, and never induces any
morphogenetic process. This observation indicates that the pedal ganglia
have the essential role in the elaboration of the morphogenetic factor
of the penis. The ocular tentacle seems to be no more than a storage
zone for the morphogenetic factor produced in the pedal ganglia.
Experiments carried out in order to localize the cells responsible for

the elaboration of the morphogenetic factor show the existence of neu-
rosecretory cells in the active zone of the right pedal ganglion. An
experimental correlation seems to exist between synthetic activity of
elementary granules in neurosecretory cells and morphogenetic effec-
tiveness of the ganglia. The latter disappears during the resting
period of the neurosecretory cells.
These observations support the assumption that the morphogenetic factor

is synthetized by the right pedal ganglia.
A structural study of the tentacle shows no particular secretory cell

from which the morphogenetic factor could originate. On the other hand,
the right ocular tentacle - and only the right one - has in its basal
region (besides the normal circulatory system) a dense lacunar network.
This is linked up to the arteries responsible for buccal irrigation by
a narrow duct controlled by a sphincter. Such a special semi-closed
system resembles a portal system and seems to constitute an anatomical
basis for the function of a storage place assigned to the right ocular
tentacle (Le Gall, S., and Le Gall, P., 1974).
176
Fig. 4 A-E. Demonstration of the respective role of the right ocular

tentacle and pedal ganglia in the morphogen e sis of the penis (from
Le Gall, 1974). (A) Penis regeneration in presen£e of pedal ganglia;
(B) absence of penis regeneration after removal of pedal ganglia;
(C) penis regeneration after grafting of pedal ganglia; (01) penis
and tentacular regeneration in presence of pedal ganglia; (02) ten-
tacle regenerated without removal of pedal ganglia, induces penis
morphogenesis in culture; (El) absence of penis regeneration, in spite
of tentacular regeneration after removal of pedal ganglia; (E2) when
regenerated after removal of pedal ganglia, the tentacle has no mor-
phogenetic activity in culture; P penis; Px pedal ganglia; T tentacle
The morphogenetic system seems to be organized in the following way:

the morphogenetic factor is produced by the medial part of the right
pedal ganglion. It is then accumulated in the tentacular "portal system",
177
from where it diffuses to tpe adjacent morphogenetic territory. This

factor, very active in the young male, disappears progressively in the
female.
The cycle of the external male genital tract in CY'epidula is governed

by variations of two antagonistic substances that diffuse into the
haemolymph:
a) a pedal morphogenetic factor, mainly concentrated in the right ocu-
lar tentacle, brings about morphogenesis in the young and disappears
in the female when the period of sex inversion starts,
b) a retrogressive factor released by the cerebropleural junction is
present in the young male, becomes preponderant during sex inversion
and induces the involution of the external male genital tract in the
female phase.
According to variations of the ratio between these two antagonistic

factors, important differences may occur in the cycle of the external
male genital tract. Thus, the penis can disappear or be maintained
during the female phase depending on local ecological factors varying
from station to station (Le Gall, 1974).
E. Conclusion
From the results of previous experiments some fundamental aspects of

the process of sex differentiation and of the realization of protandry
in prosobranch gastropods is becoming apparent.
All the processes are controlled by the nervous system which leads the
gonadal primordia either into male or female direction and achieves
gametogenesis by the liberation of essential function-promoting factors.
The nervous system in a synchronous way, also directs differentiation
and morphogenesis of the successive genital tracts.
The harmonization in the differentiation of the gonads and the genital

tracts in protandric hermaphrodites suggests the existence of control-
ling and synchronizing factors for the differentiating activities of
the central nervous system. Present studies pursued along these lines
should soon provide results about the modalities of action of ecolog-
ical factors in sex differentiation.
References
Choquet, M.: Contribution a l'etude du cycle biolegique et de l'inver-

sion du sexe chez Patella vulgata L. (Mollusque Gasteropode Proso-
branche). These doct. Sci. Nat. Lille, no. 185, 1-234, 19 pl. (1969)
Coe, W. R.: The reproductive organs of the prosobranch mollusk CY'epidula
onyx and their transformation during the change from male to female
phase. J. Morph. 70, 501-512 (1942).
Le Gall, S.: Determinisme de la morphogenese et du cycle du tractus
genital male externe chez un Mollusque Hermaphrodite Protandre
CY'epidula fomicata Phil. These doct. Sci. Nat. Caen (1974). A.O.:
10.739.
Le Gall, S., Le Gall, P.: Existence d'une lacune sanguine au niveau
du tentacule droit et du penis chez CY'epidula fOY'nicata Phil. Role
dans la concentration du facteur morphogenetique d'origine pedieuse.
Soc. Zool. Fr. (1974).
178
Le Gall, S., Streiff, W.: Presence du facteur morphogenetique du penis

au niveau des ganglions pedieux chez des Mollusques Prosobranches
hermaphrodites (Crepidula, Calyptraea) et gonochoriques (Littorina,
Buccinum). C.R. Acad. Sci. Paris 279, 183-186 (1974).
Gould, R.N.: Conditions affecting the development of the male phase in
CrepiduZapZana. Biol. Bull. 93, 194 (1947).
Lubet, P., Streiff, W.: Etude experimentale de l'action des ganglions
nerveux sur la morphogenese du penis et l'activite genitale de
CrepiduZa forniaata Phil. (Mollusque Gasteropode). (Ed. Gordon and
Breach). Cours et Documents de Biologie, I., pp. 141-159. Paris
1969.
Streiff, W.: Recherches cytologiques et endocrinologiques sur le cycle
sexuel de CaZyptraea sinensis L. (Mollusque Prosobranche hermaphrodite
protandre). These doct. Sci. Nat., Toulouse, no. 294, 1-26 L, 29 pl.
(1967).
Streiff, N., Le Breton, J.: Etude comparee en culture in vitro des fac-
teurs responsables de la morphogenese et de la regression du tractus
genital male externe chez deux Mollusques gasteropodes prosobranches:
CrepiduZa forniaata Phil. (espece protandre) et Littorina Uttorea L.
(espece gonochorique). C.R. Acad. Sci. Paris 270, 632-634 (1970).
Wilczynski, J.Z.: On sex behaviour and sex determination in CrepiduZa
forniaata L. Publ. Univ. Libanaise, SC. Nat. 31 p. (1958).
Wolff, E., Haffen, K.:·Sur une methode de culture d'organes embryon-
naires in vitro. Texas Rep. Biol. Red. l.Q., 463-472 (1952).
The Sexuality of Pelecypod Molluscs
P. Herlin-Houtteville and P. E. Lubet
Several older studies are devoted to the sexual organization (herm-

aphroditism or gonochorism) of many pelecypod species. However, the
study of the sequence and the modalities of the sexual cycle has been
performed in a satisfactory way (to a very limited extent) by the use
of thorough cytological investigations.
The study of internal factors controlling the sexual differentiation

and the activity of gonads is poorly known, because of the technical
difficulties inherent in such ~ type of research. Therefore, our at-
tention will be focused on the analysis of the few experimental results
that have been achieved in this field.
A. Gonochorism and Hermaphroditism in Pelecypods
Earlier reviews in this field have been provided by Coe (1943), Franc
(1960), Fretter and Graham (1964).
I. Gonochorism
According to Coe (1945) gonochorism seems to be the rule among pelecy-

pods. Less than 400 out of 10 000 known species are hermaphrodites.
However, only some of those which are actually known as gonochorists
have been studied. The possibility cannot be excluded that a part of
these will be recognized as hermaphroditic.
Placopecten magellanicus described as gonochoristic by Drew (1907), Dakin

(1908), Coe (1949) is, in fact, hermaphroditic (Merril and Burch,
1960) .
Anodonta cygnea is gonochoristic in running waters, whereas hermaphro-

ditic animals can exist in different environmenta~ conditions (e.g.
ponds, Weisensee, 1916).
A very short juvenile hermaphroditic phase (Glycimeris glycimeris, Lucas,

1964), or a longer one (Mercenaria mercenaria, Loosanoff, 1937), may pre-
cede a final gonadal differentiation into males and females (consecu-
tive sexuality, according to Coe, 1943).
II. Hermaphroditism
Distinctions between the various types are somewhat artificial because

of the many gradual transitions that are known.
180
1. Simultaneous Hermaphroditism (Functional Hermaphroditism, Coe, 1943;

Synchronous Maturation, Bacci, 1951)
The coexistence of both germinal lineages (male and female) at the

same stage of cytological development, can be seen within the annual
sexual cycle. The morphological aspects of this hermaphroditism can
vary greatly from one species to another. In Tered£J digenis (Coe, 1941),
both germinal cells coexist within the same gonadal follicles, whereas
Laevicardium crassu.m produces male or female gametes in adjacent folli-
cles. Two distinct male and female gonads can exist in Pandora elongata
(Pelseneer, 1894) or in Musculinum heterodon (Okada, 1935). Finally, the
gonad can show'two distinct areas, a proximal male part and a distal
female one (Pecten).
2. Successive Hermaphroditism
a) The asynchronous maturation of the germinal cells in the same go-

nadal follicle, during the annual sexual cycle, was found in shipworms
and oysters (Ostrealurida, Coe, 1932; Ostreaedulis, Cole, 1942; Teredo
navalis, Coe, 1944). This is known as "rhythmical consecutive sexuality"
(Coe, 1 943) .
b) Finally, some pelecypods can show an entirely male or female gonad

lasting one or several annual sexual cycles and, eventually, perform
a sex change later on. Thus, there is a maturation (Bacci, 1951) known
as "alternative sexuality" according to Coe (1945). This was found in
Crassostrea virginica (Coe, 1938; Galtsoff, 1938), in Chlamys varia (Lubet,
1959; Lucas, 1965) and in Crassostrea angulata (Le Dantec, 1968).
Bacci (1951) attempted to give a genetic explanation of this different

sexual behavior. Recent advances performed on the breeding of pelecypod
larvae give way to an experimental approach of this problem.
B. The Internal Factors Controlling the Sexual Differentiation and the

Gonadal Activity
In Gastropod molluscs, the experiments carried out by Streiff (1967)

on Calyptraea sinensis, by Lubet and Streiff (1969) on Crepidula fornicata,
by Griffond (1969) on Vivipara vivipara and by Guyard (1971) on Helix
aspersa, showed an ovarian self-differentiation. Such a study has not
been undertaken in pelecypods and nothing is known about the factors
controlling the sex differentiation in lamellibranchs. The few experi-
mental results obtained so far are related to the gonadal activity;
they were found by studying gonochoristic species (Mytilus edulis and
Fig. 1. (A) Oogenesis in the female mantle (stage III B and III C in ~
January). Formalin - Gabe's trichrom staining x 150. A female acinus;
c.v. vesicular cell; adpg adipo-granulous cells. (B) Storage tissue
during rest stage (stage 0), x 3750. adpg adipo-granulous cells;
gp oblong granule; L lipid droplet; cv vesicular cell filled with poly-
saccharids; mb cytomembrane (phosphotungstic acid). (C) Lysis of the
storage tissue in the female (Richardon x 600). au adipo-granulous cell
with a big autophagosome. (D) Coalescence of lysis material from the
storage tissue around clusters of germ cells (Richardson x 450).
mg clusters of germ cells; cv vesicular cells; adpg adipo-granulous
cells; ml lysis material
181
Fig. 1 A-O
182
Mytilus gaUoprovincalis, Lubet, 1959, 1966, 1969; Houtteville and Lubet,

1974; Houtteville, 1974).
In Mytilidae, the gonad invades the connective tissue of the mantle

(Fig. 1). The primordium of the gonad, which is located medio-dorsally
in the juvenile animals, spreads into the visceral mass and both lobes
of the mantle in. the mature animals.
The different kinds of cells making up the connective tissue show a

very complex development during the annual sexual cycle. On the other
hand, the gonad develops in a totally reverse manner (Lubet, 1959;
Houtteville, 1974). Studies by Dupouy and Martinez (1973) and our-
selves, on parasitized mussels, as well as the results of the first or-
gan cultures performed in our laboratory (Houtteville and Lubet, 1974)
allowed us to demonstrate a close interdependence between the storage
tissue and the gonad - which seems to be controlled by neuro-endocrine
factors.
I. Interrelationship between the Sexual Activity and the Storage Tissue
The annual sexual cycle in Mytilus edulis varies from one geographical
region to another. However, four distinct and successive phases can
be seen (Chipperfield, 1953; Lubet, 1959). Lubet calls the sexual rest
phase in summer, stage 0; during this period the gonad does not present
any functional acini. The highly developed interfollJ..cular storage
tissue is represented by two kinds of resting cells; the vesicular cell,
an enormous bag of polysaccharids; and the granulous adipous cell,
filled with lipid droplets, protein granules and glycogen (Fig. 2).
Stage I is a phase of differentiation and active multiplication of
gonia. It occurs in the Channel around September-October. During stage
II (November through December), the mantle presents a typical aspect
of the differentiation and maturation phase. The storage tissue is
then profoundly altered. As the acini develop, the volume taken up by
the interfollicular tissue gets smaller. Some areas of lysis appe~r
in vesicular and adipo-granulous quiescent cells (Fig. 1C): in the
-
Storage progressive
, restoration
tissue max~.. disappe.arance .. disappearance beginning of
restoration
Gonad ---..multiplicat. .. , genital instability .....--

rest .. different. .. . " rest
gonia maturation(A)discharge(B.D)restor. (C)
- -
Reproductive
III
~
stages
------0--- 'B '\
! ! ! !
atter l U B ET B B 0 0
1959 • • II
C C
A C C ~
month
A s o N o M A M
Fig. 2. Sexual cycle of Mytilus edulis (L.)

183
areas of quiescent cells filled with lipids, glycogen and protein gran-
ules, homogeneous pools can be seen in which pycnotic nuclei, cell bodies,
protein granules and isolated lipid droplets are identified.
Within such a material or in its immediate environment, germinal cells

develop and multiply (Fig. 1D). Pinocytosis vesicles have so far been
identified, only in the male, at the basis of germ cells at the periph-
ery of the acinus (Herlin, 1974). Such observations lead us to think
that gonia could use the material freed by storage tissue in order to
make their own synthesis. Stage III is related to the reproductive
period. On our coast it occurs from January to June. Gametes mature
(stage III A) and then, are discharged. The mussel has several cycles
of maturation and successive spawnings (stage III B for early ones and
III D for the later ones). Between each discharge of gametes, a re-
covering phase (III C) occurs; during this stage (III C) gonia that
were immature during the previous discharge, start or achieve their
differentiation and maturation.
When the period of sexual activity is reached the necrosis of the

storage tissue increases. A gradual disappearance of this tissue can
be noticed and it is confined to a small number of flattened cells
between the acini, or pushed under the epithelium. This almost total
disappearance can be noticed for both males and female in February -
March, i.e. in the middle of stage III (Fig. 3A).
Late maturations of the second half of stage III only concern a very
small number of gametes and the acini only reach a smaller size. The
storage tissue quickly rebuilds itself in the intertubular space,
which is empty because of the successive discharges of gametes (Lubet,
1959) .
A more detailed study (Houtteville, 1974) allowed us to observe that
the alternating mechanism, storage tissue-gonad, is stricter than Lubet
(1959) thought, since a certain number of storage cells reappears be-
tween the acini at the beginning of each recovering stage (III C),
and this occurs from the very beginning of each stage III (Fig. 4).
Young vesicular, then adipo-granulous cells differentiate, probably

from amoebocytes and macrophages which appear inside and outside the
acini after each spawning (Fig. 3C and D). These amoebocytes secure
the resorption of residual or atretic gametes (Fig. 3B) and stock
their metabolites. During the first part of stage III, the sexual ac-
tivity is so important that the storage tissue does not have the time
to build at the same speed as it undergoes lysis. Moreover, during
the second part of stage III, gametic activity is lower and very likely
requires less metabolites. Therefore, the formation of the storage
tissue prevails over the lysis.
Consequently, the storage tissue and the germinal tissue grow in two
opposite directions (Fig. 4).
II. Survey of Different Studies Related to Parasitic Castration
Mytilus gallopY'ovincalis (from the "Etang de Thau"), is used an an inter-

mediate host for the trematode PY'octoeces maculatus. The sporocysts of the
parasite settle preferentially in the storage tissue of the mantle of
the mussel.
In classical histology, Dupouy and Martinez (1973) studied the influ-

ence of the parasite on the sexual cycle of the host.
184
Fig. 3 A-D
185
,, i , 1 1 ,
~ ,
", ' "~,1 w,1
E'I
1
1
1
~ ,
'"
1
,, .c'",
~ I I I ~t "'I
.c:0u ,,,
, cl
.. ,
1 ~ I
l .cu ' l 1 .c ,
,
, ~
1 -
I
:, .."-; : u,
M,
1 ; 1 :; .
, 1
qu iescent 1 1 1
, :
1 , 1
storage l issue 1 ,
1 1 '
11 1'
1 ,
Fig. 4. Scheme of the relations between gonad cycle and changes in

the storage tissue
Castration only occurs when two conditions are found: the invasion by
miracidia during the sexual rest period (stage 0); and the maximal
proliferation rate of sporocysts before the proliferation of gonia.
This is an indirect castration, since the germ cells do not seem to
be affected by the parasite. The sporocysts simply consume the storage
substances of the intertubular tissue (especially glycogen) and one
notices that the s e xual cycle is stopped later on.
These observations suggest that the sexual cycle of the Mytilidae must
~e closely related to the storage state of the animal (Daniel, 1921).
The storage tissue very likely has a trophic control over the gonad,
which takes its metabolites and uses them for its own synthesis.
This perfect alternation of the gonad and storage tissue cycles led
us to assume that such a mechanism could be controlled by neuro-endo-
crine factors. The first results from organ cultures seem to agree
with this idea.
III. Results of Organ Culture Experiments
Mantle fragments of male Mytilus eduli s were taken from individuals

during gametogenesis (stages II and III) and maintaine d in culture,
~ Fig. 3. (A) Spermatogenesis in a male mantle (stage III in February).

Formalin, Gabe's trichrom staining x 38. A male acinus; ec connective
tissue-space; e mantle epithelium. (B) Resorption of two atretic 00-
cytes in th e storage tissue during stage III C (Richardson x 375) .
A female acinus; R storage tissue; ov residual oocyte; N oocyte nucleus
and cytoplasm (e) within the storage tissue. (C) Spermatozoids near a
cluster of amoebocytes. Formalin - Gabe's trichrom staining x 300.
sp spermatozoids within the storage tissue; a cluster of amoebocytes;
cv vesicular cells; A male acinus
186
either associated to cerebro-pleural or visceral ganglia (cultures were

performed on Streiff's (1967) medium).
A histological and ultrastructural study of the explants is proceeding.

The first results obtained, after a 13 day's culture, were reported in
a former publication (Houtteville and Lubet, 1974). These results agree
with the experiments on ganglia extirpations performed by Lubet (1959-
1966) .
In cultures of isolated mantle fragment (taken during stage II) only

the presence of cerebro-pleural ganglia can maintain the genital matu-
ration. Storage tissue is better preserved in brainless aminals than
in sham-operated controls. The brainless specimens (during stage III A)
show a heavy infiltration of amoebocytes into the lumen of gonadal tu-
bules as it occurred in cultures (mantle + visceral ganglia) .
In conclusion, a survey of our experiments shows that the internal

factors controlling the sexual differentiation and sex changes in
pelecypods are still unknown. The only results obtained so far concern
the gonadal activity. As for the mussel, the storage tissue is abso-
lutely necessary for the development of germinal lineages (spermato-
genesis and oogenesis). The visceral ganglia seem to have a neuro-
endocrine control over the storage tissue, the cerebro-pleural ganglia
over the germinal lineage.
References
Bacci, G.: Ermafroditismo ed intersessualita nei Gastropodi e Lamelli-

branchi. Archiv Zool. Ital. 7, 57-151 (1951).
Chipperfield, P.N.J.: Observations on the breeding and on the settle-
ment of Mytilus edulis in the British waters. J. Mar. Biol. Ass. U.K.
32, 449-476 (1953).
Coe~W.E.: Development of the gonads and the sequence of the California
oyster (Ostr'ea lurida). Bull. Scripps Inst. Ocean. l (6), 119-144
(1932) .
Coe, W.E.: Primary sexual phases in the oviparous oyster (Ostrea virgi-
nica). Biol. Bull. 74 (1),64-75 (1938).
Coe, W.E.: Sexual phases in wood boring Mollusks. BioI. Bull. 81 (2),
1 78-1 86 (1 941 ) .
Coe, W.E.: Sexual differentiation in Mollusks. Quart. Rew. BioI. ~
(2),154-164 (1943).
Coe, W.E.: Development of primary gonads and differentiation of sexual-
ity in Teredo navalis and other Pelecypod Mollusks. Biol. Bull. ~ (2),
178-186 (1944).
Coe, W.E.: Development of reproductive system and variations in sexual-
ity in Pecten and other Pelecypod Hollusks. Contr. Scripps Oceanogr.
Inst. no. 225 (1949).
Cole, H.A.: Primary sex phases in Ostreaedulis. Quart. J. Micr. Science
83, 317-356 (1942).
Dakin, W.J.: The edible scallop. Trans. Liverpool BioI. Soc. 11, 333-
468 (1908).
Daniel, R.J.: Seasonal changes in the chemical composition of the
Mussel. Lancashire Sea, Fish. Labo. Rep., p. 205 (1921).
Dupouy, J., Martinez, J.C.: Action de Protoecesmaculatus (Looss-1901)
(Trematoda, Fellodistomatida 1) sur Ie developpement des gonades
chez Mytilus gaUoprovincialis Lmk. C.R. Acad. Sci. Paris 277, 1889
(1973) .
Franc, A.: Classe des Bivalves. In: Traite de Zoologie, V (II), pp.
845-2133 (1960).
187
Fretter, V., Graham, A.: Reproduction. In: Physiology of Mollusca

(ed. Wilburg, K., Yonge, C.M.), I, pp. 127-164. London: Academic
Press 1964.
Galtsoff, P.S.: Sex change and physiological intersexuality in Ostrea
virginica. Anat. Rec. Suppl. 72 (a), 42-43 (1938).
Griffond, B.: Survie et evolution, en culture "in vitro", des testi-
cules de Viviparus viviparus L. Gasteropode prosobranche a sexes se-
pares. C.R. Acad, Sci. Paris 268, 963-965 (1969).
controlant l'orientation sexuelle des gonocytes de l'escargot HeZix
aspersa Muller. These Doc. Sci. Univ. Besanpon, no. 56 (C.N.R.S. -
A.O. 5939), pp. 1-183, 43 pI. (1971).
Herlin, P.: Absorption des metabolites par les cellules germinales
m~nes chez MytiZus eduUs (L) (Mollusque-pelecypode). C .R. Acad. Sci.
Paris (1974, in press).
Houtteville, P.: Contribution a l'etude cytologique et experimentale
du cycle annuel du tissu de reserve du manteau de MytiZus eduUs (L).
These Doc. 3eme cycle, Univ. Caen, pp. 1-98, 32 pl. (1974).
Houtteville, P., Lubet, P.: Analyse experimentale, en culture organo-
typique, de l'action des ganglions cerebropleuraux et visceraux
sur le manteau de la moule I).lale MytiZus eduUs (L.) (Mollusque Pele-
cypode). C.R. Acad. Sci. Paris 278, 2469-2472 (1974).
Le Dantec, J,.: Ecologie et reproduction de l'huitre portugaise (Cras-
sostrea anguZata LMK) dans le bassin d' Arcachon et sur la rive gauche
de la Gironde. Rev. Trav. Inst. Peches Marit. Paris 32, 1-126 (1968).
Loosanoff, V .L.: Primary gonad and sexual phases in Venus mercenaria L.
Biol. Bull. 72, 408-416 (1937).
Lubet, P.: Recherches sur le cycle sexuel et l'emission des gametes
chez les Mytilidae et les Pectinidae (Moll. Bivalves). Rev. Trav.
Inst. Peches maritimes 23 (4), 387-548 (1959).
Lubet, P.: Essai d'analyse-experimentale des perturbations produites
par les ablations des ganglions nerveux chez MytiZus eduUs L .. et
Myti'Zus gaUoprovinciaUs LMK. Ann. Endocrinol. Paris 27 (3), 353-365
(1966) .
Lubet, P.: Donnees experimentales sur les effets des ablations de
ganglions nerveux chez les Mollusques Pelecypodes. Cours et Doc.
BioI. Gordon Breach, I, 161-173 (1969).
Lubet, P., Streiff, W.: Etude experimentale de l'action des ganglions
nerveux sur la morphogenese du penis et l'activite genitale de
CrepiduZa fornicata Phil. (Mollusque gasteropode). Cours et documents
de Biologie, 1, pp. 141-159. Paris: Gordon et Breach 1969.
Lucas, A.: Mise en evidence de l'hermaphrodisme successif chez GZycimeris
gZycimeris L. par l' analyse des pourcentages sexuels. C • .R. Acad. Sci.
Paris 285, 5742-5744 (1964).
Merrill, A:S., Burch, J.B.: Hermaphroditism in the scallop PZacopecten
mageUanicus Gm. BioI. Bull. ill, 197-201 (1960).
Okada, K.: Some notes MuscuUnum heterodon Philb., a freshwater bivalve.
Sc. Rept. Tohokn Imp. Univ. 1-25 (1935).
Pelseneer, M.J.: Hermaphroditism in Mollusca. Quart. J. Micr. Sci. 37,
19-46 (1894).
Streiff, W.: Etude endocrinologique du determinisme du cycle sexuel
chez un mollusque hermaphrodite protandre CaZyptraea sinensis L.
III. Mise eh evidence par culture "in vitro" de facteurs hormonaux
conditionnant l'evolution de la gonade. Ann. Endocrinol. Paris 28,
641-656 (1967).
Weisensee, H.: Die GeschlechtsverhiH tnisse und der Geschlechtsapparat
bei Anodonta. Z. wiss. Zool. 115 (1916).
Hermaphroditism In Echinoderms. Studies on Asteroids
R. Delavault
A. Generalities and Definitions
I. Hermaphroditism in Echinoderms
Hermaphrodites (individuals with germ cells of both sexes) are found

among all classes of Echinoderms. Such individuals do occur even in
gonochoristic species. Examples are found in Hyman's book (1955: Aster-
oids, cf. pp. 287-288; Echinoids, cf. pp. 478-479; Ophiuroids, cf.
p. 625) and, more recently, in Delavault's work (1966, cf. p. 616).
Other examples have been provided in the meantime, among the more re-
cent ones, those by Komatsu and Oguro (1972) and Achituv (1973) should
be mentioned.
In some species, however, this particularity affects the totality of

individuals; these species are called hermaphroditic (cf. Hyman, 1955,
Holothuroids, pp. 173, 176 and 198; Ophiuroids, pp. 623 and 625, and
Delavault, 1966: Asteroids, p. 617).
According to a previously adopted terminology, the first type mentioned

above was designated as "labile gonochorism" (Cognetti and Delavault,
1960a). The other type corresponds to "strict hermaphroditism". But if
we refer to studies on hermaphroditism in Echinoderms during the last
decade, two main aspects arise: the possibility of functional and po-
tential hermaphroditism.
II. Types of Hermaphroditism
Functional hermaphroditism is defined by a complete development of the

two gametogenetic tissues until the maturation of the gametes. It cor-
responds to the hermaphroditism sensu stricto.
As "potential hermaphroditism" are described such cases when in a spe-

cies nearly all individuais function like in gonochoristic species al-
though an examination of the gonads reveals that in reality germinal
cells of the other sex do occur also. To a certain extent, there are
similarities between labile and potential hermaphFoditism.
B. Hermaphroditism in Asteroids
This class is certainly the one which was the object of the largest
number of investigations on the problem of sexuality in Echinoderms.
I. Functional Hermaphroditism
During the last years most studies were devoted to Asterina gibbosa, but
other work nas also been done on Asterina pancerii and Fromia ghardaqana.
189
1. Sexuality in Asterina gibbosa, A. pancerii and Fr omia ghardaqana
Works which are to be taken into consideration are the following:

- Asterina gibbosa : BruslE~(1968a,b; 1969a,b)
Asterina pancerii : Delavault, BruslE~ and Pierre (1969)
- Fromia ghardaqana : Achituv and Delavault (1972)
Details in the sexuality of th e se three species are extremely complex

but, nev ertheless, some mai n features can be elaborated:
The first genital activity is characterized by a complete spermatogen-

esis (Fig. 1a). This stage seems to be limited to one year in Asterina
pancerii, whereas it is repeated again during 2 or 3 years at least in
A. gibbosa . Therefore, at the beginning of their life, the individuals
behave as functional males and hermaphroditism is protandric. But it
is ess e ntial to point out that the gonads of these animals always con-
tain young oocytes which never undergo a development beyond a certain
stage (Fig. 1a). Therefore, these gonads are hermaphroditic from a
cytological point of view.
Fig. 1 a and b. Successive aspects of the gonadaL activity in her-

maphrodi tic Asteroids. (a) Functional male of Asterina pancerii , Naples.
Initial spermatogenesis (spg .). Young oocytes (y . ov .) arre sted at an
early stage of their developme nt. (b) Later oogenetic activity with
matur i ng oocytes (ov .). Reappearance of slight spermatogenetic activity
(spg.) which leads to the maturation of spermatozoids (s pz.). The ani-
mal b e come s a functional hermaphrodite (Asterina gibbosa , Banyuls). (By
courtesy of Brusle)
This funct i onal spermatogenetic activity is followed then by an inversion

of sex with oogenesis resulting from maturation of the oocytes (Fig. 1b).
But, a spermatogenetic activity, usually discrete, reappears as sus-
pected by Hauenschild (1954) and confirmed later by Cognetti (1956,
190
1958) who calls it "the secondary spermatogenesis" ("resurging" sper-

matogenesis according to the more recent studies of Brusle, 1967,
1970a). Therefore, during this stage, the individuals function as true
hermaphrodites with the possibility of self-fertilization as it was
indicated for the first time by Cognetti (1956).
The change from the functional male stage to functional hermaphroditism

may show differences in the same species according to the geographical
localization of the populations. For example, in Asterina gibbosa, this
change takes place very regularly at the same age in all the individuals
which live at Plymouth (Bacci, 1949), Dinard (Delavault, 1960), Banyuls
(Brusle, 1967, 1970a), Marseille (Brusle, 1969a), whereas the change
extends over a much longer period of time for the populations of Naples
(Bacci, 1951), Roscoff (Neefs, 1959; Brusle, 1968a) where individuals
can be found which remain exclusively male or female during all their
life.
We know (Bacci, 1950, 1951) that these two categories of populations

correspond to "sexual races". The first type is called a "balanced
race", the second one an "unbalanced race". Delavault (1966, cf. p.
618) suggests a genetic explanation of the fixation of sex in these
races.
As to Asterina pancerii and Frorma ghardaqana, we recently related them

to the unbalanced protandric hermaphroditism (Delavault et al., 1969;
Achituv and Delavault, 1972).
2. The Origin and the Behavior of the Gonocytes
This problem has mainly been studied in Asterina gibbosa. It concerns,

on the one hand, the research on the first origin of the germinal tis-
sue, and on the other hand, the localization of the center of primor-
dial germ cells in grown up animals.
a) The Origin of the Germinal Tissue
On this object we refer to the statements by Lender and Delavault

(1964) on all the Echinoderms (cf. pp. 177-189).
Approximately 12 days after the hatching, during the larval develop-

ment before metamorphosis, gonocytes can be recognized cytologically
in Asterina gibbosa. These are grouped near the anlage of the hydrophoric
canal and near the anlage of the aboral part of the axial organ (cf.
the anatomy of this organ, Fig. 2b). We also see in close vicinity to
this aggregation ("genital bud"), mesenchymatous cells of which we
could presume that they might represent the origin of the gonocytes.
The topographical proximity between the axial organ and the genital
bud (Fig. 2a) persists until 4 or 5 months of age. At that time two
extensions grow from this bud in opposite directions (Fig. 2a). Later,
in adults, it never appears clearly that the genital bud had subsisted
in its initial position.
Therefore, former authors had concluded that the gonocytes, supposedly

of mesenchymatous origin, regrouped at first in the bud and would leave
it then in order to colonize the gonads directly.
The permanent center of the primordial germ cells in adults was sup-
posedly located in a "genital aboral ring", derived from the genital
bud. It is at least this opinion which became apparent from the first
investigations by Lender and Huet (1962a,b).
191
b
Fig. 2 a and b. First localization of the gonocytes. (a) The gonocytes
(g.) form a group in the genital bud (g. b .) located very closely to
the aboral part of the axial organ (a. p .). The bud pushes two exten-
sions (e x t .) in opposite directions. h . c . hydrophoric canal. (b) Anat-
omy of the axial organ. a . p . aboral part; h . c . hydrophoric canal;
o.p. oral part; T test. (By courtesy of Leclerc)
b) Origin and Localization of the Permanent Center of the Germ Cells

in the Adults
Experiments by Lender and Huet (1962a,b) and Huet (1965) showed clearly
that the regeneration of the gonads, in Asterina gibbosa , is possible only
when .t he center of germ cells which exists in the aboral part of the
individuals, in the genital ring, is not destroyed.
In this way we have a proof that a stock of germinal elements subsists
at any age in this area.
Nevertheless, the precise origin of this stock as well as the behavior

of the gonocytes was still to be studied thoroughly and investigations
were made in this way by Delavault and Leclerc (1967).
The starting point for this research was a study of the relations be-
tween the genital and axial complexes in Asteri na gibbosa of which we al-
ready know the topographical conditions (cf. Fig. 2a) during the larval
and post-larval development. This study confirmed entirely the existence
of these close relations which are established between the two complexes
and the authors conclude that, in these conditions, the axial organ was
liable to take part in the physiology of sexuality.
A new study of the morphogenesis of the genital complex first stated

precisely that the genital bud appears at the very same place where
the formation of the aboral part of the axial organ takes place. There-
fore, this part, indeed, contains the genital bud. From this bud is
built a "genital rachis" pushing strips which surround the madreporic
interradius. It is in these strips that the gonocytes are located.
192
Later, a "mantle" is formed which derives from the 'somato-pleura cov-

ering the dermis and which extends from the aboral to the oral. pole.
This formation represents the place of transit of the germ cells to-
wards the gonads. Therefore the germinal elements migrate at all levels
and not, as we thought before, via the interradial "genital cords"
after they left the "genital ring".
In the adults, we see that the aboral part of the axial organ does not
show germ cells any longer, apart from rare exceptions.
We could conclude from these observations that if the initial center

of the germinal cells forms a bud within the axial organ, this center
then disappears. Consequently, the aboral part of the axial organ could
not be considered as the stocking place of the gonocytes.
Nevertheless, we know that a potential source of new germinal elements

exists since the regeneration of the gonads is impossible without their
participation.
For this reason our work was devoted then to the search for this source.
Leclerc (1974a,b) has studied what happens in the aboral area when ger-
minal cells. are lacking as, for example, in gonads after the natural
emission of their products, or after experimental destruction of the
genital organs. In order to elucidate this latter point various experi-
mental procedures were used: X-irradiations of the gonads, cultures in
vitro of the aboral part of the axial organ after X-irradiation of the
gonads. Details will be found in Leclerc's survey (1974a, cf. pp. 26-27).
The results obtained demonstrate that approximately two days after the
emission of the gametes, or after the destruction of the gonads, big
cells are observed which, in the aboral region, derive from the splanch-
nopleura (Fig. 3a) and migrate towards the aboral part of the axial
organ where they accumulate and soon show the characteristic features
of gonocytes (Fig. 3b).
Later, in irradiated animals, a genital rachis is formed identical to

the one of young animals, and the migration of the gonocytes begins as
we have already described it.
In summary: It is clearly shown that the permanent source of gonocytes

derives initially from elements of the mesenchym which leave the coe-
lomic splanchnopleural sheet to reach the aboral part of the axial or-
gan and then rejoin the gonads.
Consequently, the axial organ corresponds to a relay-organ during the

germinal cytodifferentiation which seems to be a notable particularity
among Metazoa.?
3. The Evolution of Germinal Cells Inside the Gonads
The behavior of the germinal cells contained in gonads was the subject
of numerous investigations. They also concern Astenna gibbosa.
The guiding idea of these studies was to remove the gonads from their
usual environment in order to investigate the possible modifications
that could then affect their normal activity which we described above.
The first experiments were made by Delavault (1963). They essentially

consisted in transplanting gonads between subjects presenting different
sexual activities. This idea was taken up again by Brusle who added
193
many technical improvements (1967b) to the methods of transplantation

(1967b, 1970b), and associated them with cultures in vitro of gonads
(Delavault and Brusle, 1969, cf. pp. 242-244).
Fig. 3 a and b. Origin and localization of the gonocytes in adult ani-

mals. (a) Cellular elements (c.) derive from the splanchnopleural sheet
(spl) and leave it (arrows) for the aboral part of the axial organ.
(b) These cells, arrived in the axial organ, have the aspect of gono-
cytes (g.). (Same scale as in a). (By courtesy of Leclerc)
These are the essential points which emerged from the work by Brusle
(1968b) .
Whatever the experimental intervention was, the results are remarkably

homogeneous:
a) Gonads of Asterina gibbosa must be removed between two stages of their

development in order to follow their further growth or the modifications
which they undergo in experimental conditions. If they are removed too
early, they become unable to develop, and if the operation happens too
late they show a development similar to the gonads remaining in situ.
b) If the conditions of gonad extirpation are fulfilled, the gonads

belonging to the first period of sexual activity in functional males
only display a reduced spermatogenetic activity, but they show evidence
of an oogenetic growth (Fig. 4a,b).
In subjects which are functional hermaphrodites, gonads going through

a period of dominating oogenetic activity also show a certain oogenetic
growth.
Considering these results one of the most important questions was to

know whether the cells which perform oogenetic growth after experimen-
tal intervention in functional males derived from undifferentiated gono-
194
cytes which have undergone a sexual orientation diametrically opposed

to their former determination or whether a "waking up" of pre-existing
oogonia that were "asleep" in the gonads has to be assumed.
Fig. 4 a and b. Aspect of gonads from functional males. (a) Control

gonad in situ with active spermatogenesis. (b) Gonad of the same animal
after treatment. Spermatogenetic activity much reduced, advancing oogen-
esis (same scale as a). (From Delavault, Physiology of Echinodermata,
p. 628. New York-London: John Wiley 1966)
The answer to this question is in favor of the second hypothesis.
Many investigations with the electron microscope (Brusle, 1968c,d;

Brusle and Delavault, 1968; Delavault and Brusle, 1968) have led to
the conclusion that there is no germinative epithelium containing a
cellular category similar to undifferentiated gonocytes. Gonads contain
a double stock of germinal cells: spermatogonia and oogonia (Fig. 5a
and b) and the cells of each category are already fixed in their mor-
phogenetic destiny which irreversibly leads to a differentiation into
spermatozoa and eggs. In other words gonads are sexually bipotential,
but with separated potentialities (Brusle, 1970c).
In summary: The development of germinal cells inside the gonads is de-

finitively fixed from the moment they have penetrated into them. Trans-
plantations and culture experiments have only modified the normal
rhythms of sexuality, without any effects on the differentiation char-
acteristic of these cells.
195
Fig. 5 a and b. The two types of germinal cells in the gonads. (a) Oogo-
nium. In the nucleus (N) the chromatin (chr .) is generally rather dif-
fuse, and we often observe two nucleoli (nu .). (b) Spermatogonium.
The chromatin is more densely arranged in clusters and only one nucle-
olus occurs. (By courtesy of Brusle)
4. Sexualization of the Gonocytes
We know that germinal cells have a mesenchymateous origin and reach

the gonads after a transit into the axial organ's aboral part. They
achieve their sexualization during this way.
When we keep in mind that gonads only show spermatogonia and oogonia,
the conclusion is that the cell's sex is fixed outside of the gonads.
Then, of course, a possible role of the axial organ immediately has
to be taken into consideration, since at a certain period it takes up
the mesenchymal cells from which the gonocytes derive.
We have some evidence for supporting this assumptIon by observations

with the electron microscope. Leclerc and Delavault (1972) have dis-
covered the presence of cells which appear very similar to spermato-
gonia and oogonia in the aboral part of the axial organ. Moreover,
Leclerc (1974a) has shown more recently that some tests for alkaline
phosphatase do not contradict the interpretation that these cells be-
long to the germinal lineage. Other authors (Brusle, 1970c; Huet, 1972)
have reported earlier the occurrence of oocytes during the passage
towards the gonads in old animals, or during the gonads' regeneration.
Therefore, we conclude that a "primary" fixation of the sex of the
gonocytes is followed by a "secondary" cytodifferentiation which takes
place in the gonads together with some classical manifestations of
spermatogenesis and oogenesis.
196
II. Potential Hermaphroditism
Potential hermaphroditism has been defined by Brusle (1969c).
Observations with the light and electron microscope have shown that
all gonads of male animals in three gonochoristic Asteroids (Echinastel?
sepositus, Henricia sanguinolenta and Marthastenas glacialis) are in reality
ovotestes.
These perfectly functional testes all have oogonia in a parietal posi-

tion (Fig. 6a) and sometimes even premeiotic oocytes the cytological
characters of which are comparable to those of cells which belong to
the same sex in the hermaphroditic Asterina gibbosa (Brusle, 1968c).
Fig. 6 a and b. Potential hermaphroditism. (a) Male gonad of Echinaster

sepositus with spermatozoids (spz.). The parietal area contains numerous
oogonia (ovog.). (Half-thin section. By courtesy of Brusle). (b) Testis
of Echinaster sepositus with oocytes (ov.) which developed from oogonia
as shown in a). (spg. spermatogenesis; spz. spermatozoids)
The oogonia are more voluminous than the spermatogonia. They contain
one or two nucleoli in very clear nucleoplasm belonging to a large
sized nucleus. They are visible, either at the beginning of spermato-
genetic activity or during full spermatogenesis, or even at its end.
Mitotic divisions of oogonia can also be observed.
The cytological criteria which allow to recognize these oogonia are

approximately the same for each species. Similar investigations have
not been carried out with females of those species.
These observations are of great interest, since they enable us to under-

stand better the occurrence of occasional hermaphroditism in some gono-
choristic species, and they can be compared with what we know about
hermaphrodi tism sensu stricto.
197
In the past, the presence of germinal cells of both sexes in the gonads
of gonochoristic species was regarded as purely accidental. As early
as 1960, however, Cognetti and Delavault (1960b) had noted the compara-
tively high rate (5%) of subjects showing such atypical gonads in Echi-
naster sepositus. This was later reconfirmed (Delavault, 1961) for the
same species (estimated rate 4%) as well as for Marthasterias glacialis
(9% according to Delavault and Cognetti, 1961).
For such reasons we formerly thought that these species seemed to show
a "labile gonochorism" and we placed them between strict gonochorists
and complete hermaphrodites. An attempt to interprete these phenomena
has been made by Cognetti and Delavault (1962).
Now we can approach some aspects of the problem much better. The obser-
vations by Brusle (1969c) show that all male gonads contain cellular
elements of the other sex. They do not develop but under certain cir-
cumstances only which cannot yet be defined. Specimens can be found
in which the testes contain oocytes of various developmental stages
(Fig. 6b).
Consequently "labile gonochori.sm" is a certain expression of a latent

hermaphroditism. The proof for this could be firmly established with
regard to the male sex.
As far as females are concerned no parallel study has been carried out
up to now. But a few hypothetical ideas appear to be meaningful.
The presence of spermatozoa in ovaries has been reported among some

gonochoristic species (cf. particularly Cognetti and Delavault, 1962;
Pl. I, Fig. 4). At first sight there seems to be no reason why sperma-
togonia should not occur in ovaries of such species, but direct cyto-
logical evidence remains to be provided.
With regard to true hermaphrodites one is struck by the histologicat

similarity of hermaphroditic gonads during the functional male phase
and those mentioned in Brusle's descriptions (1969c). In both cases
oogonia and young oocytes whose development has stopped always occur
near to active spermatogenetic territories, and therefore all these
gonads show the characters of potential hermaphroditism.
Hermaphrodi tic specimens of Asterina gibbosa and male individuals of

gonochoristic species differ only by the way in which the two germinal
potentialities are expressed. In hermaphrodites a twofold manifestation
is the rule whereas only one type of sexual potentiality is generally
observed in gonochorists.
The assumption of "intragonadic sexual bipotentiality", as Brus1E~

defines it (1970c), must necessarily be extended to certain gonocho-
ristic species.
The nature of the mechanisms which allow the expression or the non-
expression of this double potentiality remains to be elucidated.
C. Conclusion
At the present stage of our knowledge we can describe a number. of new

informations concerning hermaphroditism in Echinoderms. In Asteroids
it is represented by two cellular types in the gonad, each of them
being strictly determined in its later sexual morphogenesis as male
or female.
198
Depending on whether each of the two categories expresses its own po-
tentialities or not, the hermaphroditism is entirely functional or
simply potential.
The undifferentiated gonocytes from which derive subsequently the ele-

ments of these two cellular categories are found outside of the gonads.
They derive from the mesenchyme and travel through the axial organ be-
fore reaching the gonads.
Consequently the sexualization of the gonocytes is brought about in

two successive steps, and at two different levels. The first one is
the same for potential and functional hermaphrodites. During the mi-
gration of the gonocytes to the gonads their sex is determined, prob-
ably in the region of the axial organ.
The second step resides in the gonads themselves. In functional her-

maphrodites a double gametogenetic cytodifferentiation is taking place,
whereas generally in potential hermaphrodites this cytodifferentiation
appears to be restricted to only one of the two germinal categories.
The mechanisms that participa~e in the determination of sexual poten-

tialities and in their morphogenetic expression remain unknown. How-
ever, the observed sexual variability which is shown by the existence
of sexual races (balanced or not) and the differences in the expression
of cellular potentialities are clear indications of genetical and phys-
iological processes involved in sex determination.
References
Achi tuv, Y.: The genital cycle of Asterina burtoni Gray (Asteroidea) from
the gulf of Elat, Red Sea. Cah. Biol. mar. 14, 547-553 (1973).
Achituv, Y., Delavault, R.: Nouvelles recherches sur l'hermaphrodisme
de Fromia ghardaqana Mrtsn (Echinoderme, Asteride). Cah. Biol. mar.
13, 433-442 (1972).
Bacci, G.: Ricerche su Asterinagibbosa (Penn.). II: L'ermafroditismo in
una popolazione di Plymouth. Arch. zool. ital. li, 49-73 (1949).
Bacci, G.: Alcuni problemi dell'ermafroditismo negli Invertebrati.
Boll. Zool. 17, 193-212 (1950).
Bacci, G.: On two sexual races of Asterina gibbosa (Penn.). Experienta 7...'
31-37 (1951).
Brusle, J.: Recherches complementaires sur la sexualite d'Asterina gib-
bosa P. de Banyuls. Vie et Milieu 18 (1A), 133-141 (1967a).
Brusle, J.: Homogreffes eOt heterogreffes reciproques du tegument et
des gonades chez Asterina gibbosa Pennant et Asterina pancerii Gasco
(Echinodermes, Asterides). Cah. Biol. mar. 8, 417-420 (1967b).
Brusle, J.: Nouvelles recherches sur l' hermaphrodIsme d' Asterina gibbosa
de Roscoff. Cah. Biol. mar. 9, 121-132 (1968a).
Brusle, J.: Recherches sur la sexualite d'un hermaphrodite fonctionnel:
Asterina gibbosa Pennant (Echinoderme, Asteride). These Doct. Etat.
Orsay, 173 p., 71 pl. (1968b).
Brusle, J.: Aspects ultrastucturaux de la differenciation ovogenetique
chez un hermaphrodite fonctionnel, Asterina gibbosa Pennant. Ann. Sci.
nat. Zool. Biol. animale 10, 545-561 (1968c).
Brusle, J.: Aspects ultrastructuraux de la differenciation spermato-
genetique chez un hermaphrodite fonctionnel, Asterina gibbosa Pennant.
Comparaison des deux lignees gametogenetiques. Ann. Sci. nat. Zool.
Biol. animale 10, 563-578 (1968d).
Brusle, J.: Sexuali te d' Asterina gibbosa, Asteride hermaphrodite, des
c6tes de Marseille. Mar. Biol. 1, 276-281 (1969a).
199
BrusH~., J.: Les cycles g€lUi taux d' Asterina gibbosa. Cah. BioI. mar. 10,
271-287 (1969b). -
Brusle, J.: Hermaphrodisme potentiel chez les males de trois especes
d'Asterides. Ann. Ernbryol. Morph. 2, 445-450 (1969c).
Brusle, J.: Ovogenese dominante et sperrnatogenese secondaire resurgente
chez Asterina gibbosa Penn. des cotes de Banyuls. Vie et Milieu 21 (2A),
495-504 (1970a). -
Brusle, J.: Auto-, homo- et heterotransplantations de gonades chez
Asterinagibbosa P. Arch. Zool. exper. gen. 111, 159-180 (1970b).
Brusle, J.: Les potentialites germinales intragonadiques d'Asterina
gibbosa P. Cah. BioI. mar. 11, 35-42 (1970c).
Brusle, J., Delavault, R.: Recherches sur la cytodifferenciation des
gametes chez un herrnaphrodi te fonctionnel: Asterina gibbosa. Ul tra-
structure des ovogonies et des ovocytes en premeiose. C.R. Acad.
Sci. Paris 266, 21-23 (1968).
Cognetti, G.: Autofecondazione in Asterina. Boll. Zool. 1], 275-278
(1956) •
Cognetti, G.: La spermatogenesi secondaria in Asterina e la colorazione
degli individui proteroginici di Asterina gibbosa. Accad. Nazion.
Lincei, Rend. Class. Sc. Fis. Mat. Nat. 24, 325-327 (1958).
Cognetti, G., Delavault, R.: Gonocorismo stabile, gonocorismo labile
ed ermafroditismo in Asteroidi dell' Atlantico e del Mediterraneo.
Accad. Na'zion. Lincei, Rend. Class. Sc. Fis. Mat. Nat. 28 (ser. 8),
82-85 (1960a). -
Cognetti, G., Delavault, R.: Recherches sur la sexualite d'Echinaster
sepositus (Echinoderme, Asteride). Etude des glandes genitales chez
les animaux des cotes de Livourne. Cah. BioI. mar. 1, 421-432
(1960b). -
Cognetti, G., Delavault, R.: La sexualite des Asterides. Cah. BioI.
mar. 3, 157-182 (1962).
Delavault, R.: Les cycles g€lUitaux chez Asterina gibbosa de Dinard. C.R.
Acad. Sci., Paris 251, 2240-2241 (1960).
Delavault, R.: La sexualite chez Echinaster sepositus Gray du Golfe de
Naples. Pubbl. Staz. zool. Napoli 32, 41-57 (1961).
Delavault, R.: Recherches experimentales sur la sexualite des herm-
aphrodi tes chez Asterina gibbosa: Greffes de glandes geni tales. Arch.
Anat. micro Morphol. exper. 52, 469-496 (1963).
Delavault, R.: Determinism of Sex. In: Physiology of Echinodermata,
822 p. New York-London: John Wiley 1966.
Delavault, R., Brusle, J.: Recherches sur la cytodifferenciation des
gametes chez un hermaphrodite fonctionnel: Asterina gibbosa. Ul tra-
structure des cellules de la lignee sperrnatogenetique et comparaison
spermatogonies-ovogonies. C.R. Acad. Sci. Paris 266, 710-712 (1968).
Delavault, R., Brusle, J.: Analyse experimentale de la sexualite
d' Asterina gibbosa en particulier par les cultures org anotypiques •
In: Cultures d'organes d'Invertebres (publ. sis direct. H. Lutz),
265 p. Paris-London-New York: Gordon and Breach 1969.
Delavault, R., Brusle, J., Pierre, G.: La sexualij::e d'Asterinapancerii
Gasco (Echinoderme, Asteride) des Golfes de Naples et de Marseille.
Pubbl. Staz. zool. Napoli 37, 140-155 (1969).
Delavault, R., Cognetti, G.: L'hermaphrodisme chez Echinaster sepositus
Gray du Golfe du Naples. C.R. Acad. Sci. Paris 245, 2545-2547 (1957).
Delavault, R., Cognetti, G.: L'hermaphrodisme chez Marthasterias glaciatis
L. C.R. Acad. Sci. Paris 253, 2771-2772 (1961).
Delavault, R., Leclerc, M.: Le complexe axial d' Asterina gibbosa (Echino-
derme, Asteride) et ses rapports avec Ie complexe genital. Bull.
Soc. zool. Fr. ~, 603-611 (1967).
Hauenschild, C.: Zur Frage der Geschlechtsbestirnrnung bei Asterina gibbosa.
Zool. Jahrb. 65, 43-53 (1954).
Huet, M.: Action~es rayons X sur la lignee germinale et la regenera-
tion de I' appareil genital d' Asterina gibbosa Penn. (Echinoderme).
C.R. Acad. Sci. Paris 260, 707-709 (1965).
200
Huet, M.: Etude experimentale de la regeneration du-bras et de l'appa-

reil genital d'Asterinagibbosa Penn. These Doct. Etat. Orsay, 79 p.,
30 pI. (1972).
Hyman, L.H.: The Invertebrata. Vol. IV, Echinodermata, 763 pp. New York-
Toronto-London: McGraw Hill 1955.
Komatsu, M., Oguro, C.: Notes on the hermaphroditic specimen of the
Sea-star, Certonardoa semireguZaris (MUller et Troschel). Proc. Jap.
Soc. System. Zool. no. 8, 49-52 (1972).
Leclerc, M.: L'organe axial et ses relations avec la sexualite et
l'immunite chez les Asterides. These Doct. Etat. Orleans, 74 p.,
18 pl. (1974). Publ. dans: Ann. Sci. nat. Zool. BioI. anima Ie 16,
285-360 (1974a). -
Leclerc, M.: Proliferation naturelle et experimentale de cellules ger-
minales chez une Asteride: Asterina gibbosa (Echinodermes). Ann. Sci.
nat. Zool. BioI. animale (1974b, in press) .
Leclerc, M., Delavault, R.: Morphologie et ultrastructure du complexe
axial d' Asterina gibbosa (Echinodermes, Asterides), etudiees en micro-
scopie electronique. Ann. Sci. nat. Zool. Biol. animale 14, 193-202
(1972). -
Lender, Th., Delavault, R.: La lignee germinale chez les Echinodermes.
In: L'origine de la lignee germinale (publ. sis direct. E. Wolff),
370 p. Paris: Hermann 1964.-
Lender, Th., _ Huet, M.: Sur la regeneration des gonades d 'Asterina gibbosa
(Penn.). C.R. Acad. Sci. Paris 254, 2447-2449 (1962a).
Lender, Th., Huet, M.: Etude de la regeneration de l'appareil genital
d' une Asteride: Asterina gibbosa (Penn.). Bull. Soc. zool. Fr. 87,
191-196 (1962b).
Neefs, Y.: Developpement et evolution sexuelle chez Asterina gibbosa.
Proc. 15th Int. Congr. Zool. London, pp. 286-288 (1959).
The Gonadal and Adenohypophysial Functions of Natural
Sex Reversal *
S. T. H. Chan, Wai-sum 0, and S. W. B. Hui
A. Introduction
Hermaphroditism is not uncommon among the lower vertebrates, especially

in teleosts which exhibit the most diversified expressions of sex, with
all types of intersexuality and hermaphroditism (Atz, 1964). Natural
sex reversal is generally referred to as hermaphroditism functional in
time, i.e. an organism functions both as female and as male in its
life history, but at anyone time it would function only as one sex.
An intersexual stage invariably occurs during transformation and is
always transitional (Chan, 1970). In view of the fact that in many
protogynous forms (including Monopterus) the presumptive male germ cells
pre-exist before the onset of change, natural sex reversal, or sex
inversion (Reinboth, 1970), is in reality a natural biological process
of sequential maturation of the female and male gonadal sex in succes-
sion, relating to the gonadal ontogeny and life history (Chan, 1970,
Chan et al., 1972).
Despite being an unique biological phenomenon, natural sex reversal

(protandry and protogyny) in teleosts aroused little interest among
endocrinologists until the last decade, and many problems concerning
the physiological functions of the dynamic gonad remain unanswered.
For instance, it is not known how the gonad of an organism can produce
mature germ cells of the two opposite sexes in succession. In the two
successive different sexual phases, it is also not understood what
mechanism exists in regulating the functional aspects of the two nor-
mally antagonistic groups of sex steroid hormones, and how the produc-
tion of only the appropriate kind of hormone can be ensured. The ques-
tion of whether the process of natural sex reversal at gonadal level
is accompanied, or caused by changes of adenohypophysial or hypothal-
amic nature in their secretions or in their feedback threshold sensi-
tivity or in the intrinsic mechanism of priming the CNS-hypothalamo-
hypophysis axis also remains unresolved. These questions, with due
inference to the existence of an integrated hypothalamo-hypophysio-
gonadal axis in most vertebrates, bring forth a concept of broader
dimensions for natural sex reversal - projecting it as a series of
integrated events in the gonad and adenohypophysis irrespective to
whether the initial trigger arises within or without this axis. Taking
into consideration the current concepts on reproductive physiology in
vertebrates, the present report attempts to correlate some aspects of
the gonadal and adenohypophysial forms and functions of spontaneous
sex reversal in fishes, with special reference to Monopterus.
*Supported by a Research Grand on Sex Determination from the Nuffield

Foundation, London
202
B. The Gonad
The gonad of all vertebrates comprises two basic components, the germ
cells and the somatic tissue. Differentiation of the gonadal sex is
controlled by the naturally destined and genetically predetermined
development of these two components. Witschi's observations on the
existence of embryonic medulla and cortex in the primordial gonad and
his conceptual ideas of hypothetical inductor theory for amphibians
(Witschi, 1957) has markedly influenced subsequent studies in other
vertebrate groups; the two embryonic soma appear to be invariably
present in all amniotes (Burns, 1961) and in elasmobranchs (Chieffi,
1959). In cyclostomes and teleosts, however, such separation of embry-
onic somatic tissues is claimed to be absent and that in the salmonids,
cyprinids and cyprinodonts, the whole genital ridge is reported to
derive from one origin, believed to be homologous to the gonadal cortex
of amphibians and amniotes (D'Ancona, 1949). Such an assumption has led
to the suggestion by some workers that this is the cause for the rela-
tive high frequency of hermaphroditism in bony fish (Dodd, 1960). How-
ever, D'Ancona (1949) postulated a "gynogenine and androgenine system"
for the serranids and sparids - where a separate localisation of the
male. and female germinal tissue exists and hermaphroditism and sex re-
versal prevail. Whether for most teleosts there exists a dual induction
system in a common gonadal substratum or a gonadal soma with dual em-
bryonic origin awaits further elucidation, especially when our existing
knowledge on this area includes information from only a few species.
I. Structural Basis for Succession of Sexes?
The concept of "sexuality succession", or sequential expression of the

female and male sex in succession, in relation to life cycle (Chan,
1970), in defining the process of natural sex reversal in Monopterus,
is based on the observation that the potential male germ cells exist
along the inner edge of the gonadal lamellae before the onset of
changes (Chan and Phillips, 1967a). There is at present no direct
confirmative evidence on tracing the development of these male gonia
to spermatogenetic cells. However, observations on biopsy studies
(Chan et al., 1972) and on specimens with lesser degrees of spermato-
genetic activities provide strong affirmative indication. Thus, for
certain hermaphrodites at least, the presence of heterologous germinal
tissues even at the mature state of one sex, appears to .provide a
structural basis for sex reversal. In the protandric sparids Sparus,
Mylio and PageUus, and the flatheads, CocieUa and Rogadius, preformed
and non-functional ovarian tissue exists early in the male phase.
Similarly, in some protogynous species, e. g. Sacura, Spondyliosoma and
Coris, potential testicular tissues are performed in the ovary (Rein-
both, 1962a, 1970). However, in the protandric Gonostoma and Cobitis,
and protogynous Thalassoma, the potential female and male cells respec-
tively, have not yet been identified before sex reversal (Reinboth,
1970) .
Further work is needed to clarify various problems in this area, espe-

cially for the close interrelationship between the germ cells and the
somatic tissue in their development and function. It will be inter-
esting to know whether two types of gonia co-exist in a common somatic
enviroruaent, each with its own predetermined epigenesis; or whether
only one type of bipotential gonia exists whose differentiation would
be influenced entirely by the differential induction capacity of the
somatic substratum - originating either from a dual embryonic origin
or from the regional or zonal orientation of somatic tissue of common
origin. It could also be that some bisexual mechanisms operate in both
the gonia and the soma.
203
somatic tissue of common origin. It could also, however, indicate bi-

sexual mechanisms operate in both the gonia and soma.
It is up to the cytogeneticists and the embryologists to elucidate

these aspects of the problem and provide some information of funda-
mental importance to our understanding of sex reversal.
II. Sex Steroid Hormone Sites and Productions
Evidence from histo-enzymology studies in teleosts shows that the

theca or the granulosa of maturing follicles are the main steroid tis-
sue in the ovary (Lambert, 1970; Yaron, 1971; Bara, 1965); while in
the testes, the interstitial Leydig cells are the major sites of ste-
roidogenesis in a number of teleosts (Stanley et al., 1965; Yaron,
1966; Yamazaki and Donaldson, 1969; Bara, 1969; Wiebe, 1969). However,
little is known about the steroid tissues in hermaphrodites. Following
earlier histological studies on the possible endocrine sites in Monop-
terus (Chan and Phillips, 1967b; Chan et al., 1967), histo-enzymological
evidence (using standard techniques for 3S-HSD and 17S-HSD activities)
is now available that the interstitial cells in the male and the granu-
losa cells in the female are the major steroidogenic sites (Tang et al.,
1974a,b). Hence, the steroidogenic sites in Monopterus female and male
gonads are similar to those in the ovary and testis of other gonochor-
istic fishes. The marked increase in steroid enzyme activities observed
in parallel to the extensive development of interstitial Leydig tissue
during sex change indicates a rather unique feature in this species.
This observation, together with the scarcity of 3S-HSD positive cells
in the gonadal interstitium at the female phase, suggest that inter-
stitial cell development might precede testicular lobule formation and
that a change in the steroid endocrine environment in the gonad might
be a possible cause of natural sex reversal in Monopterus (Chan and
Phillips, 1969; Tang et al., 1974b).
Despite the volumninous literature on steroid production by the gonads

of gonochoristic teleosts (Gottfried, 1964; Lofts, 1968; Barr, 1968),
similar work on hermaphroditic species is still fragmentary. The ear-
liest attempts were on Serranus scriba, where Lupo and Chieffi (1965)
reported the isolation of testosterone, androstenedione, andrenosterone,
estrone, estradiol-17S, estriol and progesterone; and on Centropristes
striatus, in which Reinboth et al. (1966) reported the incubations of
progesterone- 14 C and testosterone-1 4 C with ovarian tissue in vitro
producing only reductive metabolites. Chan and Phillips (1969) studied
the in vitro steroidogenesi.s of Monopterus gonads at various phases and
identified, among other metabolites, progesterone, androstenedione,
testosterone, estradiol-17S and estrone, the latter two being in trace
amounts (except in large females). Based on the ratio of estrogen and
androgen production in various sexual phases, the~ concluded that a
shift to "hormonal maleness" occurred during natural sex reversal.
On protandric species, Colombo et al. (1972) reported that androstene-
dione, testosterone, estradiol-17S and estrone was formed in both the
mature testis and mature ovary of Sparus (with greater estrogen forma-
tion in the latter) and that this protandrous hermaphrodite was similar
to the protogynous Monopterus in the types of steroids produced. During
sex reversal in Sparus, the testicular part produces only progesterone,
17u-hydroxyprogesterone and androstenedione while the ovarian part
yields only progesterone and 17u-hydroxyprogesterone (Colombo et al.,
1972). More recently, Reinboth (1974) reported that in the protandrous
PageUus, progesterone is metabolized more rapidly by ovarian tissue
in vitro than by testicular tissue of functional males or inverting
specimens. Reinboth (1974) found that the number of metabolites in
the testicular part of inverting specimens gave no indication for a
204
decrease in the enzyme activities, contrasting the data in Sparus

(Colombo et al., 1972). It appears that, as far as steroid hormone
production is concerned, most evidence now available indicates that
the two mature sexual phases of the gonad in a sex-reversing species
are similar to the ovary and testis of gonochoristic species. Thus,
in sex-reversing forms, the functional endocrine system in the male
and female phase is in concord with the maturing germinal tissues of
the corresponding sex. This is not surprising, since as far asMonopterus
is concerned, the major anatomical units for steroidogenesis in the
two sexual phases differ and seldom co-exist.
Owing to some obvious limitations in the in vitro studies, the in vivo

approach is always preferred because this will provide more reliable
information on the actual secretory pattern of hormones. Following
the technique of Exley (1968), investigation is being carried out in
this laboratory to measure the plasma level of testosterone and estra-
diol-17f', in various sexual phases of Monopterus, taking into account also
the seasonal cyclical changes of reproductive activities. Results on
the postnuptial specimens are shown in Fig. 1. It is clear that the
in vivo pattern is similar to the in vitro pattern of hormonal shift.
The testosterone levels in both the early and mid intersexes are sig-
nificantly higher than that in the male (P < 0.05). This may be either
the true pattern or is more likely related to the fact that the male
testis is relatively inactive in postnuptial period whereas in the
same season, the intersexual gonad is always actively engaged in
transformation activities. On the other hand, the plasma level of
estradiol-17f', is high in female but very low in male, the mean value
of the latter is just above the sensitivity limit of the technique for
estradiol-17f', (2ng/ml plasma); estrone cannot be accurately measured
by this technique.
100
(Postnuptial samples I 0---0 Estradiol -1713
e - - e Testosterone
..§ 80
01
C
'>"
~ 60
"0
c
o
E
~
o 40
.&;
CI
E
III
E
0... 20
O-L--------------------------------------
Early Mid Late
Fmale Male
Intersex
Fig. 1. Plasma levels of testosterone and estradiol-17f', in Monopterus

at various sexual phases
More will be known on the in vivo steroid pattern when our current
research reaches its completion.
205
1.11. Some Aspects of the Gonadal Functions
Existing knowledge concerning the gonadal functions in lower verte-

brates is confined mainly to gonochoristic species (Pickford and Atz,
1957; Barr, 1968; Lofts, 1968); little is known about hermaphrodites.
As the two cardinal concerns of the vertebrate gonads are production
of mature gametes and secretion of sex hormones, the hermaphroditic
fishes should not show any exceptions. However, unlike most vertebrates
which exhibit a clear dichotomy of male and female sexes, synchronous
and protogynous or protandrous hermaphrodites confront with the most
challenging problem of performing the functions of two opposite sexes
either simultaneously or in succession by the gonad of an individual
organism - a condition rarely possible in the higher forms.
Information concerning the anatomical aspect of gamete maturation in

hermaphroditic fishes is comparatively abundant (Atz, 1964; Reinboth,
1970). Most studies suggest that these gonads function in this aspect
as do the gonochorists, e.g. in Monopterus, the oocytes and their ma-
turation in the female phase are identical to those of gonochoristic
fishes (Chan et al., 1967), and in the male, the cystic spermatogenesis
(Chan and Phillips, 1967b) conforms to the typical pattern of fish
testes (Lofts, 1968). Studies on hormonal secretions (as already dis-
cussed) also· indicate that endocrinologically, the sex-reversing gonad
functions successively as the ovary and testis of gonochorists. Hence,
as far as the fundamental functions are concerned, Monopterus exercises
its gonad in two separate, independent entities, sequentially spaced
and no mature gametes of both sexes occur concurrently. Sexuality suc-
cession thus provides a system as efficient as gonochorism in prevent-
ing self-fertilization (Chan, 1970).
As the hormonal type is normally harmonious with the gonadal sex (even
in sex-reversing species), one major aspect of the gonadal function,
therefore, concerns the possible facilitation by hormonal secretions
on germinal tissues of the corresponding sex. The exact functional
role of the major sex steroids found in the sex-reversing heramphro-
dites is little known to date. In vertebrates, there exists a close
interrelationship between the gonia and the associated endocrine tis-
sue and, especially in females, they are indispensably interdependent
in their normal function. It follows that one probable function of
the endocrine secretions and their changes during sex reversal, may
relate to germ cells development - possibly providing a compatible
environment facilitating their maturation and function. Thus, the
estrogen at the female phase in Monopterus might be involved in some
interaction with the maturing oocyte - for it is maintained that in
some vertebrates at least, the granulosa cell secretion exerts effects
on both meiotic and pre-ovulatory changes in the oocyte (Foote and
Thibault, 1969). In addition, action of estrogen on ovarian follicles
can be indirect, via its influence on the hypothalamo-hypophysial axis,
for among higher vertebrates increase in estrogen~itres triggers the
LH surge - resulting in ovulation and luteinization (Greep, 1973),
although at present, little is known about the endocrine mechanism
involved in ovulation in lower vertebrates (Barr, 1968) and in sex-
reversing fishes. A further aspect concerning estrogen function in
facilitating oocyte development can be derived from its action in
poikilotherms and birds, concerning calcium mobilization and vitello-
genesis. The seasonal patterns of plasma calcium levels in Monopterus
(Fig. 2) clearly show comparatively higher levels in the female and
in the prenuptial period, correlating with the higher titre of estro-
gen in females and the period of active vitellogenesis. This, together
with the fact that estrogen treatment in females only causes hyper-
calcaemia (0 and Chan, 1974c), suggests a possible role of estrogen
on vitellogenesis. The experimental finding that cyanoketone adminis-
206
tration (hence blocking all steroidogenesis) to females, results in a

massive degeneration of the maturing follicles (Tang et al., 1974c)
also lends support to the belief that estrogen in Monopterus maintains
the normal maturation of female germ cells .
5.0 • __e male

0--0 female
o intersex
4. 0
...I
.......
~
E
c
3.0
o
E 2.0
'"o
ii:
F A J J A S 0 o J F
( Month)
Fig. 2. Seasonal levels of plasma calcium in Monopterus at various sexual
phases
Due to their capacity in influencing gonadal differentiation during

undifferentiated stages in amphibians and teleosts, resulting in 8X-
perimental functional sex inversion in gonochorists (Yamamoto, 1958;
Burns, 1961), sex steroids have been considered as one of the probable
causatives of spontaneous sex reversal - mainly by their general na-
ture in being antagonistic to tissues of the opposite sex and, as
discussed above, an enhancement to those of the same sex. Such aspects
of the hormonal differential effects on heterologous germinal tissues
and the probable causative role of androgen in Monopterus have been
extensively investigated by administrations of hormones or drugs (Chan
et al., 1972b: 0, 1973; Tang et al., 1974c,d) coupled with studies by
the biopsy techniques established by Chan et al. (1972a). In general,
administration (at different dosages) of various androgens, including
testosterone, methyl-testosterone and 11-keto-testosterone, fail to
bring about the precocious development of the male gonia in the female
phase. This differs from some other protogynous species, Coris, Thalassoma
and Halichoeres, where Reinboth (1962b, 1963 a,b) reported some success
in causing precocious sex reversal. Implantation of testosterone to
male and mid-intersexes, however, enhances spermatogenetic activities
(Fig. 3 a-d). This is in agreement with the spermatokinetic effects
of androgen in other gonochoristic species (Pickford and Atz, 1957;
Lofts, 1968). Estrone implantation to female causes no obvious changes
(except some with increase of small oocytes), but to the male and inter-
sex results in marked destruction of the interstitial Leydig tissue
and spermatogenetic cells, leaving behind only spermatogonia in col-
lapsed lobules (Fig. 3a, b, e, f). Thus, in Monopterus, estrogen appears
to be antagonistic to the male gonia - suppressing their spermatogenetic
acitivites.
207
On the contrary, effects of androgen on the cryptic male gonia and on

ovarian follicles in female Monopterus seem insignificant. This failure
of androgen in bringing about precocious sex reversal indicates that
the endogenous androgen increase might not be a primary causative fac-
tor of sex reversal, but rather as a secondary event responding to
other triggering factor(s). In addition, it has been demonstrated by
use of cyanoketone (with or without androgen) that the failure of exo-
genous androgens is not related to any antagonistic effect against them
by endogenous estrogen (Tang et al., 1974c). This, together with the
fact that at a later ontogenetic stage in mid-intersex the male germ
cells become responsive to androgen stimulation, clearly indicates a
possible "age-dependent" nature of the male gonia in their epigenesis,
and/or in their sex-specific receptor system.
C. The Adenohypophysis
The anterior pituitary has been reported to play no direct role on the
differentiation of gonadal sex in vertebrates (Burns, 1961). This is
shown in hypophysectomized embryos of amphibians (Puckett, 1940);
chicks (van Deth et al., 1956); and mammals (Jost, 1970). Though the
pituitary apparently exerts no decisive influence on primary gonado-
genesis, its presence is essential for the germ-cell maturation and
normal endocrine function of the gonad in both sexes. Hence, the role
that the pituitary could play in the successive maturation of the fe-
male and male germ cells in sex-reversing fishes, such as Monopterus,
cannot be overlooked, especially when the structural sex change is
accompanied by ,extensive interstitial Leydig development (Chan and
Phillips, 1967a). Such tissue is known to be controlled by adenohypo-
physial secretions in vertebrates (Gabrilove, 1973).
I. Anatomical Structure
Despite the accumulating information on the gonad and sex steroids in

various hermaphroditic fishes, little is known about their pituitary
glands. 0 and Chan (1974a) report a cytological investigation on the
pituitary structure and cell types in the protogynous hermaphrodite,
Monopterus, and find that it essentially comprises a highly interdigi-
tating neurohypophysial core, ramifying extensively into the other
component, the adenohypophysis. By a series of standard ~taining tech-
niques and various experimental verifications on their physiological
functions, seven adenohypophysial cell types have been identified as
being related to various adenohypophysial secretions (0 and Chan,
1974b,c; Chan et al., 1975).
In general, the structural organization of the pituitary in Monopterus

conforms to the typical anatomical patterns found in those advanced
teleosts so far studied, except that the Monopterus adenohypophysis has
a characteristic feature - in that each type of adenohypophysial cell
is not confined to one distinct zone but ocvurs, often as symmetrically ,
as small cell groups; the three basophils intermingle and are separated
from the neurohypophysis by acidophils and amphiphils. Hence, the gland
exhibits great variations in cell type distribution, although the over-
all pattern remains more or less unchanged.
II. The Gonadotrops

Among the various adenohypophysial cells, the basophils (because of
their glycoprotein content) deserve special attention with regard to
208
Fig. 3 a-f. Effects of steroid hormone implantations on Monopterus gonads

showing, after 60 days, the control intersex (a) and control male (b),
the 5 mg-testosterone-implanted intersex (cl and male (d) both with
enhanced spermatogenetic activities, and the 2 mg-estrone-implanted
209
intersex (e) and male (f) both with spermatogenesis suppressed, leav-
ing only resting spermatogonia in collapsed lobules and severely af-
fected interstitial cells
210
gonadotropic functions. In Monopterus, 0 and Chan (1974a) report three

types of basophils, the type 5, 6 and 7 cells, which exhibits clear
differences in cell sizes, nuclear sizes, shapes and tinctorial char-
acteristics. By experiments with radioactive iodine, thiourea, thyrox-
ine and TSH, it has been demonstrated that the type 5 cells are con-
cerned mainly with thyrotropic function (0 and Chan, 1974b). On the
other hand, using injection experiments involving testosterone, estrone,
mammalian LH and FSH, it is shown that the remaining two basophils are
related to gonadotropic functions, and that these two basophils appear,
cytophysiologically, to be two types of gonadotrops (0 and Chan, 1974c);
a result similar to those found in the European eel and in salmonids
(Olivereau and Herlant, 1960; Olivereau and Ridgeway, 1962; Olivereau,
1963, 1967; van Overbeeke and McBride, 1967). Moreover, in treatment
experiments the two gonadotropic cells exhibit sex-related differential
responses to the different sex steroid hormones (0 and Chan, 1974c);
this observation, like the condition found in male gonia responding to
androgen and in plasma calcium level reacting to estrogen, leads to
the speculation on the existence of "sex specifity", or the presence
of different sex-specific receptors in tissue of MonopteY'us at different
sexual stages. This aspect of the problem awaits further elucidation.
On the experimental evidence available, 0 and Chan (1974c) speculate
that these two gonadotropic cells probably secrete different types of
secretions, the type 6 cells a "LH-like" hormone, while the type 7
cells a "FSH-like" hormone (see later section) .
III. Gonadotropic Functions of the Adenohypophysis
In all vertebrates, the gonad in both sexes is under the control of

the pituitary, which secrete specific hormones - the gonadotropins.
Experimental evidence in mammals has established the existence of two
types of gonadotropins (Fevold et al., 1931): FSH and LH. Also the
patterns of gonadotropin secretion in the male and female are differ-
ent in rat and in some other vertebrates (Harris, 1970; Faiman and
Winter, 1971; Johnson, 1972). Whether any differences in gonadotropin(s)
secretion exist in teleosts between the two sexes, or in the two sexual
phases of a sex-reversing hermaphroditic species, such as MonopteY'Us,
ramains a subject awaiting investigations. However, before answering
this question, two major problems in studies of teleost gonadotropins
have to be considered: first, the presence of one or two types of go-
nadotropin, and secondly, the functions of the gonadotropin(s)in
teleosts.
The role of gonadotropin in teleosts has been demonstrated by experi-

mental approaches, such as hypophysectomy followed by replacement
therapy. In the male, the primary role concerns steroidogenesis, while
the androgen elaborated is responsible for the subsequent spermato-
genic effects (Sundararaj et al., 1971, 1972). In the female, there
is no unanimity of opinion regarding gonadotropin for oogenesis, vi-
tellogenesis and ovulation (Barr, 1968). The partially purified salmon
gonadotropin (Donaldson et al., 1972) has been found to be effective
in re-initiating oogenesis and vitellogenesis, maintaining yolky 00-
cytes and in inducing ovulation and spawning (Yamazaki and Donaldson,
1968b; Sundararaj et al., 1972). Thus, the "single" salmon gonadotro-
pin extract appears to be responsible for both vitellogenesis and ovu-
lation. In addition, Yamazaki and Donaldson (1968a) and Sundararaj et
al. (1972) suggest the presence of a single gonadotropin in salmons.
This disagrees with previous suggestions for 2 gonadotropins in tele-
osts (Otsuka, 1956; Ball, 1960; Barr, 1968; Yamazaki, 1969).
Despite the recent advances in total purification and characterization

of the two gonadotropins in mammals (Li, 1973), pioneering work on the
211
isolation of gonadotropins in fishes has been carried out only in two

fishes, Cyprinus carpio (Burzawa-Gerard, 1971) and Oncorhynchus tschCOilytscha
(Donaldson et al., 1972) and their evidence points to the existence of
only one gonadotropin. However, in these biochemical studies different
extraction, purification and bioassay methods were employed; such dif-
ferent methods have not been considered as totally satisfactory by
some workers (Reinboth, 1972). It has also been pointed out by Pick-
ford and Atz (1957) and Barr (1968) that it is difficult to assess the
different end-points used, e.g. ovulation, spermiation, increase in
gonad weight - and further complications may arise from variations
with the differences in techniques. Therefore, at this stage it appears
premature to draw any generalized conclusion on the existence of one
or two types of gonadotropic secretion in such a diversified group as
the Teleostei, and further investigations on more species seem neces-
sary.
Preliminary studies on the possible gonadotropic secretions in Mo nopterus

have been made in our laboratory. The extensive development of inter-
stitial tissue during sex change clearly indicate a possible direct
involvement of the pituitary secretion(s) in the process. In addition,
the fact that two gonadotropic cells are distinguished in the adeno-
hypophysis (0 and Chan, 1974a) and that the female gonad responds to
treatments of mammalian LH and FSH (see Fig. 4, 0, 1973) resulting in
some structural sex changes, provide additional evidence. It should be
mentioned, however, that mammalian-LH causes extensive precocious de-
velopment of interstitial Leydig cells in all experimental animals.
Fig. 4. The exten-

sive development
of Leydig cells
and some spermato-
genetic activities
in the distal re-
gion of gonadal
lamellae of ovine-
LH treated female
Monopterus (0. 1 IU /
fish every 12 hour-
ly for 10 days)
resemble to a cer-
tain degree the
natural sex reversal
However, this interstitial tissue development in the LH-induced sex

reversal is not matched with comparable magnitude by the male-gonia
proliferation and differentiation though some spermatogenetic activi-
ties are observed especially in animals with more frequent treatment
(Fig. 4). Whether this unresponsiveness is due to the lack of appro-
priate specificity of the exogenous hormone, to the absence of addition-
al stimulatory factor(s), or to the intrinsic age-dependent and sex-
specific mechanism of the male gonia in their development and/or in
their responsiveness to stimulation, is at present uncertain. Reinboth
(1962a; see Atz, 1964) has tentatively suggested that increase in gona-
dotropin in growing fish is responsible for the eventual sex reversal
in protandrous species; experiments with injection of fish pituitary
212
homogenates, however, do not successfully stimulate" Serrar!us (Reinboth

and Simon, 1963; quoted by Atz, 1964). This result, contrasting to the
condition in Monopterus, could be due to various factors including spe-
cies variations.
As an initial direct investigation on Monopterus pituitary function,

the existence of any gonadotropic activities in the gland is best
measured by bioassays, the choice of which remains a major problem.
With due references to earlier discussion on studies of fish gonado-
tropin(s), and also to similar work in mammals (Rosenberg, 1968) the
two bioassay techniques chosen for Monopterus are the rat ovary ascorbic
acid depletion test (Parlow, 1961) for "LH-activity" and the rat ovar-
ian augmentation test (Steelman and Pohley, 1953) for "FSH-activity",
both being simple, sensitive and uninfluenced by other tropic hormones
(Parlow, 1961; Christiansen, 1972). Despite reports on the phylogenetic
specificity regarding the biological activities of salmon gonadotropin
(Yamazaki and Donaldson, 1968a,b; Channing et al., 1974), the choice
of bioassay with mammalian tissue for gonadotropic activities in Mon-
opterus was based on the fact that with the difficulties in assessing
different end-points used in different assays (Pickford and Atz, 1957;
Barr, 1968), the use of some reliable and standardized techniques
seemed to be essential as a start for comparative evidence in future
applications of other bioassays.
Special precautions are taken in collecting Monopterus pituitaries for

gonadotropic activities. Fish are collected in the reproductively ac-
tive season (May) only, and mature, healthy and living specimens are
used; pituitaries are collected in the shortest time (in all cases no
more than one minute after decapitation)J while the isolated gland is
in chilled redistilled acetone and immediately stored in deep-freezer
(- 20°C) until use. For studies related to sex differences, microscopic
and histological studies on the gonad are always used and the individ-
ually stored pituitaries are then pooled according to the sexual status.
Standard gonadotropins (ovine, NIH-LH-S 18 and NIH-FSH-L 9 ) 1 are used as
references.
Assays for gonadotropic activities in the 0.9% saline total pituitary

homogenate, irrespective to sex, show the presence of "LH activity"
and "FSH activity" in equivalent to 0.081 Ilg NIH-LH-S18 per pituitary
(0.506 Ilg/mg pituitary) and 0.67 Ilg NIH-FSH-S9 per pituitary (4.277 Ilg/
mg) respectively (Fig. 5). In addition to the OAAD test, the "LH-activ-
ity" in Monopterus pituitary homogenate produces also statistically
significant responses in tne biological tests involving: (a) the ~n vivo
stimulation of plasma testosterone level on mature male rats; (b) the in-
crease of conversion of 3H-cholesterol to progesterone by rabbit corpora
lutea tissue-homogenate in vitro; and (c) the increase of prostate weight
in immature male rats (Chan et al., unpublished results). Preliminary
investigations on the difference of gonadotropin activities in pitu-
itaries of different sexual phases have also be made, using crude go-
nadotropic fractions extracted according to Otsuka (1956) with modifi-
cations (Saxena and Rathnam, 1968).
Results are summarized in Table 1. The gonadotropic activities (after

compensating for loss) in Monopterus seems to be different in fish at
various sexual statuses. While the "FSH activity" appears to be more
or less similar in all three sexual stages, with the male and intersex
(mainly late intersexes in May) slightly higher than the female, the
"LH-activlty" is found to be much higher in the female than the inter-
sex and male - whose "LH-activities" are too low to be measured with
IFree gifts from the Endocrinology Study Section, N.I.H.

213
500
a
a 100 200 300 '00
NIH - F5H -59 In "'9
100
C
I-
z 1:
lLI
I-
£"
90
n T
OJ
80
~T
Z
~
0
u
I~~~A
c:
11\ 70
0
11\
U >
0
<l
60
01
['1
!::! E
II)
a: 00 50 B
0
....01
"I
u
VI
<l .3.. 40
0·' 0·5 10 5-0 10·0
NI H-LH-S I8 In log fig
Fig. 5_ Dose-response curves and bioassays for gonadotropic activities

in Monopterus pituitary using the HCG augmentation test for FSH (upper)
and OAAD test for LH (lower). (C contro l; A 28.75 pit./100 gm rat;
B 57.5 pit./100 gm rat)
the OAAD test .. Two points should be mentioned. First, the presence of
both "LH" and "FSH" activities in Monopterus pi tui tary does not refute
the view of single gonadotropin in fishes, because a single gonadotro-
pin may exhibit both activities in the test system. However, prelimi-
nary data on the gonadotropic activities in Monopterus at various sexual
214
Table 1. Gonadotropic activities in pituitary of MOYlOpterus at various

sexual phases
Activities in equivalent of
NIH-FSH- S 9 NIH-LH-S 18
()lg/gm-pit) ()lg/pit) ()lg/gm-pit) ()lg/pit)
Homogenate total-pitiC 4.277 0.671 0.506 0.081

Female pit-extract a 2.568 0.395 3.575 0.550
Intersex pit-extract b 2.854 0.495 0.342 0.059
Male pit-extract C 3.885 0.500 0.023 0.0035
* All sexes b 133 intersexual fish

a 104 female fish c 113 male fish
phases indicate a high "FSH" and low "LH" per pituitary in male and
a high "LH" and low "FSH" in female. Hence, this difference found in
the same species at different sexual stages could only be best explained
by assuming the presence of two gonadotropins from the two different
gonadotrops reported. Secondly, both seasonal and cytological data on
pituitary cells indicate the gonadotrops are active in May (0 and Chan,
1974c), thus, the high activities measured in the gland represent ac-
tive secretory activity. Considering the data as a whole, it appears
that there are obvious differences between the mature female and male,
at least in their "LH" activity. It should be emphasized that these find-
ings are preliminary and have many limitations. Further "investigations
on individual fish, using other bioassays with non-mammalian system,
or larger amounts of purified pituitary materials, or radioimmunoassays
on plasma gonadotropin(s), are necessary. Some aspects of research
along these lines are being carried out in this laboratory.
D. Summary and Discussion
As far as the production of mature gametes and sex hormones is con-

cerned, the function of the gonad in the female and male phase of the
sex-reversing protogynous hermaphrodite, Monopteru2, is closely similar
to the ovaries and testes in gonochoristic teleosts. During the process
of structural transformation, both in vitro and in vivo steroid studies
clarly indicate a decrease in estrogen and an increase in androgen se-
cretions. The role of steroid hormones in the process of natural sex
reversal was studied extensively by using implantations and injections
of estrogenic and androgenic steroids. The rise in androgen during
natural sex change apparently was not the primary-causative factor trig-
gering the onset of the proliferation and differentiation of the resting
male gonia, but rather was a secondary event parallel to the development
of male germinal tissues, perhaps providing them with a compatible
environment in the gonadal substratum and also an endocrine secretory
pattern in concord with the new gonadal sex, which in Monopterus could
not be brought about precociously by simply steroid treatments. In
this respect, the process of natural sex succession in Monopterus seems
to differ from the condition in the embryonic undifferentiated gonads
of those gonochoristic amphibians and teleosts, where estrogen and
androgen could influence (mainly by their antagonistic actions), the
differentiation of the bisexual gonad to either of the two alternativ~
sexes. However, the suppressive effects of estrogen on spermatogenic
cells in the intersex and male phase remain very obvious in Monopterus,
215
indicating that endocrine-environment vs. germinal-tissue interaction

remains one of the important factors controlling sexuality succession.
Unlike its spermatokinetic effect in the intersex and male, the inef-
fectiveness of exogenous androgen (during the female phase) in causing
the destruction of developing oocytes and in enhancing the quiescent
male gonia into spermatogenetic activities suggests, on one hand, the
"age-dependence" of the male gonia in their respective responsiveness
to androgen, and on the other hand, the possible presence of "sex-
specificity" or a "sex specific receptor mechanism" in sex-related
tissues or functions at different sexual phases of Monopterus. In all,
present information and experimental results appear to suggest that
the fundamental basis, within the gonad, of natural sex reversal in
Monopterus resides in part in the innate nature of germ cells and in the
gonadal endocrine interaction and its effects on germ-cell differen-
tiation.
The invariably extensive development of interstitial Leydig tissue

during spontaneous sex reversal in Monopterus indicates a probable in-
volvement of the adenohypophysial secretion(s) as one of the higher
level events controlling, in part, the structural and functional trans-
formation of the gonad. Investigations on the forms and functions of
the adenohypophysis of Monopterus are in progress. Among various aspects,
the identification of adenohypophysial cell types, the effects of mam-
malian gonadotropins on precocious sex change and on the gonadotrops,
have to date reached some definite findings. Conclusive elucidations
on the quality, quantity and patterns of the gonadotropic secretion(s)
and functions in various sexual stages of this sex-reversing teleost
may prove to be very difficult, especially in view of the lack of unan-
imous opinion in the existing information concerning fish gonadotro-
pins. In spite of all recent biochemical studies favoring the idea
of a single gonadotropin in teleosts, there is, however, the problem
of two gonadotropic cell types, as shown by cytological and cytophys-
iological evidence in certain teleosts such as AnguiHa (Olivereau
and Herlant, 1960); Oncorhynchus (Olivereau and Ridgeway, 1962); MugU
(Leray, 1963); Lepomis (Simon, 1972, quoted by Reinboth, 1972); Carassius
(Leatherland, 1972); and Monopterus (0 and Chan, 1974a,c). Electron
microscopic studies on Zoarces (oztan, 1966) and AnguiHa (Knowles and
Vollrath, 1966a,b) also corroborate this duality concept. On the con-
trary, in some fishes, only one type of gonadotropic cell has been
identified (Matty and Matty, 1959; Sage and Bromage, 1970). The possi-
bility of species variations should not be overlooked.
The preliminary biochemical results concerning gonadotropin(s) inMon-

opterus must be interpreted with caution especially in referring to
their LH or FSH nature. However, irrespective of types, the gonado-
tropic activities in OAAD tests have been found to be much higher in
mature females than in the intersexes and males. Whether this high
gonadotropic activity in a mature female is the characteristic pattern
for that particular phase, and being maintained a~ a constant high
level throughout the year, or only represents a short stage marking
a "surge" in secretion during the breeding season is not certain -
though in vertebrates, sexual maturation is known to be related to
gonadotrops (Kallman and Schreibman, 1973) and with increase in gona-
dotropins levels (Tanner, 1967). The problem in Monopterus can only be
resolved by measuring the pituitary or plasma gonadotropic activities
(e. g. by radioimmunoassays). As sex reversal in Monopterus usually is
mainly a postnuptial event and the onset of the process is marked by
interstitial tissue development (which has been shown under the in-
fluence of LH), it is therefore not unreasonable to believe that there
is a natural rise in gonadotropic activities as a normal event in
the life cycle of Monopterus at the mature female phase during the
216
Envi ronment al
( PrImIng on th e HypothalamuS) _ _ _ _-, factors
, - -- - - tor o('w GIH patt ern ,
eNS
GONAD
Germ-Cells Hypothalamus •
: : : : Oogonia ( Inductors?)
Somatic
. t
Adenohypophy si s
.............. .
(Hormonal) o -gonio P )
. . ................ .
...................
:::: :Ovarian
Inductor ,
:::: :growth
...................
................ Age F.? GTH 1"FSH"? I
: : : : : Mature Estrogen I" LH "?I
: : : : : foll icle t---T""":"--'
d -gonia
.. .. .. • .. • • .. • .. • .. 1-
:::::::::::: . ....... ~ ....... --- Age fo ci or

'-:~;.:.:.:. .:.:-:
. :i-':i-i:i-i:,;:,;:-r:-r:f---~O---- .Sp',~'gon'i a ' )~:Wf~ Other toctor ?
: : : : : : : :: : :: + proli ferotion :::::::::::::: (. FSH·? I
:~ ~ :~~~l~!i~?: Interstit ial :~ !J~!~J
Androge n
'---:'::-;:~:H::~:~:::-;:~:~::+--+--....:..;.~ Leydig -cell s ~~!~~~~~~~~~~
.....!"'"G-r-H-(-..L-H ,,?-)-4
IIIIIIi
... . ..................
....................
Ovarian + Sp -gonia :t~:W~:
: : : degeneration sp?~matogene~'i-;'-@:
jliSteroidogenesis
l:1ilil:1:1:1:1:1:1:1:1:1:1:1ililil:lij~J:!
~:j:
& .:-: 6TH
:::: -~l·-:-
FS-H·-.'-lH-'-',-)- --I
Spermatogenesis ~:~:
of the ~-phasd:
f~IIIIIIII'1
Fig. 6. A proposed scheme for the endocrine events of natural sex
reversal in Monopterus showing the various possible- controls in the
life cycle on tne process, involving the nature of the germ-cells,
the gonadal endocrine interaction and the hypothalamo-adenohypophysio-
gonadal axis; genetic and age factor being the ultimate basis of con-
trol
breeding season. Based on our experimental evidence at present avail-

able, it is very likely that, as one of the series of events in the
process of natural sex change in Monopterus, the endocrine function of
tne adenohypophysis may contribute a significant role - in that the
rise in gonadotropic secret.ion in the mature female towards the breed-
ing period could (in addition to its other probable functions such as
217
inducing ovulation), trigger the development of the interstitial Leydig

cells and perhaps to certain degrees also that of the resting male
gonia hence, marking the onset of natural sex reversal. As a tentative
view, the hypothetical scheme on the possible endocrine changes in the
adenohypophysis and the gonad (Fig. 6), forming part of the series of
sequential events in Monopterus during the successive maturation of the
male and female germinal cells in relation to gonadal ontogeny and
life cycle (Chan, 1970), offers a concept of wider perspectives on the
phenomenon of natural sex reversal. It incorporates the hypothalamo-
adenohypophysio-gonadal axis function, the gonadal endocrine inter-
actions and their effects on germinal tissues, and the germ-cell age-
dependent responsiveness to hormonal stimulations. That sexuality suc-
cession in Monopterus may have close relationship with age and develop-
ment and with the function of the adenohypophysio-gonadal axis, does
not necessarily refute the current concepts maintaining a genetic basis
for the process (Atz, 1964; Chan, 1970). As every trait should have
its genetic background and the adult phenotype is the result of a se-
ries of developmental events, it is believed that sexuality succession
in Monopterus is possibly governed by a system of multiple factors con-
trolling the sequential events in gonadal ontogeny and its physiolo-
gical function throughout the life history (Chan, 1970). In teleosts,
there is direct evidence demonstrating the presence of a sex-linked
gene controlling gonadotrop differentiation and determining the age of
sexual maturation in the male Xiphophorus maculatus (Kallman and Schreib-
man, 1973) - indicating the direct link between genetic control, adeno-
hypophysial function and onset of gonadal function. Moreover, the pres-
ent hypothetical scheme with the involvement of the adenohypophysio-
gonadal axis, which is intimately associated to and controlled by the
hypothalamus and CNS, would also open some venues for possible neuro-
endocrine pathways whereby the environmental factors, such as tempera-
ture and light (Harrington, 1971), sex-ratio or social stimuli (Fishel-
son, 1970; Robertson, 1972) could bring about or influence the process
of spontaneous sex reversal in teleosts.
References
Atz, J.W.: Intersexuality in fishes. In: Intersexuality in Vertebrates

Including Man (eds. C.N. Armstrong and A.J. Marshall), pp. 145-232.
London-New York: Academic Press 1964.
Ball, J.N.: Reproduction in female bony fishes. Symp. Zool. Soc. Lon-
don, I, 105-135 (1960).
Bara, G.: Histochemical localization of n 5-3S-hydroxy-steroid dehydro-
genase in the ovaries of a teleost fish, Scomber scomber L. Gen. Compo
Endocr. 5, 284-296 (1965).
Bara, G.: HIstochemical demonstration of 3S-, 3a-, J1S-, and 17S-
hydroxysteroid dehydrogenases in the testis of Fundulus heteroclitus.
Gen. Compo Endocr. 17, 189-200 (1969).
Barr, W.A.: Patterns o~ovarian activity. In: Perspective in Endocri-
nology (eds. E.J.W. Barrington and C. Barker J¢rgensen), pp. 164-
238. London: Academic Press 1968.
Burns, R.K.: Role of hormones in the differentiation of sex. In: Sex
and Internal Secretion (ed. W.C. Young), Vol. 1, pp. 76-158. Balti-
more: Williams and Wilkins 1961.
Burzawa-Gerard, E.: Purification d'une hormone gonadotrope hypophysaire
de pOisson telE~osteen la carpe (Cyprinus carpio L.). Biochem. 53, 545-
552 (1971). -
Chan, S.T.H.: Natural sex reversal in vertebrates. Phil. Trans. R. Soc.
Lond. B. 259, 59-71 (1970).
218
Chan, S.T.H., 0, Wai-sum, Hui, W.B.: The interrenal gland and the ACTH
and prolactin cells in the adenohypophysis of Monopterus and their
roles in osmoregulation. Gen. Compo Endocrinol. (1975, in press).
Chan, S.T.H., 0, Wai-sum, Tang, F., Lofts, B.: Biopsy studies on the
natural sex reversal in Monopterus albus (Posces: Teleostei). J.
Zool. Lond. 167, 415-421 (1972a).
Chan, S.T.H., Phillips, J.G.: The structure of the gonad during natu-
ral sex reversal in Monopterus albus (Pisces: Teleostei). J. Zool.
Lond. 1 51, 1 29 -1 41 ( 1 9 67 a) .
Chan, S.T~, Phillips, J.G.: Seasonal changes in the distribution of
gonadal lipids and spermatogenetic tissue in the male phase of
Monopterus albus (Pisces: Teleostei). J. Zool. Lond. 152, 31-41
(1967b). -
Chan, S.T.H., Phillips, J.G.: The biosynthesis of steroids by the go-
nads of the ricefield eel Monopterus albus at various phases during
natural sex reversal. Gen. Compo Endocrinol. ~, 619-636 (1969).
Chan, S.T.H., Tang, F., Lofts, B.: The role of sex steroids on natural
sex reversal in Monopterus albus . Excerpta Medica Inti. Congr. 256,
348 (1972b). -
Chan, S.T.H., Wright, A., Phillips, J.G.: The atretic structures in
the gonad of the ricefield eel (Monopterus albus), J. Zool. Lond.
ill, 527-539 (1967).
Channing, C,P., Licht, P., Papkoff, H., Donaldson, E.M.: Gen. Compo
Endocrinol. 22, 137-145 (1974).
Chieffi, G.: SeX-differentiation and experimental sex reversal in
elasmobranch fishes. Arch. Anat. Microsc. Morph. esp. (suppl.) 48,
21-36 (1959).
Christiansen, P.: Studies in the rat ovarian augmentation method for
follicle stimulating hormone. Acta Endocrinol. 70, 636-652 (1972).
Colombo, L., del Conte, E., Clemenze, P.: Steroid biosynthesis in vitro
by the gonad of Sparus auratus L. (Teleostei) at different stages
during natural sex reversal. Gen. Compo Endocrinol. 12, 26-36 (1972)
D'Ancona, U.: Ermafroditismo ed intersessualita nei Teleostei. Experien-
tia 5, 381-391 (1949).
Deth, J~H.M.G. van, van Limborgh, J., van Faassen, F.: Le role de
l'hypophyse dans la determination du sexe de l'oiseau. Acta Morphol.
Neerl.-Scandinav. 1, 70-80 (1956).
Dodd, J.M.: Gonadal and gonadotrophic hormones in lower vertebrates.
In: Marshall's Physiology of Reproduction. Vol. 1, pp. 417-582
(ed. A.S. Parkes). London: 'Longmans, Green & Co. 1960.
Donaldson, E.M., Yamazaki, F., Dye, H.M., Philleo, W.W.: Preparation
of gonadotropin from salmon (Oncorhynchus tscha:wythscha) pi tui tary
glands. Gen. Compo Endocrinol. 18, 469-481 (1972).
Exley, D.: Ultramicrodetermination-of plasma testosterone by electron-
capture detection of testosterone diheptaflurobutyrate. Biochem. J.
107, 285-292 (1968).
Faiman, C., Winter, J.S.D.: Sex differences in gonadotrophin concen-
trations in infancy. Nature (Lond.) 232, 130-131 (1971).
Fevold, H.L., Hisaw, F.L., Leonard, S.L~The gonad stimulating and
the luteinizing hormones of the anterior lobe of the hypophysis.
Amer. J. Physiol. 97, 291-301 (1931).
Fishelson, L.: Protogynous sex reversal in the fish Anthias squamipinnis
regulated by the presence of absence of a male fish. Nature (Lond.)
227, 90- 91 ( 1 970) .
Foote, W.D., Thibault, C.: Recherches experimentales sur la maturation
in vitro des oocytes de trine et de veau. Ann. Bioi. animo Biochim.
Biophys. 9, 329-349 (1969).
Gabrilove, J~L.: In: Methods in Investigative and Diagnostic Endocri-
nology (eds. S.A. Berson and R.S. Yalow), Vol. 2A, pp. 541-546.
Amsterdam: N. Holland Publishing Co. 1973.
Gottfried, H.: The occurrence and biological significance of steroids
in lower vertebrates. Steroid 1, 219-242 (1964).
219
Greep, R.O.: Nature, role and control of the gonadotropins. J. Reprod.

Fert., Suppl. li, 1-13 (1970).
Harrington, R.W., Jr.: How ecological and genetic factors interact to
determine when self-fertilizing hermaphrodites of Rivulus marmoratus
change into secondary males, with a reappraisal of the mode of
intersexuality among fishes. Copeia 1971, 389-432.
Harris, G.W.: Hormonal differentiation of the developing central ner-
vous system with respect to patterns of endocrine function. Phil.
Trans. Roy. Soc. Lond. B. 259, 165-177 (1970).
Johnson, D.C.: Sexual differentiation of gonadotrophin patterns. Amer.
Zool. 12, 197-205 (1972).
Jost, A.::Hormonal factors in the sex differentiation of the mammalian
foetus. Phil. Trans. Roy. Soc. Lond. B. 259, 119-130 (1970).
Kallman, K.D., Schreibman, M.P.: A sex-linked gene controlling gonado-
trop differentiation in determining the age of sex maturation and
size of tile platyfish, Xiphophorus maculatus. Gen. Compo Endocrinol.
21, 287-304 (1973).
Knowles, F., Vollrath, L.: II. The structure and innvervation of the
pars distalis at different stages of the life-cycle. Phil. Trans.
Roy. Soc. Lond. B. 250, 329-342 (1966a).
Knowles, F., Vollrath,~: Changes in the pituitary of the migrating
European eel during its journey from rivers to the sea. Z. Zell-
forsch. 75, 317-327 (1966b).
Lambert, J.G.D.: The ovary of the guppy, PoeciUa reticulata. The granu-
losa cells as sites of steroid biosynthesis. Gen. Compo Endocrinol.
15, 464-476 (1970).
Leatherland, J.F.: Histophysiology and innervation of the pituitary
gland of the goldfish, Carassius auratus L. Can. J. Zool. 50, 835-
844 (1972).
Li, C.H.: Hormonal Proteins and Peptides. New York-London: Academic
Press 1973.
Lofts, B.: Patterns of testicular activity. In: Perspectives in Endo-
crinology (eds. E.J.W. Barrington and C. Barker J¢rgensen), pp.
239-304. New York-London: Academic Press 1968.
Lupo Di Prisco, C., Chieffi, G.: Identification of steroid hormones
in the gonadal extract of the synchronous hermaphrodite teleost,
Serranus scriba. Gen. Compo Endocrinol. 5, Abstr. no. 71 (1965).
Matty, A.J., Matty, J.M.: A histochemical-investigation of the pitu-
itary glands of some teleost fish. Quart. J. Microscop. Sci. 100,
257-267 (1959). -
0, Wai-sum: A study on the relationship of the pituitary gland and
natural sex reversal in Monopterus albus. Thesis, University of Hong
Kong, 1973.
0, Wai-sum, Chan, S.T.H.: A cytological study on the structure of the
pituitary gland of Monopterus albus. Gen. Compo Endocrinol. l±, 208-
222 (1974a).
0, Wai-sum, Chan, S.T.H.: A study on the thyroid gland and the TSH-
cells in the pituitary of the ricefield eel, Monopterus albus. Gen.
Compo Endocrinol. 24, 99-112 (1974b). -
0, Wai-sum, Chan, S.T~.: Some aspects of the cytophysiology of the
gonadotropic cells in the adenohypophysis of Monopterus albus. (197 4c,
in prep.).
Olivereau, M.: Cytophysiologie du lobe distal de l'hypophyse des
Agnathes et des Poissons, a l'exclusion de celle concernant la fonc-
tion gonadotrope. In: Cytologie de.l'adenohypophyse (eds. J. Benoit
and C. DaLage), pp. 315-329. Paris: C.N.R.S. 1963.
Olivereau, M.: Observations sur l'hypophyse de l'Anguille femelle, en
particulier lors de la maturation sexuelle. Z. Zellforsch. 80,
286-306 (1967).
Olivereau, M., Herlant, M.: Etude de l'hypophyse de l'Anguille male au
cours de la reproduction. C.R. Seanc. Soc. Biol. 154, 706 (1960).
220
Olivereau, M., Ridgeway, G.J.: Cytologie hypophysaire et antigen serique

en relation avec la maturation sexuelle chez Oncorhynchus species.
C.R.Seanc. Soc. Biol. 254, 753-755 (1962).
Otsuka, S.: On the extraction and bioassay of the follicle stimulating
and luteinizing substances of the salmon. Endocrinol. Japan. 3,
272-277 (1956). -
Overbeeke, A.P. van, McBride, J.R.: The pituitary gland of the sockeye
(Oncorhynchus nerka) during sexual maturation and spawning. J. Fish.
Res. Bd. Can. 24, 1791-1810 (1967).
oztan, 1'1.: The fine structure of the adenohypophysis of Zoarces vivi-
parus L. Z. Zellforsch. £2., 699-718 (1966).
Parlow, A.F.: Bioassay of pituitary luteinizing hormone by depletion
of ovarian ascorbic acid. In: Human Pituitary Gonadotrophins (ed.
A. Albert, C. Charles), pp. 300-310. Springfield/Ill.: Charles C.
Thomas 1961.
Pickford, G.E., Atz, J.W.: The Physiology of the Pituitary Gland of
Fishes. New York: N.Y. Zool. Soc. 1957.
Puckett, W.O.: Some effects of crystalline sex hormones on differen-
tiation of the gonads of an undifferentiated race of Rana catesbiana
tadpoles. J. expo Zool. 84, 39-51 (1940).
Reinboth, R.: Morphologische und funktionelle Zweigeschlechtlichkeit
bei marinen Teleostiern (Serranidae, Sparidae, Centracanthidae,
Labridae). Zool. Jb. (Physiol.) 69, 405-480 (1962a).
Reinboth, R.: The effects of testosterone on female Coris juUs (L.),
a wrasse with spontaneous sex-inversion. Gen. Compo Endocrinol. ~,
abstr. no. 39 (1962b).
Reinboth, R.: Experimentall induzierter Geschlechtswechsel bei Fischen.
Verh. Dt. zool. Ges. Munchen, pp. 67-73 (1963a).
Reinboth, R.: Naturlicher Geschlechtswechsel bei Sacura margaritacea
(Hilgendorf) (Serranidae). Annotnes zool. jap. 36, 173-178 (1963b).
Reinboth, R.: Intersexuality in fishes. Mem. Soc. Endocr. ~, 515-544
(1970) .
Reinboth, R.: Hormonal control of the teleost ovary. Amer. Zool. if,
307-324 (1972).
Reinboth, R.: Studies on gonadal steroid metabolism in ambisexual fish.
Gen. Compo Endocrinol. 22, 342 (abstract) (1974).
Reinboth, R., Callard, loP., Leathem, J.H.: In vitro steroid synthesis
by the ovaries of the teleost fish, Centropristes striatus (L.). Gen.
Compo Endocrinol. 7, 326-328 (1966).
Robertson, D.R.: Social control of sex reversal in a coral-reef fish.
Science 177, 1007-1009 (1972).
Rosemberg, ~ Gonadotropins 1968. Los Altos: Geron-X. 1968.
Sage, M., Bromage, N.R.: The activity of the pituitary cells of the
teleost Poecilia during the gestation cycle and the control of the
gonadotrophic cells. Gen. Compo Endocrinol. L1, 127-136 (1970).
Saxena, B.B., Rathman, P.: Purification and properties of human pitu-
itary FSH. In: Gonadotropins (ed. E. Rosemberg), pp. 3-12. Los Altos:
Geron-X 1968.
Stanley, H., Chieffi, G., Botte, V.: Histological-and histochemical
observations on the testis of Gobius paganeUus. Z. Zellforsch. ~,
350-362 (1965).
Steelman, S.L., Dohley, F.M.: Assay of the follicle stimulating hor-
mone based on the augmentation with human chorionic gonadotropin.
Endocrinology 53, 604-616 (1953).
Sundararaj, B.I., Anad, T.C., Donaldson, E.M.: Effects of partially
purified salmon pituitary gonadotrophin on ovarian maintenance,
ovulation and vitellogenesis in hypophysectomized catfish, Hetero-
pneustes tossilis (Bloch). Gen. Compo Endocrinol. ~, 102-114 (1972).
Sundararaj, B.I., Nayyar, S.K., Anad, T.C., Donaldson, E.M.: Effects
of salmon pituitary gonadotropin, ovine-luteinizing hormone and
testosterone on the testes and seminal vesicles of hypophysectomized
catfish, Heteropneustes tossiUs (Bloch). Gen. Compo Endocrinol. 11,
73-82 (1971).
221
Sundararaj, B.I., Nayyar, S.K., Anad, T.C., Donaldson, E.M.: Effects

of partially purified salmon pituitary gonadotrophin on re-initia-
tion of spermatogenesis in the long-term hypophysectomized catfish,
Heteropneustes fossilis. J. Reprod. Fert. 27, 292-293 (1972).
Tang, F., Lofts, B., Chan, S.T.H.: ~5-3~-hydroxysteroid dehydrogenase
in the ovary of the ricefield eel, Monopterus albus • Experientia 30,
316 (197 4a) .
Tang, F., Lofts, B., Chan, S.T.H.: A study on the 3~- and 17~-hydroxy
steroid dehydrogenase activities in the gonads of Monopterus albus
at various sexual phases during natural sex reversal. J. Zool.
(1974b, in press).
Tang, F., Chan, S.T.H., Lofts, B.: The effect of steroid hormones on
the process of natural sex reversal in the ricefield eel, Monopterus
albus. Gen. Compo Endocrinol. 24, 227-241 (1974c).
Tang, F., Chan, S.T.H., Lofts, B~ The effect of mammalian LH on the
natural sex reversal of the ricefield eel, Monopterus albus. Gen.
Compo Endocrinol. 24, 242-248 (1974d).
Tanner, J.M.: Puberty~In: Advances in Reproductive Physiology (ed.
A. McLaren), Vol. II, pp. 311-347. London: Academic Press 1967.
Wiebe, J.P.: Inhibition of pituitary gonadotropic activity in the
viviparous sea perch, Cymatogaster aggregata Gibbons, by dithiocarb-
amoylhydrazine derivative (I.C.I. 33,828). Can. J. Zool. 46, 751-
758 (1969').
Witschi, E.: The inductor theory of sex differentiation. J. Fac. Sci.
Hokkaido Univ. 13 (5), 428-439 (1957).
Yamamoto, T.: Artificial induction of functional sex reversal in geno-
typic females of the medaka (Oryzias latipes). J. exp. Zool. 137,
227-262 (1958).
Yamazaki, F.: The gonadotropin of fishes. Bull. Jap. Soc. Sci. Fish
35, 695-709 (1969).
Yamazaki, F., Donaldson, E.M.: The spermiation of goldfish (Carassius
auratus) as a bioassay for salmon (Oncorhynchus tschCJJi)ytscha) gonadotropin.
Gen. Compo Endocrinol. 10, 383-391 (1968a).
Yamazaki, F., Donaldson, E.M.: The effect of partially purified salmon
pituitary gonadotropins on spermatogenesis, vitellogenesis, and
ovulation in hypophysectomized goldfish (Carassius auratus). Gen. Compo
Endocrinol. 11, 292-299 (1968b).
Yamazaki, F., Donaldson, E.M.: Involvement of gonadotrophin and steroid
hormones in the spermiation of the goldfish (Carassius auratus). Gen.
Compo Endocrinol. ~, 491-497 (1969).
Yaron, Z.: Demonstration of 3~-hydroxysteroid dehydrogenase in the
testis of TUapia mossambica. J. Endocrinol. 34, 127-128 (1966).
Yaron, Z.: Observation on the granulosa cellso f Acanthobrama terraesanctae
and TUapia nilotica (Teleostei). Gen. Compo Endocrinol. .l2, 247-252
(1971).
Ovarian and Testicular Intersexuality in Two Protogynous
Mediterranean Groupers, Epinephe/us aeneus and
Epinephe/us guaza
J. Brush~ and S. Brush~
A. Introduction
Groupers are perciform teleost fishes, of the family Serranidae. They

are found in tropical Atlantic and Mediterranean regions, where they
occur typically on the continental shelf and are fished commercially
by line-fishing and trawling.
Sexuality in the Epinephelinae (sea-basses, groupers, cerniers, garru-

pas, etc.) has been the subject of a limited number of studies, since
the first obpervations by van Oordt (1933). For example, Smith (1959,
1965) studied Epinephelus adscensionis and Epinephelus striatus in Bermuda,
and Reinboth (1968), Epinephelus chlorostigma from Japan. All the obser-
vations on Polyprion (Kirincic and Lepetic, 1955), Mycteroperca, Cephalo-
pholis and Petrometropon (MacEr lean and Smith, 1964; Smith, 1959, 1965)
led to the conclusion that protogynous hermaphroditism seems to be the
rule in groupers.
This paper is concerned with the sexuality (previously unknown) of the

groupers of E. aeneus and E. guaza, which are abundant all the year in
Tunisian fish-markets.
B. Observations
The observations deal with 854 individuals (6'72 E. aeneus and 182 E.
guaza) whose lengths (LT) and weights (W) were measured and the gonads
histologically studied.
The gonads are ovotestes without distinct ovarian and testicular loca-
lization. They occur as mixed germinal cells in both sexes ("Epinephelus
type": Smith, 1965) and are therefore, different from the "hetero-
sexual" type distinguished in other serranids, such as Serranus scriba,
S. cabrilla and S. hepatus which are synchronous hermaphrodites.
Fishes of these two species were ranged in weight from 0.125 to 20 kg

and were categorized into 7 arbitrary weight classes (I to VII)
(Tables 1 and 2) .
From the histological analysis, it was possible to distinguish three

natural divisions:
a) Weight < 3 kg: All the fishes in the two first classes (I - II) are
non-functional, 100% young and immature females. The ovaries do not
participate in the seasonal cycle of oogenesis.
b) Weight 3 to 9 kg: The greatest number of fishes in the classes III
and IV (80% in E. guaza and 83% in E. aeneus) are functional females.
Table 1. Sexuality in EpinepheZ.us guaza
Weight classes Sex
'i' 'i' (young) 'i' 'i' d'd' 1i
Number Number Number Number
% % % %
of fishes of fishes of fishes of fishes
I o a kg 18 100 0 0 0
II a 3 kg 3 37,5 5 62,5 0 0
III 3,5 a 6 kg 0 22 88 3 12 0
IV 6,5 a 9 kg 0 34 72,3 11 23,4 2 4,2
V 9,5 a 12 kg 0 27 49 28 51 0
VI 12,5 a 15 kg 0 7 43,7 9 56,3 0
VII 15,5 a 20 kg 0 4 30,7 7 53,8 2 15,3
Total = 182 ex. Females: 120 65,9% Males: 62 = 34%
Table 2. Sexuality in Epinephelus aeneus

Weight classes Sex
'i' 'i' (young) 'i' 'i' d'd' 1i
Number Number Number Number
of fishes % of fishes % % %
of fishes of fishes
I o a kg 2 100 0 0 0
II a 3 kg 6 66 3 33 0 0
III 3,5 a 6 kg 0 127 83 22 14,3 4 2,6
IV 6,5 a 9 kg 0 202 63,9 91 28,8 23 7,2
V 9,5 a 12 kg 0 64 38,3 93 55,6 10 6,1
VI 12,5 a 15 kg 0 4 19 14 66,6 3 14,3
VII 15,5 a 20 kg 0 0 3 75 25
Total = 672 ex. Females: 408 = 60,7% Males: 264 = 39,2%
r->
r->
w
224
The ovaries show seasonal maturation at the end of spring and at the
beginning of summer and from July, mature oocytes are ready for spawn-
ing. The ovaries are at rest during autumn and winter.
c) Weight> 9 kg: Some females in this group are still functional, with
the number progressively falling off with increasing weight (E. guaza:
49% class V, 43% class VI and 30% class VIIi E. aeneus: 38% class V,
19% class VI and 0% class VII). On the other hand, the proportion of
males is steadily increasing, becoming more important in the high
weight classes (more so in E. aeneus than in E. guaza, where class VII
still shows 30% females).
c. Discussion
The observations have shown that in E. aeneus and E. guaza all young
fishes are females - the first functional activity being oogenetic.
Males occur among the largest individuals after a sex-reversal. This
confirms that protogynous hermaphroditism seems to be the general
situation in groupers, as has also been asserted by Smith (1959).
Sex-reversal occurs during sexual inactivity, generally at weights

between 5 - 20 kg. Similar variability has been shown in E. merra and
E. sVlTlmanaby van Oordt (1933) and in E. guttatus by Smith (1959). Thus,
some females are larger than males although males are more common in
the higher weight classes.
The period of sex-reversal (ovary + ovotestis + testis) frequently

shows transitional stages in the form of ovarian and testicular inter-
sexuality (Fig. 1-4).
Ovarian intersexuality, with "male nests" inside the female tissues,

has been observed by Smith (1 965) in Petrometropon and by Reinboth
(1967, 1968) in CheUdoperca and E. chlorostigma. It seems to correspond
to the first spermatogenetic activities inside the ovarian gonad and
would appear to be evidence of the forthcoming sex-reversal. Thus,
in some cases classes IV - V may be considered as precursor stages of
sex-reversal.
Testicular intersexuality, characterized by oocytes present in func-

tional testes ("testicular oocytes", Atz, 1964), has been recorded in
E. merra, E. pachycentrum and E. sexfasciatus by van Oordt (1933) and also
by Smith (1959, 1965, 1971) in E. guttatus, E. striatus, E. itajara and
Mycteroperca tigris.
Fig. 1. Testicular intersexuality in E. aeneus. Functional testis con- ~

taining small basophilic oocytes (TL = 80 cm, W = 7 kg, 25.4.71)
Fig. 2. Testicular intersexuality in E. guaza. Functional testis, at
the beginning of spermatogenetic activity, with two basophilic oocytes
(TL = 98 cm, W = 16 kg, 16.2.71)
Fig. 3. Ovarian intersexuality in E. aeneus. Ovary, during sexual rest,
with small basophilic oocytes and some seminiferous cysts with sperma-
togenesis (TL = 85 cm, W = 8 kg, 14.11.70)
Fig. 4. Ovarian intersexuality in E. guaza. "Male nests" inside ovarian
gonad (TL = 78 cm, W = 7 kg, 18.12.70)
225
Fig. 1 - 4
226
These testicular oocytes never reveal auxocytosis or vitellogenesis

and persist "blocked" in the previtellogenetic stage. Later, they
become atretic. The same situation occurs in Cephalopholis (Smith,
1965), Chelidoperca (Reinboth, 1967) and Ps eudogramma (Smith and Atz,
1969), where "2-stage" oocytes are able to survive for a long time
inside the testis. Non-spawned mature and sub-mature oocytes are re-
sorbed at an earlier stage. Testicular intersexuality gives evidence
for a former oogenetic activity and progressive oocyte regression -
it may be considered a late stage of reversal.
Thus, ovarian and testicular intersexuality (class IV) would consti-

tute two successive transitional stages between primitive ovarian
and definitive testicular differentiation (Fig. 5).
However, it seems also possible that some cases of ovarian intersexual-
ity (class VII) may be considered as a partial sex-reversal, sperma-
togenetic manifestations being expressed only in a discrete and limited
manner inside the gonad, where female influence still continues. The
existence, in E. guaza , of large sized females may be the consequence
of an inability to undergo even partial sex-reversal, masculinization
being never revealed in these fishes.
young
OVARY TESTIS
functional
OVARY
1::::1 indiff .
3 ovarian
intersexuality
Fig. 5. Hypothetical scheme of the gonadal evolution in Epinephelus

227
On the other hand, it is possible that testicular intersexuality may

n0t be limited to recent derivation from a previous oogenetic stage,
and that previtellogenetic oocytes may be differentiated periodically
for several consecutive years inside functional testes. This suggestion
presupposes the permanent existence of an oogonial stock inside the
seminiferous cysts. In this case, in "pure males", a potential inter-
sexuality, permanent at the cellular level, would be hidden by sperma-
togenetic activities. In the gonochoristic asteroids, Eehinaster, Henrieia
and Marthasterias, we have shown by electron microscopy of the gonads
(Brusle, 1969) that functional males are, in fact, potential hermaph-
rodites of this type.
Only ultrastructural analysis of precocious germinal cells in these

fishes, will permit the distinction of oogonia and spermatogonia and
confirm this hypothesis.
This problem is at present being investigated.
Aeknowledgements. The authors are very grateful to Mr. Ben Attigue (Tunis)
for his technical assistance and to Mr. Lecomte (Banyuls) for his pho-
tographic work.
References

London: Academic Press 1964.
Brusle, J.: Hermaphrodisme potentiel chez les males de trois especes
d'Asterides. Ann. Embryol. Morph. 2, 445-450 (1969).
Kirincic" J., Lepetic, V. : Recherches-sur l'ichthyobenthos dans les
profondeurs de l'Adriatique meridionale et possibilite d'exploita-
tion au moyen de palangres. Acta adriat. 7, 3-113 (1955).
MacErlean, A.J., Smith, C.L.: The age of sexual succession in the pro-
togynous hermaphrodite Myeteroperea mierolepis. Trans. Amer. Fish.
Soc. 93, 301-302 (1964l.
Reinboth-;-R.: Protogynie bei Chelidoperea hirundinaeea (C. et V.) (Serra-
nidae). Ein Diskussionsbeitrag zur Stammesgeschichte amphisexueller
Fische. Annot. zool. jap. 40, 181-186 (1967).
Reinboth, R.: Zum Problem der-aIDphisexuellen Fische. Zool. Anz., Suppl.
31, 316-325 (1968).
Smith, C.L.: Hermaphroditism in some fishes from Bermuda. Pap. Michigan
Acad. Sci. Arts Lett. 44, 111-119 (1959).
Smith, C.L.: The pattern of sexuality and the classification of Serra-
nid fishes. Amer. Mus. Novit. ,2207, 1-20 (1965f.
Smith, C.L.: A revision of the american Groupers: Epinephelus and allied
genera. Bull. Amer. Mus. nat. Hist. ~, 67-242 (1971).
Smith, C.L., Atz, E.H.: The sexual mechanims of the Reef Bass Pseudo-
gramma bermudensis and its implications in the classification of the
Pseudogrammidae (Pisces: Perciformes). Z. Morph. Tiere 65, 315-326
(1969). -
Van Oordt, G.J.: Zur Sexualitat der Gattung Epinephelus (Serranidae,
Teleostei). Z. mikr. anat. Forsch. 11, 525-533 (1933).
Hermaphroditism In Cichlid Fishes *
H. M. Peters
I would like to discuss a peculiar type of hermaphroditism, which has

been discovered in the 'Mbuna'. The Mbuna are a group of cichlid fishes
endemic to Lake Malawi. They are unique in various aspects of taxonomy,
so that Fryer (1959), who made a fundamental ecological study of them,
proposed to grant them the rank of a tribus. The Mbuna are small fish
living around the shore of the lake, mostly in shallow water. The dif-
ferent species are in a fascinating way adapted to the many special
niches within the great diversity of ecological conditions along the
rocky and sandy seashore, where they live in extremely dense popula-
tions. Some Mbuna species have been introduced as aquarium fish, be-
cause of their beautiful coloration and their interesting mode of re-
production; all the females being mouth-brooders.
There are 9 genera of'Mbuna, comprising 28 species. The present study

refers to 7 of these genera and to 14 species. They are listed in
Table 1. It may be noted, that only the two monotyic genera Cynot'ilapia
and Gephyrochromis have not been studied.
By far the most of my material was collected during two trips to the
southern part of Lake Malawi. Some specimens, which had been brought
back from Malawi were kept and observed in our laboratory before fixa-
tion, whereas a few others were given to me by aquarists (Table 1).
I suspected some form of hermaphroditism in the Mbuna-group, due to
the fact that whereever sexual polymorphism occurs in teleost fish,
something unusual is likely to happen with the gonads. In fact, dif-
ferent sexual morphs exist in some Mbuna species (Fryer and Iles,
1972). For instance, in Lcibeotropheus there is one type of female with
a rather uniform blue-grey body. The body of the other type is marked
by black blotches on an orange ground.
Although at least in this case the different morphs can probably not
be related to differences in the state of the gonads, this starting
point was nevertheless a 'fruitful one which provided findings that
were of some interest.
I have examined 65 Mbuna ovaries and 63 testes (Table 1) by means of

histological preparations. Most of the specimens were sexually mature
or almost mature; in all specimens ovaries or testes had at least
reached their typical structure.
The structure of the ovaries follows the general pattern known from
other cichlids. It may be mentioned, that the left shank of the organ
(as similar with testes) is very much reduced in size. No ob~ervations
could point towards hermaphroditism in females. Furthermore, no differ-
ences between the ovaries of representatives of the two female morphs
of Lcibeotropheus fueUeborni could be detected (Table 11. On the other
*Dedicated to Professor Dr.Dr.h.c. Bernhard Rensch for his 75th birthday

229
hand, all males turned out to be hermaphroditic, insofar as all of

them contained oocytes in their testes. The quantity of these oocytes
seems to be different, depending on the species and to a lesser degree,
on the individuals. Fig. 1 shows a part of a testis, which may be con-
sidered as typical. One may find testes, where the inclusion of oocytes
is much denser, whereas sometimes their number is less striking. Com-
pared with all other species only in Pseudotropheus livingstonii were
testis-oocytes (with one exception) relatively scarce. It became ob-
vious that counts would be necessary to provide more precise informa-
tion.
The oocytes lie everywhere in the spermatogenetic tissue attached to

the wall of the tubuli. They already become distinguishable from sper-
matogonia at the very onset of spermatogenesis (Fig. 2). With the pro-
gress of spermatogenesis they grow in a given part of testis. Finally,
when the surrounding mass of spermatogenetic tissue has been trans-
formed into spermatozoa (accumulated in the lumen of the tubulus) the
oocytes are to be seen isolated along its wall (Fig. 3).
When they have reached a certain siz.e they stop growing. By far the
most of the largest oocytes, which I found in testes of mature Mbuna,
had a diameter of about 0.05 - 0.08 mm. In relatively large males one
may find individual oocytes which had grown larger (Fig. 1), the maxi-
mum size found was approximately 0.2 mm. Even this size is much less
than the oocytes in the ovaries, when they start yolk deposition.
The fact that larger oocytes have never been found in testes is in
favour of the assumption, that there is no transformation of the herm-
aphroditic males into sexually functioning females. Of course, one may
question whether the examined animals really had reached the size
(age?) in which such a transformation could perhaps take place. The
answer is that there were at least some males which had reached or
surpassed the species-specific maximum body size, as reported in liter-
ature (Table 1). These individuals also showed no signs at all of
transformation. Two of them apparently were sterile. The testes con-
tained little sperm and very little spermatogenetic tissue. In one,
Pseudotropheus auratus (see Addenda no. 2), most of the very numerous
oocytes were very deformed, indicating degeneration. The other testis,
Laheotropheus trewavasae (Fig. 4, see also Addenda no. 1), was even more
inactive, and again there was no sign of further development in the
female direction. The spermatogenetic tissue had largely disappeared
and most of the tubuli were practically empty.
When one looks at such a section, where most of the oocytes are con-
fined to the outer layer of the testis (the more median parts of the
tubuli being more or less devoid of them) one might suggest that at
least a portion of the more centrally lying oocytes had been ejected.
But it is still an open question whether such ejeGtions really take
place. There were occasionally sections to be seen which showed oocytes
detached from the wall of tubuli and lying within the central sperm
mass. But since this sperm mass shrinks by fixation, it cannot be ex-
cluded that the detachment of those oocytes took place in connection
with that shrinkage.
I have not compared the structure of testis-oocytes with ovary-oocytes

in much detail. However, at first glance one difference is striking.
The nuclei of teleost oocytes in the first meiotic prophase are char-
acterized by numerous small nucleoli beneath the nucleus envelope.
Such peripheral nucleoli begin to appear very early in the development
of ovary-oocytes (Fig. 6). In contrast to this, testis-oocytes in the
corresponding stages of size, contain very large individual nucleoli
Table 1. List of material examined in this study. The localities refer to the places where fishes were ~
collected by the author; all these places are situated in the southernmost part of Lake Malawi, in the 0
vicinity, or not far from, Monkey Bay. The locality "Monkey Bay" includes also some places situated
very close to that village, such as Tumbi Island. The dates refer to the times when the fish were caught
and the gonads fixed. Lengths are given as total lengths; maximum lengths recorded are cited after
Jackson (1961). Four of the 25 females Labeotropheus fueZlebomi (lengths 54, 57, 60, 99 rom) belonged to the
"blotched" type. It has become doubtful that MeZanoehromis meZanopterus and MeZanoehromis vermivorus can be
regarded as different species (Trewavas, pers. corom.); so these names are used here provisionally
Species Number Length Number Length Locality Date Max.length

of (rom) of (rom) recorded
females males (rom)
Cyatoehromis 3 102, 111, 11 5 8 94, 103, 106, Monkey Bay 22/23. 150
obZiquidens 113, 114, 116, Dec. 69
Trewawas 117, 122
Genyoehromis 65 Monkey Bay 1 5 . Aug. 68 1 26

mento
Trewavas
Labeotropheus 25 39, 41, 42, 43, 8 40, 54, 58, 65, Monkey Bay Aug. 68 110
fue Uebomi 44, 45, 45, 47, 66, 68, 70, 108 and Mwalamba
AHL 54, 55, 56, 57, Peninsula 4.Jan.70
60, 62, 62, 66, d' 70 aquarium
67, 73, 74, 76, ~~ 54, 57, 60,
80, 93, 96, 99, 62, 66, 74, 99,
106 (39-106) 106 aquarium
Labeotropheus 114 aquarium 117
trewavasae
Fryer
Labidoehromis 3 42, 48, 58 45 Monkey Bay 23.Dec.69 seldom

veUieans exceeds
Trewavas 70 rom
MeZanoehromis 67 Monkey Bay 25.Dec.69 120
brevis
Trewavas
Melanoahromis 109 Monkey Bay 23.Dec.69 129
melanopteT'Us
Trewavas
Melanoahromis 55 3 48, 54, 75 Monkey Bay 23./24. 95

vennivoT'Us Dec. 69
Trewavas
Petrotilapia 5 44, 47, 53, 95, 6 47, 49, 53, 54, Monkey Bay 23-30. 250
tridentiger 133 55, 86 Dec. 69
Trewavas if 86 aquarium
Pseudotropheus 7 44, 52, 55, 59, 7 57, 61 , 64, 67, aquarium 90

auratus 95, 99, 100 73, 92, 115
(Boulenger)
Pseudotropheus 7 58, 59, 70, 70, 3 56, 61 , 106 Monkey Bay 23.-30. 82
elongatus 73, 77, 83 Dec. 69
Fryer
Pseudotropheus 7 77, 101 , 107, 8 83, 88, 97, 104, Mazinzi 29.Dec.69 150
Uvingstonii 108, 114, 115, 130, 140, 147,
(Boulenger) 121 152
Pseudotropheus 5 53, 62, 68, 89, 7 59, 74, 100, Monkey Bay 23-30. about 130
tropheops 92 103, 104, 112, Dec. 69
Regan 143 if 143 aquarium -
Pseudotropheus 2 85, 86 8 86, 96, 96, 97, Monkey Bay 23./24. 110
zebra 100, 101, 102, Dec. 69
(Boulenger) 108
~
'"
232
Fig. 1. Pseudotropheus tropheops, part of testis. (For explanation see

Addenda 1(5)
Fig. 2. Pseudotropheus fuelleborni, length 70 mm, part of testis showing
onset of spermatogenesis and small oocytes
Fig. 3. Pseudotropheus auratus, length 64 mm, isola ted oocytes at the wall
of a testis tubulus
233
instead of the small peripheral ones; one may call them "giant" nu-
cleoli (Fig. 5). Although it can certainly be said that giant nucleoli
are characteristic of Mbuna testis-oocytes, small peripheral nucleoli
are not always absent. In the relatively rare cases where testis-
oocytes had grown larger than those usually found, they in fact showed
small peripheral nucleoli (Fig. 1). Thus, giant nucleoli are surely
not due to a complete blockage, but more likely to a temporal delay
of the p'rocess of formation of peripheral nucleoli.
These observations point towards one of the many open question offered
by the Mbuna gonads. It is tempting to speculate about these questions,
especially the basic question, namely: why among gonocytes lying side
by side in the same group, do some develop in the female direction
whereas others concurrently become spermatocytes and finally spermato-
zoa? Since these differentially developing gonocytes seem to be ex-
posed to the same external influences, it may be necessary to take
into account some kind of differential cell division of the gonocytes.
I myself do not plan to pursue the study on the Mbuna gonads and con-
sider this paper only as a preliminary note.
In the laboratory, I was chie'fly interested in the behavior of Mbuna.
From the vLewpoint of ethology I was surprised to notice that the
sexual behavior of Mbuna males, whose testes are usually overcrowded
with oocytes, followed exactly the male behavior pattern of other
mouthbrooding cichlids, whose testes are, so to speak, "normal".
Addenda
I. State of the gonads of some large Mbuna males
1 . Laheotropheus tl'eWavasae (Fig. 4)

Length 114 rom (max. length recorded 117 rom, see Table 1)
Weight 26,2 g
Age not known
From an aquarium, where it had spawned several times
State: Testis apparently sterile, almost no sperm; many tubuli
practically empty, without spermagenetic tissue; very
little active spermiogenesis. Oocytes in the lateral
bording zone extremely numerous, but not larger than
usual, the largest measuring about 0.05 - 0.08 mm.
2. Pseudotl'opheus auratus
Length 92 ,rom (max. length recorded 90 rom, see Table 1)
Age not known
From an aquarium
State: Testis had become surprisingly brown colored, almost
opaque. Little sperm and very little spermiogenetic tis-
sue; interstitial tissue very conspicuous. Oocytes very
numerous, largest about 0.06 rom, most of them much de-
formed in an irregular fashion.
~ Fig. 4. Laheotl'opheus tl'ewavasae, part of testis. (For explanation see
Addenda 1/1)
Fig. 5. Laheotl'opheus fueUeborni, same individual as Fig. 2, testis-oocyte
with a "giant" nucleolus. Inset: nucleus of another oocyte of this
testis showing an extremely large "giant" nucleolus
Fig. 6. Labeotl'opheus fueUeborni, length 66 rom, young ovary oocyte show-
ing small peripheral nucleoli
234
3. Pseudotropheus auratus
Length 115 nun
Age: The fish had been caught at Monkey Bay, Sept. 1968 (length
estimated at about 65 nun) and then reared in our laboratory,
where it spawned several times: at least 2 clutches (July and
August 1969) were definitely fertilized. The fish was fixed
6.12.1969, thus its age was more than 15 months; estimated age
about 21 months.
State: Testis overmature, sperm plentiful, but very much of the
spermatogenetic tissue had disappeared. Spermiogenesis
confined to the lateral zones. Numerous oocytes in these
parts. Most of the larger oocytes measured about 0.08 nun,
one was found with a diameter of 0.125 nun.
4. Pseudotropheus elongatus
Length 106 nun (max. length recorded 82 nun, see Table 1)
Caught and fixed Monkey Bay, 30.12.69
State: similar to no. 3
5. Pseudotropheus tropheops··
Length 143 nun (max. length recorded 'about 130 nun', see Table 1)
Age: The fish had been caught Aug. 1968 at Monkey Bay (estimated
at 65 nun) and was then reared in our laboratory. First spawning
act 21.10. 1968, spawned about 10 times, at least several clut-
ches were fertilized. Last spawning act about 4 weeks before the
fish was killed (25.11.70). Thus its age was more than 2 1/4
years, most probably nearer to 3 years.
State: Testis very mature, with much sperm; spermiogenesis in

the broad lateral zones still active. Oocytes very numer-
ous; most of the larger ones were about 0.08 nun; at one
place a group of 3 oocytes was found, measuring about
0.125 nun each (Fig. 1); another large oocyte measured
0.19 nun.
II. Number of "giant" nucleoli

In a def ini te part of the testis of the Pseudotropheus tropheops male
just mentioned (no. 5) the number of the nucleoli were counted.
There were 55 oocytes, ranging from 0.02 to 0.07 nun. The nuclei
of 50 of these cells contained only one large nucleolus; in 5 nu-
clei there were two nucleoli of different size. Three of the 50
single nucleoli seemed to be in the state of dividing. (For the
general problem, see Busch and Smetana, 1970,)
III.Hermaphroditism in other cichlids

Some observations on hermaphroditism in tank feared cichlids (pro-
togyny in Aequidens; 'ovotestes' in Cichlasoma) were reported by Polder
(1971). I have occasionally found a few small oocytes in testes of
cichlids other than Mbuna (e.g. Tilapia). The peculiarity of the
Mbuna seems to lie in the fact, that in this group testis-oocytes
occur regularly and that their quantity usually is extremely high.
Acknowledgements. I would kike to thank my friend David Eccles, Senior

Research Officer in Monkey Bay, Malawi, without whose enthusiastic help
this study would never have been possible. Mr. Eccles not only provided
the technical equipment for collecting the material, but also put his
profound knowledge at my disposal. He also identified the specimens.
With reference to some special cases, Dr. Ethelwyn Trewavas (British
235
Museum, London) definitively cleared up some taxonomic questions.

I have also to thank Dr. Sylvia Berns, not only for her ceaseless
assistance during our field work in Malawi, but also for the intelli-
gent way in which she cared for and observed the Mbuna, which we had
brought back from Malawi to our laboratory.
References
Busch, H., Smetana, K.: The Nucleolus. New York-London: Academic Press
1970.
Fryer, G.: The trophic interrelationships and ecology of some littoral
communities of Lake Nyasa with special reference to the fishes, and
a discussion of the evolution of a group of rock-frequenting CiehZi-
dae. Proc. zool. Soc. Lond. 132, 153-281 (1959).
Fryer, G., Iles, T.D.: The Cichlid fishes of the Great Lakes of Africa.
Edinburgh: Oliver and Boyd 1972.
Jackson, P.B.N.: Check List of the fishes of Nyasaland. Occ. Pap. Nat.
Mus. S.Rhodesia 3, No. '25 B, 535-621 (1961).
Polder, J.J.W.: On gonads and reproductive behaviour in the Cichlid
fish Aequidens portaZegrensis (Hensel). Netherl. J. Zool. 21, 265-365
(1971). -
An Unusual Approach to Experimental Sex Inversion in the
Teleost Fish, Betta and Macropodus
P. Becker, H. Roland, and R. Reinboth
A. Introduction
In 1937, Noble and Kumpf reported briefly that in juvenile female

fighting fish (Betta sptendens), spaying may induce the regeneration of
testicular tissue and the development of male secondary sex characters
in some individuals which have undergone this operation. In 1951,
Kaiser and Schmidt obtained corresponding results after simular exper-
imental intervention. Although occasional occurrence of various forms
of intersexuality is well known from teleosts (Atz, 1964) and spontane-
ous sex inversion as well as simultaneous hermaphroditism does occur
among bony fish (Atz, 1964; Reinboth, 1970), those phenomena in Betta
are quite unusual. Since the forementioned papers by Noble and Kumpf
(1937) and Kaiser and Schmidt (1951) do not provide sufficient tech-
nical data, we started to repeat and extend those studies, in order
to learn more about the mechanisms that might be involved in this
strange process. Since the results of a thesis by Becker (1969) and
a further study by Eckert-Roland (1969) on the same topic have not
been published, the following report summarizes the essential data
that we obtained.
1. Juvenile females of Betta sptendens measuring 2,0 - 3,5 cm were oper-

ated under a stereo-microscope after anaesthesia in a 2,5% solution
of ethyl carbamate (urethane). A dorso-caudal incision, between 2 and
4 mm in length and starting near the white genital papilla (the genital
papilla of males is greyish), enabled us to remove the genital organs.
These were preserved in Bouin's fixation-fluid for later histological
examination.
2. In a similar operation one half of the ovary was used for histolog-
ical studies, the other one was squashed between two slides and the
resulting tissue-brei re-implanted into the genital cavity of its
donor.
3. After spaying, one half of the ovary was implanted into the body
cavity of another individual. (In Betta, the genital and body cavity
is separated from each other.) After 6, 8 and 10 days respectively,
the implanted tissue was removed from its host; a small piece was used
for histological inspection; the larger part was re-implanted auto-
plastically into the original donor (mostly back into its genital cav-
ity, partly, however, also into the body cavity) .
All the treated animals were recognized, either by their particular

color pattern or by tagging them individually (Reinboth, 1954).
237
In some cases "Terrarnycin"-Pfizer was successfully used as an anti-

biotic agent against infections.
4. A smaller number of normal animals received a single intraperitoneal

injection of 0,5 mg androgenic hormone ("Perandren"-Ciba = testosterone-
isobutyrate) .
Most of our experiments were carried out with Betta splendens. But the
paradise fish, MacY'opodus opeY'cularis , has also been used for experiments
of type 1 and 2.
C. Results
From 87 specimens of Betta in which simple castration was performed

(procedure 1), ovarian tissue regenerated in 23 fish (33 animals proved
to be sterile at later histological inspection). Nearly half of those
(15) exhibited definite male secondary sex characters. Four fishes
could not be distinguished externally from true males after several
months, while one of the two built a typical foam nest after 8 months
and paired normally with a female - but at.e the eggs 24 hours later.
Histological examination of its genital region revealed that no germ
cells were present but in regenerated tissue, clusters of cells were
easily discernible, which must be considered as interstitial cells
from a comparison with interstitial cells in a normal testis (Fig.1a,b)
Fig. 1 a and b. Interstitial cells (ic) in Betta. (a) sterile regener-

ated tissue from a spayed female, (b) testicular tissue from a normal
male
238
In the remaining 26 individuals, sex inversion did occur (as judged

from the appearance of testicular tissue with various stages of active
spermatogenesis) .
In 12 animals which were allowed to survive longer than 3 months, full

development of all male sex characters was achieved. Since in two of
these, the regenerated testicular tissue was in direct contact with
the former oviduct (being still open), the animals successfully fertil-
ized the eggs of females with which they had mated (cf. Fig. 6).
In 5 cases, testicular and ovarian tissue was present side by side,

but with variations in the topographical distribution of the hetero-
logous germinal tissues and the conditions of the ovarian component
(being in degeneration or with large growing oocytes; Fig. 2).
Fig. 2. Ambisexual gonadal tissue with a functional testicular part

and maturing oocytes in Betta
Fig. 3. Regenerated ambisexual gonadal tissue in Macropodus. Small oocytes
among testicular lobules with full spermatogenetic activity
An explanation for the fact that all the five fish with ambisexual
gonadal tissue (including male interstitial cells) retained their fe-
male characters, remains to be found (suppressive influence of the
ovarian elements?) .
The use of procedure 2, as described proviously, brought about a sig-

nificant increase of sex inversion (Table 1).
239
Table 1. Effects of autoplastic reimplantation of ovarian tissue-"brei"
Total number Number of specimens with

of fish sterile ovarian testicular ambisexual
treated gonadal tissue
a) Betta
splendens 47 4 12 23 8
b) Macropodus
opercularis 49 2 20 27
Testicular tissue developed in 48% of the treated fish and according

to the length of time before sacrifice, the male secondary sex charac-
ters appeared more or less fully. Increasing aggressiveness, about
1 month after the operation, proved to be one of the first symptoms
for the onset of sex inversion. Moreover, in no case was a connection
between the testicular tissue and a genital opening established in
this series. Therefore, there had been no successful mating, although
courtship-behavior, pair formation and its frequency did not differ
from normal males.
In a pilot study preceding this experiment, an attempt was made to

investigate the details of tissue regeneration and its histological
transformation. From our observations we propose the working hypothe-
sis that follicular cells might transform into gonocytes. In specimens
which were sacrificed at different intervals after reimplantation of
the "tissue-brei", no gonocytes were discernible after two weeks in
the mass of tissuo consisting of many cells - the features of which
are hard to define.
In Macropodus, 49 specimens were investigated after reimplantation of

squashed ovarian tissue (Table 1). The results obtained are basically
similar to those described above for Betta.
The two animals with sterile tissue regenerates acquired the male ha-
bitus within 2 months, due to the development of aggregations of inter-
stitial cells. The highly aggressive fish built foam nests but did not
mate. The operated fish, which regenerated pure ovarian tissue, partly
reached sexual maturity and eggs could be squeezed out by slight pres-
sure on the belly.
It is a remarkable difference that in Macropodus the development of

testicular tissue never occurred without the presence of ovarian ele-
ments of different size in various stages of sexual maturity (Fig. 3).
However, in each case the animals acquired male secondary sex charac-
ters and displayed male behavior. Nine inverted specimens performed
mating with females and 5 of them were able to fertilize the eggs suc-
cessfully. From 4 matings of 2 "secondary males" which yielded more
than 10 larval fish, we obtained altogether 223 FI-individuals. For
histological examination they were fixed between day 40 and 103, that
is beyond the age at which Macropodus has passed a stage of juvenile
hermaphroditism, according to Schwier (1939). All of them proved to
be females. Therefore, it is concluded that in Macropodus the male sex
is heterogametic and that true sex-inversion had taken place.
From many histological pictures of the germinal tissues of the treated

animals our assumption is substantiated - that follicular cells might
play an important role in sex-inversion. We frequently observed the
occurrence of spermatocytes in clefts between oocytes and their en-
larged follicular envelope (Fig. 4).
240
Fig. 4. Cleft between oocy.te and enlarged follicular cells (fe).

Spermatocytes (ee) are clearly visible in the lumen .
Fig. 5. Grafted ovarian tissue of Betta in heterotopic position, 10
days after the operation. The gonadal tissue had previously been in
another animal for about 1 week
In experiments with Betta, in which the transplantation-technique was

used (procedure 3), we observed a marked hypertrophy of the nuclei of
the follicular cells and heavy degeneration of the oocytes, when the
graft remained for 6 days or more in the host (Fig. 5). Table 2 summa-
rizes the results which were obtained 45 - 65 days after autoplastic
re-implantation of the tissue (which had suffered an immunological
reaction because of the previous homoplastic grafting) .
Table 2. Development of tissue regenerates after transitory homoplastic

transplantation
Type of Number of specimens with re-implanted

regeneration grafts in the
gonadal general body
cavi ty
No regeneration 3 7%) 2 11 %)
Sterile tissue 9 21 %) 3 17%)
Ovarian tissue 7 17%) 7 38%)
Testicular tissue 19 45%) 3 17%)
Ambisexual tissue 4 10%) 3 17%)
Total 42 (100% ) 18 (100%)
241
In about 50% of the total number of treated fish we discovered various

stages of testicular development, partly with full spermatogenetic
activity (Fig. 6). Since the development of germinal tissue was also
observed in those cases where the tissue had been implanted in a heter-
otopic position in the general body cavity, it seems certain that the
newly formed germinal elements had their origin in the tissue which
was removed from its donor. The inversion rate was larger when the
period of the homograft-condition lasted for only 6 or 8 days. Adminis-
tration of androgenic hormone in 22 females of Betta (hav ing the same
size as those which underwent surgical intervention) had no masculi-
nizing effect on the gonad. After initial damage of the germinal tis-
sue, in which the larger oocytes were more strongly affected and became
atretic (Fig. 7), the ovary fully recovered after about 4 weeks.
Fig. 6. Regenerated testicular tissue of Betta with direct connection

to the genital papilla
Fig. 7. Degenerated ovary of Betta 1 1/2 weeks after injection of andro-
genic hormone
D. Discussion
The results of the foregoing studies raise two major problems: (1) what
type of cells are the precursors for spermatogenic dev elopment?
(2) which mechanisms are causative for the onset of testicular develop-
ment? In amphibia, Furnari-Savoca (1967) claimed somatic cells of the
sperm duct to form the crop of new spermatogonia, after complete cas-
tration of male Discogl ossus. From the particular design of our experi-
ments (gonad development in the body-cavity in a number of cases), we
242
infer that parts of the gonad itself must contain the elements for the
newly developing germinal tissue and many details of our histological
preparations support the assumption that follicular cells transform
into gonocytes. Differentiation of oocytes from follicular cells has
been reported by various authors (e.g. Wheeler, 1924; Craig-Bennett,
1931, Yamamoto, 1956, Srir~mulu and Rajalakshmi, 1966). If this is
correct, and when we consider such cells as sexually bipotential, the
mechanism remains to be found which favors male differentiation. We
have been unable to pinpoint any environmental influence (temperature,
light, supply of food, etc.) as a sex-inducer. Even infection by Ichthyo-
sporidium, which according to Wurmbach (1951) might help to explain sev-
eral cases of teratological sex-inversion in fish, cannot be considered
as responsible for the high rate of sex-inversion in our animals. In
no case was external masculinization observed when ovarian tissue re-
generated, although heavy infection by Ichthyosporidium could be seen.
Nevertheless, it is worthwhile to mention that the percentage occur-
rence of testicular tissue in Betta was twice as high in infested spe-
cimens. We tend to assume that our experimental interventions created
some kind of a hormonal imbalance - which favored male differentiation.
The chemical features of the change in the internal milieu remain to
be established.
References

Becker, P.: Experimentell induzierter Geschlechtswechsel bei Anaban-
tiden. Diss. Mainz (1969).
Craig-Bennett, A.: The reproductive cycle of the three-spined stickle-
back, Gasterosteus aculeatus Linn. Philos. Trans. Roy. Soc. 219, 197-
279 (1931).
Eckert-Roland, H.: Geschlechtswechsel beim weiblichen Kampffisch nach
homoplastischer Ovartransplantation und autoplastischer Rlickver-
pflanzung des Explantats. Diplomarbeit, Univ. Main (1969, unpubl.).
Furnari-Savoca, G.: Regeneration of the testes in the Clawed Toad
(Discoglossus pictus Otth.) after complete surgical removal. Experien-
tia l.l (3), 218-219 (1967).
Kaiser, P., Schmidt, E.: Vollkommene Geschlechtsumwandlung beim weib-
lichen siamesischen Kampffisch Betta splendens. Zool. Anz. ill, 66-73
(1951).
Noble, G.K., Kumpf, K.F.: Sex reversal in the fighting fish, Betta
splendens. Anat. Rec. 1Q (Suppl. 1), 97 (1937).
Reinboth, R.: Eine Methode zur Markierung von Aquarienfischen. Zool.
Anz. 153, 190-194 (1954).
Reinboth~.: Intersexuality in fishes. Mem. Soc. Endocrinol. 18, 515
-543 (1970). -
Schwier, H.: Geschlechtsbestimmung und -differenzierung bei Macropodus
opercularis, concolor, chinensis und deren Artbastarde. Z. ind. Abst.-
Vererbl. 77, 291-335 (1939).
Sriramulu, V~ Rajalakshmi, M.: Origin of new crop of oocytes in Gobius
giuris (Hamilton-Buchanan). Z. mikr.-anat.Forsch. 12, 64-73 (1966).
Wheeler, J. F. G.: The growth of the egg in the dab (Pleuronectes limand£l) .
Quart. Journ. microsc. Sci. 68, 641-660 (1924).
Wurmbach, H.: Geschlechtsumkehr bei Weibchen von Lebistes reticulatusbei
Befall mit Ichthyosporidium hoferi Plehn-Mulscw. Roux' Arch. 145, 109-124
(1951) .
Yamamoto, K.: Studies on the formation of fish eggs. I. Annual cycle
in the development of ovarian eggs in the flounder, Liopsetta obscura.
J. Fac. Sci., Hokkaido Univ. Ser. VI, Zool. 12 (3), 362-373 (1956).
The Possible Significance of Sex-Chromatin for the
Determination of Genetic Sex in Ambisexual Teleost Fishes
J. A. P. Mehl and R. Reinboth
A. Introduction
Barr and Bertram (1949) were the first to describe a darkly-staining

body present in the interphase nuclei of female cats but absent in
male nuclei. It is known that this body is the DNA-positive sex-chro-
matin and is derived from the heterochromatic X-chromosome of the
genetic XX female, the remaining X-chromosome being euchromatic and
not visible at interphase. In the genetic XY male the single X-chro-
mosome is euchromatic and is, therefore, also not visible at inter-
phase. In recent years, the simplicity of this sex-chromatin test for
sexual dimorphism has become of vital diagnostic use in the evaluation
of genetic sex, especially in man and other mammals, but also in other
phyla as well (Mittwoch, 1973).
It is, therefore, surprising that until the report by Mehl (1973) which
described sex-chromatin from female liver tissue of the rudimentary
hermaphrodi tic sparid teleost, Lithognathus lithognathus, there has been
no other record of sex-chromatin from teleosts. Mehl was able to dem-
onstrate that the sexual dimorphism of sex-chromatin in this second-
arily gonochoristic species apparently conformed to mammalian princi-
ples - the possible genetic sex chromosome constitution being female
XX and male XY or Xo.
This line of research has been continued. The present report deals with
the results of sex-chromatin tests on the gonochoristic cichlid spe-
cies, Haplochromis burtoni (GUnther), and on six species of ambisexual
teleosts: PageUus acarne (Risso), PageUus erythrinus (L.), Diplodus sargus (L.)
and Spondyliosoma cantharus (L.) - (Sparidae); Coris julis (L.) - (Labridae) ;
and Serranus cabriUa (L.) - (Serranidae) .. The results are evaluated and
compared with what is known about the highly complex situation of sex
determination and genetic constitution in teleosts.
Small pieces of liver were collected from all the specimens examined:
Haplochromis burtoni, from stocks maintained in the aquaria of our Insti-
tute; and from the remaining six hermaphroditic species caught in the
Mediterranean Sea, at Banyuls-sur-Mer, France. Fixation, preservation,
sectioning, staining and analysis of results proceeded exactly as de-
scribed in Mehl (1973). The state of the gonads of all specimens was
recorded. In the case of H. burtoni and C. julis the sections were ran-
domly numbered to avoid biased analyses.
244
c. Results and Discussion
Table 1 summarizes the results obtained from all the species examined
for sex-chromatin (Barr) bodies in liver tissue.
Table 1. Summary of sex-chromatin diagnoses from liver tissue of one

gonochoristic and six ambisexual teleosts
Species Known type of Sex Sex-chromatin Polyploidy?

hermaphroditism
(Reinboth, 1962,
1,970)
Hap~oehromis
gonochoristic 1 2 0"0"
burtoni 37 Cj1Cj1 + +
Page nus
aearne
protandric 50"0" + +
Page nus ? (possibly 30"0" + +
erythrinus protogynous) 3Cj1 Cj1
Dip~odus ? (possibly 2 + +
sargus protandric) H
Spondy liosoma
protogynous 1 Cj1
eantharus
Coris protogynous
9Cj1 Cj1
julis 50"0" + +
(Prim.)
1 50"0" + +
(Sec.)
Serranus simultaneous
eabrina hermaphrodite + +
The results obtained for the gonochoristic cichlid teleost, H. burtoni,

proved extremely accurate, confirming the sexes diagnosed from gonad
examination. A single, large round Barr body was present in most female
liver nuclei (apposed against the nuclear membrane), which was gener-
ally absent from male liver nuclei (Figs. 1 and 2). The pattern was
essentially the same as shown for the sparid L. lithognathus (Mehl, 1973,
Figs. 24 and 25). Thus, it seems a reasonable assumption that H. burtoni
(like L. lithognathus) would also seem to have a typically mammalian-
like sex chromosome constitution of XX, XY (or XO). Hackmann and Rein-
both (1974) have also shown that the cichlid fish, Hemihap~oehromis mu~ ti-
eo~or (Hilgendorf), probably possesses male heterogamety and female
homogamety.
The capacity of thionin to stain especially for sex-chromatin, making

for easy identification of Barr bodies (Barr, 1961) would thus seem
to make it a valid statement that the stained body obtained from the
female liver nuclei. of both L. lithognathus and H. burtoni is indeed sex-
chromatin. This premise is essential when the results from the six
hermaphroditic species are analysed.
Leaving aside the simultaneous hermaphroditic serranid, Serranus eabrina

(which may be regarded as a special case), it can be seen from Table 1
that the male liver nuclei were all sex-chromatin positive while the
female liver nuclei were all sex-chromatin negative. On the basis of
245
what is known from man and other mammalian species, the above results
are indeed surprising. Admittedly, the sample number (except for C. juZis)
is small and thus, some reservation must be placed against the validity
of the data. However, in all cases the distinction between the two tis-
sues was very clear. Therefore, it would appear that in these hermaph-
roditic species at least, the male is the homogametic sex and the female
the heterogametic sex.
To ascribe a sex-chromosome constitution .(as has been tentatively done

for L. Zithognathus and H. burtoni) to these hermaphroditic species, is
clearly not yet possible from the results presented above. If it is
to be assumed that the sexual dimorphism displayed in the liver nuclei
of these hermaphroditic species does follow the typical mammalian pat-
tern (and the possibility cannot be over-ruled) then one runs into
serious problems of interpretation. The most striking of which is to
explain how a secondary male C. juZis, which is a naturally occurring
sex-inverted female (Reinboth, 1962), apparently has the ability to
produce a sex-chromatin body. It is possible that the sex-inversion
mechanism, which presumably is genetically controlled, may have re-
leased a previously inhibited X-chromosome. Barr (1961) reports that
sex-chromatin spots have been recorded by various authors from known
heterogametic female insects and possibly from birds also. Barr also
recorded that Ohno and his co-workers found a sex-chromatin mass in
the liver and skin interphase cells of the normal heterogametic female
fowl embryo, which was absent in the male homogametic sex. The possi-
bility that the sex-chromatin body could be formed from only one X-
chromosome cannot be overlooked. Moreover, the Y-chromosome may also
be able to form Barr bodies, as is the case in some Lepidoptera and
snakes (Mittwoch, pers. comm.). There may be grounds, therefore, for
assuming that similar situations could also apply for ambisexual tele-
osts. However, until it is reliably known what the genetic sex-deter-
mining mechanism of an hermaphroditic teleost species is, such a theory
remains pure conjecture.
Serranus eabriUa is a known simultaneous hermaphrodite, producing ripe

sperms and ova in its ovotestes (Reinboth, 1962). The positive inci-
dence of sex-chromatin in all 14 specimens examined again suggests
that this species is homogametic. However, absolutely no information
exists on the sex-determining chromosomes of simultaneous hermaphro-
ditic teleosts and until such information becomes available, it is
unwise to make even the slightest prediction concerning S. eabriUa.
The only direct methods of identifying heterogametic and homogametic

sex is by cytological observation of the chromosomes, studies of sex-
linkage or by breeding from sex-reversed animals (Beatty, 1964). Male
homogamety and heterogamety is already known from the same populations
of some teleosts, e.g. PoeeiZia (Lebistes) retieuZatus, Oryzias Zatipes and
Xiphophorus maeuZatus. These results are based on genetical studies of sex-
linked inheritance of color factors and on breeding experiments using
hormonally sex-reversed fish. The vast data available in this field,
by such famous reserchers as Aida, Winge, Gordon, Yamamoto, etc. are
well known and have been carefully reviewed by Beatty (1964), Atz
(1964), Yamamoto (1969) and Mittwoch (1973). Male homogamety and het-
erogamety has also been recorded in some species of Lepidoptera, Amphi-
bia, Aves and Reptilia (Beatty,' 1964). However, the question of whether
there is any direct connection between the color factors and the pro-
cess of sex differentiation in fishes still remains unanswered (Mitt-
woch, 1973). Mittwoch records that a number of species of fishes are
known to have chromosomal mechanisms of sex determination which, how-
ever, are still in a relatively undifferentiated state. Various wor-
kers have ascribed sex-linked inheritance of color factors to x- and
246
Fig. 1 - 4
247
Y-chromosomes even though the sex chromosomes were not distinguishable

in cytological preparations.
Atz (1964) stated that fish chromosomes have proved unfavorable cyto-
logical material, as they were small, little differentiated and numer-
ous. Recently, however, techniques for examining karyotypes of teleost
fishes have been greatly improved (McPhail and Jones, 1966; Howell,
1972), and clearly it is now imperative to try and determine the ka-
ryotypes of hermaphroditic teleost species. There is a good chance
that studies of the sex chromosomes of fishes will yield pertinent
information on the part played by the sex chromosomes in the process
of sex differentiation (Mittwoch, 1973). Hopefully, such future in-
formation would do much to elucidate the problems presented by the
sex-chromatin data of this paper.
All the male liver tissue of the ambisexual species examined occasion-
ally displayed the possible presence of more than one Barr body. The
incidence was highest in Sen'anus cabrilla, but was also plainly evident
in Pagellus erythrinus and Caris julis (Figs. 3 and 4). These observations
indicate a tentative case for polyploidy.
It is possible that errors may have been made in the interpretation

of the stained chromatin nuclear material. Barr (1961) reported that
although the Barr body in man and mammals is generally plano-convex
and lying apposed to the nuclear membrane, it may also be round, tri-
angular, irregular or discus-shaped. Barr even records that Barr bodies
have been determined from neurones of cows but not from liver cells,
as the chromatin material was too large. Mittwoch (1973) records from
various authors that the incidence of Barr bodies increases as the
cells become more crowded and their nuclei become smaller. Older cells
are more likely to contain Barr bodies than younger ones while, in
rapidly dividing tissues, the incidence of Barr bodies is lower than
in cells in which mitoses have become rare.
Barr (1961) states that it is rare to find more than one Barr body in
a cell which is normally diploid. However, in tetraploid cells and
especially those in which polyploidy is the rule, two chromatin bodies
may be found. Especially interesting is the fact that this situation
has been found in the liver cells of man and rat (Barr, 1961). Barr
also records that a sex-chromatin mass has been observed in the female
liver and skin interphase cells of heterogametic fowl embryos, which
was absent in the homogametic male sex.
It is unknown whether teleost liver tissue is diploid or polyploid.

However, the fact that there was positively no sign of more than a
single Barr body in the female liver cells of both L. lithognathus (Mehl,
~ Fig. 1. Nuclei of male liver tissue of the gonochoristic Haplochromis

burtoni, showing absence of sex-chromatin boaies
Fiq. 2. Nuclei of female liver tissue of H. burtoni, showing typical
solitary sex-chromatin bodies (arrowed) apposed against the nuclear
membrane
Fig. 3. Nuclei of liver tissue from a male specimen of Pagellus erythrinus,
showing the presence of two sex-chromatin bodies (arrowed) in a single
nucleus
Fig. 4. Nuclei of liver tissue from a simultaneous hermaphroditic
specimen of Serranus cabrilla, showing one nucleus containing three sex-
chromatin bodies (arrowed)
248
1973) and H. bUX'toni would suggest that the presence of multiple Barr
bodies in the examined ambisexual teleosts is a genuine case for an
abnormal (polyploid) sex-chromosome constitution. In humans, true
hermaphroditism can occur and it is a well documented fact that chro-
mosome abnormalities in patients displaying Klinefelter and Turner
Syndromes can be diagnosed by analyses of sex-chromatin (Mittwoch,
1973). Perhaps then, it is not unreasonable to assume that hermaphro-
ditic teleosts may fall into·a similar category.
Obviously, it will be necessary to examine some other somatic tissues

of these ambisexual teleost species for sex-chromatin. However, at
present it may be suggested that the apparently homogametic males of
the species in question do have an unusual sex-chromosom constitution.
References

including Man (eds. C.N. Armstrong and A.J. Marshall), pp. 145-232.
Barr, M.L.: Das Geschlechtschromatin. In: Die Intersexualitat (ed.
C. Overzier), pp. 50-73. Stuttgart: Georg Thieme 1961.
Barr, M.L., Bertram, E.G.: A morphological distinction between neurones
of the male and female, and the behaviour of the nucleolar satellite
during accelerated nucleoprotein synthesis. Nature (Lond.) 163,
676-677 (1949). -
Beatty, R.A.: Chromosome deviations and sex invertebrates. In: Inter-
sexuality in Vertebrates including Man (eds. C.N. Armstrong and
A.J. Marshall), pp. 17-143. London: Academic Press 1964.
Hackmann, E., Reinboth, R.: Delimination of the critical stage of
hormone-influenced sex differentiation in Hemihap!ochromis rrruUicolor
(Hilgendorf) (Cichlidae). Gen. Compo Endocrinol. ll, 42-53 (1974)
Howell, W.M.: Somatic chromosomes of the black ghost knifefish, Aptero-
notus albifrons (Pisces: Apteronotidae). Copeia 1.2.2£, 191-193.
McPhail, J.D., Jones, R.L.: A simple technique for obtaining chromo-
somes from teleost fishes. J. Fish. Res. Bd Can. 23, 767-769 (1966)
Mehl, J.A.: Ecology, osmoregulation and reproductivelbiology of the
whi te steenbras, Lithognathus lithognathus (Teleostei: Sparidae).
Zoologica Africana 8, 157-230 (1973).
Mittwoch, U.: Genetics-of sex differentiation. New York: Academic
Press 1973.
Labridae). Zool. Jb. (Physiol.) 69, 405-480 (1962).
Reinboth, R.: Intersexuality in fishes. Mem. Soc. Endocrinol. 18,
515-543 (1970). -
Yamamoto, T.: Sex differentiation. In: Fish Physiology (eds. W.S. Hoar
and D.J. Randall), Vol. 111, pp. 117-175. New York: Academic Press
1969.
Sex Determination and Differentiation among Uniparental
Homozygotes of the Hermaphroditic Fish Rivulus marmoratus
(Cyprinodontidae: Atheriniformes)
R. W. Harrington, Jr.
A. Introduction
The current status of knowledge and theory pertinent to sex determi-

nation and differentiation in teleost fishes was reviewed in advance
(Harrington, 1974). The review is critical and selective but based on
an inclusive examination of contingent, derivative, and source liter-
ature. It provides a cogent introduction to this study, which is a
search at the morphogenetic level for the earliest visible manifesta-
tions of the differentiation of ovotestes versus primary testes in
fish of the same genotype. The present report begins with a retrospec-
tive summary of intersexuality in RivuZus marmOl'atus Poey and of the evi-
dence for the genetic uniformity of our material and its consequences.
The new information appropriate to the title of this paper has its
context in findings published during the last 14 years by the writer
and his co-authors.
B. Intersexuality
R. marmoratus naturally reproduces uniparentally and is represented in

the wild of Florida by genetically-different groups of homozygous,
isogenic individuals (Kallman and Harrington, 1964; Harrington and
Kallman, 1968). These three attributes are shared by no other species
of fish so far as known. Self-fertilization is internal (Harrington,
1961). The eggs are incubated intraparentally, for from less than one
hour (1-cell stage) to as long. as 2 1/2 days (stage of pectoral fin
buds), then extraparentally, for about 14 days or longer (Harrington,
1963, 1967). R. marmOl'atus has not been subjected to artificial intro-
gressive hybridization. ~he first specimens kept in aquaria were wild-
caught in Florida, two years after R. marmOl'atus was first reported from
Florida (Harrington and Rivas,1958) and the Netherlands Antilles
(Hoedeman, 1958). Descendants of one of these fish are now, 14 years
later (1974), in their 27th uniparental laboratory generation. Within
the range of temperatures at which our laboratory stock lays eggs,
selfing produces an unbroken succession of uniparental hermaphrodite
generations.
Selfing in the wild had already reduced the fish within each of the
three lines of our stock to genetic uniformity, viz. isogenicity plus
homozygosity. These lines are referred to as homozygous clones. This
usage of the term "clone" avoids repeated circumlocution and means a
group or line of descent composed of uniparental, isogenic individuals.
Clones DS and NA comprise the descendants of two genetically-different
hermaphrodites, wild-caught in the same salt marsh. Clone M comprises
the descendants of a hermaphrodite captured 145 miles (233 km) to the
south. Perhaps because of their homozygosity, the fish of these clones
show, especially in response to temperature, a developmental plasticity
250
as regards sex phenotype and meristic traits (numbers of vertebrae and

fin rays) unequaled by any other fishes studied experimentally (Harring-
ton, 1971; Lindsey and Harrington, 1972).
When incubated at temperatures below those at which their selfing par-

ents lay eggs, a majority of eggs yield primary males (Fig. 2) instead
of hermaphrodites (Fig. 1). Exposure to low temperature directly after
hatching is too late to deflect sex differentiation to the primary male
phenotype from its usual goal, the hermaphrodite phenotype (Harrington,
1967, 1968). In R. marmoratus sex determination and differentiation
coincide. These two terms will be used interchangeably. The thermo-
labile interval of sex differentiation begins after stage 31b (formed
neural and hemal arches). Although the thermolability of sex differen-
tiation ends before hatching, this interval of development cannot be
delimited precisely because the time to hatching (incubation period)
varies and exposure to low temperature not only changes the direction
of sex differentiation, from hermaphrodite to primary male, but retards
over-all growth and development, and doubtless prolongs the thermo-
labile period itself (Harrington, 1968). Exposure of eggs to low tem-
perature beyond the potential duration of the temperature sensitive
period ensures a maximum yield of primary males. If this is not done,
the percentage of primary males obtained does not measure accurately
the clone-specific (genotype-specific) degree of thermolability in
primary sex differentiation (compare A and B, Table 1). The over-all
critical interval of sex differentiation may be said to extend from
earliest formation of gonadal primordia to outset of definitive gonadal
sex (Harrington, 1974). The lability of sex differentiation to extrin-
sic factors such as low temperature (Harrington, 1967, 1968) or hetero-
typic hormones (Yamamoto, 1953, 1958, 1968; Hackmann and Reinboth,
1974) lies within this interval. The same interval is examined here,
in search of differences in germ cell mitosis and in the results of
gonadal somatic mitoses correlated with the production of ovotestes
(hermaphrodites) at moderate temperatures and of primary testes (pri-
mary males) at low temperatures.
Sex succession in R. marmoratus is not a concern of this study, but

must be described so that it can be dismissed from further considera-
tion. It occurs after primary sex differentiation and is confined to
individuals (hermaphrodites) with amphisexual gonads (ovotestes). The
ease with which it can be induced in R. marmoratus depends on the geno-
type (clone). Exposure of hermaphrodites early in life to high temper-
ature sometimes causes a precocious sex succession that produces a
secondary male at first sexual maturity (= a false male gonochorist).
Secondary males produced by sex succession from either juvenile or adult
hermaphrodites can be distinguished from primary males (= true male
gonochorists) by vestigial oviducts (Fig. 3), as was first demonstrated
in protogynous fishes (Reinboth, 1962), and also by relict oocytes,
neither of which are found in gonads of primary males. Sex succession
in adult hermaphrodites may occur after a few months or several years
of selfing. It is triggered by short days in fish previously exposed
to high temperature. At the gonadal level it consists of a total in-
volution of ovarian tissue with an enlargement of testicular territory,
which changes ovotestes into secondary testes and selfing hermaphro-
dites into secondary males. Sex succession in R. marmoratus and in proto-
gynous hermaphrodites ends with secondary testes in secondary males.
Fig. 1. Ovotestis (young, adult). Testicular tissue at the lower left ~
Fig. 2. Testis of young primary male

Fig. 3. Secondary testis by sex succession from adult hermaphrodite
251
100lJm I
,
Fig. 1 - 3
252
However, the amphisexual gonads of protogynous hermaphrodites function

first as ovaries and later as testes, while R. mar-moratus functions from
the outset of sexual maturity as a selfing hermaphrodite and becomes
a secondary male by the subtraction of female structure and function.
R. mar-moratus has been designated (Harrington, 1971) a synchronous her-
maphrodite with facultative diandry (cf. Reinboth, 1967, 1968, 1970)
and protogynoid sex succession.
Secondary males are unknown in the wild. In the Netherlands Antilles

hermaphrodites predominate; males are not uncommon but do not seem to
be produced every year; they occasionally mate with hermaphrodites,
thus reducing the probability that homozygous lines of descent exist
in the Antilles as they do in Florida (Kristensen, 1965). A single
primary male has been found in the wild of Florida, where they are
probably rare, because incubation temperatures yielding primary males
are below those permitting egg laying and uniparental lines tested so
far are homozygous. Whether or not interline hybridization occurs in
the wild of Florida is unknown, but the possibility cannot be dismissed
because of already observed physiological differences even among our
three clones (Tables 1 and 2). There seem to be no females of this
species. Hermaphrodites, ~rimary and secondary males have been obtained
from eggs of the same genotype (clone) by the manipulation of environ-
mental factors. Juvenile fish have a caudal ocellus, persistent in
hermaphrodites, but obsolescent or lacking in males of both types.
Hermaphrodites have prominent body markings and barred fins. These
are retained by secondary males, which acquire in addition orange
spots and an orange wash. Primary males lack body markings and fin
bars. They are uniformly bright orange, except for the ventral margin
of the caudal fin, which is black.
Table 1. Primary males of RivuZus mar-moratus obtained by incubation at

below 20°C. (A) throughout the thermolabile period; (B) into stage 31b
but not with certainty throughout the thermolabile period
(A) Clone DS 67% n 30 (Harrington, 1967) (B) Clone DS 37% n 43

67% n 107 (Harrington, 1968)
Clone NA 88% n 17 (Harrington, 1967) Clone NA 52% n 50
Clone M 100% (all cases so far) Clone M 81% n 47
Table 2. Mean age in days postsyngamy and mean standard length in

millimeters at earliest functional sexual maturitya of 404 RivuZus
mar-moratus reared at 25 ± 1°C from date of dechorionation and transfer
from the incubation temperature
Sex Incubation Clone DS Clone NA Clone M

type temperature Days SL N D~ys SL N Days SL N
31 ± O.l°C 108 24.4 40 110 22.9 37 205 27.2 44

25 ± O.l o C 107 22.6 47 103 22.5 52 162 25.5 45
19 ± O.l°C 119 22.1 26 104 21.4 26 158 24.5 9
84 19.2 16 88 19.9 24 139 22.6 38

aHermaphrodite: first egg laid; primary male: first nuptial color.
253
C. Genetic Uniformity
Apart from presumptive homozygosity inferable from its mode of repro-

duction, the evidence for genetic homogeneity within uniparental lines
of R. marmoratus was obtained by tissue transplantation tests, alone and
combined with interclonal hybridization. Within each clone: grafts are
accepted freely; parent-to-offspring; offspring-to-parent; and sib-to-
sib; interline grafts are rejected both ways (Kallman and Harrington,
1964; Harrington, unpublished). Tissue transplantation is the most
sensitive test for homozygosity (Billingham and Silvers, 1959). Detailed
biochemical analysis of five enzymes and a preliminary survey of eight
others (Massaro et al., in preparation) revealed no consistent differ-
ences either within or among Clones OS, NA, and M. Advantage was taken
of the chance occurrence of a Clone-OS hermaphrodite that consistently
emitted unfertilized eggs to obtain interclonal (OS x NA) hybrid Fl
fish, one of which by selfing, produced an F2 generation. A graft from
a fish of the sperm-supplying clone (NA) into an Fl hybrid and grafts
from interclonal Fl and F2 hybrids into fish of both clones, as well as
grafts between F2 sibs, gave a pattern of autograft and allograft reac-
tions (Harrington and Kallman, 1968) that rigorously confirm the homo-
zygosity of Clone NA and, by implication, of other such clones. In sum,
the present study utilizes vertebrate material as genetically uniform
as is ever likely to be found, a uniformity acquired naturally in the wild
The intersexuality and genetics of R. marmoratus, their manifestations

and implications are considered at length from several Viewpoints in
the discussions of Harrington (1971) and of Lindsey and Harrington
(1972) .
D. Material and Methods
The new findings reported here are based on examinations of unbroken

series of cross sections through the entire body, from behind the exits
of the excretory' ducts usually at least to the level of the pronephroi.
A series of developmental stages were studied, from those in which germ
cells have not yet entered the genital folds to those in which the fish
are sexually functional. In a&l, 248 fish were examined, 100 (all of
Clone OS) incubated at 26 ~ 1 C, a temperature at which gnly hermaphro-
dites are obtained, and 148 incubated at within 19 ~ 0.8 C (113 of
Clone OS, 35 of Clone M). Most of these had been sectioned when it was
learned that Clone-M eggs yield 100% primary males at low temperature,
while Clone-OS eggs yield about 67% (Table 1). The cost of replacing
the Clone-OS series with a comparable series of Clone-M sections was
too great. Instead, 35 Clone-M fish of stages spanning the, by-then-
better-known, critial interval of sex differentiation were used for
comparison with Clone-OS fish of the same stages. Clone-M fish reach
functional sexual maturity as primary males much later than Clone-OS
fish (Table 2), but the fish of both clones were found to reach earli-
est sexual differentiation as primary males at the same sizes. Compar-
isons of Clone-OS and Clone-M fish at stages with Fresumed covert
sexual differentiation of the male gonad revealed a somatic feature
in putative primary testes that is the precursor of overt sexual dif-
ferentiation, as evidenced later by the sexualization of the germ
cells.
Gonadal differentiation was measured along the scale of progressive

over-all development and growth, marked by embryonic stages before
254
hatching and by standard lengths after hatching. Through the early

critical stages the gradation was in tenths of millimeters. Through
later stages it was seldom more than a millimeter. The ages of all
specimens, except 36 Clone-DS fish, were known to within one day, but
a morphogenetic scale is' more appropriate than a temporal scale because
temperature fluctuations and unknown factors accelerate the growth and
development of some fish and retard those of others. The data of Fig. 4,
however, when plotted against time give a similar pattern and lead to .
the same conclusions, but the pattern is less focused and coherent.
With this reservation, results will sometimes be compared with refer-
ence to both morphogenetic and temporal scales.
The fish were fixed in Bouin's solution. Their sections were 10 microns
thick and stained with iron haematoxylin. Only the best fixed and
stained series were used for germ cell counts.
E. Germ Cell Counts from Onset of Mitosis to Onset of Meiosis
Germ cell mitosis begins after the formation of neural and hemal arches
(Stage 31b+). It begins at about the same stage regardless of incuba-
tion temperature, but with regard to time it is postponed at low tem-
perature. Even with regard to developmental stage, ~rimordial germ
cell numbers persist longer in fish incubated at 19 C (Fig. 4). Between
fish incubated at 26°C and 19°C the rate of germ cell mitosis differed
sharply, whether the germ cell count was plotted against developmental
stage (Fig. 4) or time (days postsyngamy). At 26°C germ cell counts
rise steeply within the range of standard lengths at which hatching
usually occurs (4.5 - 5.4, average, 4.8 mm, n = 63). Maximal counts
are reached when the first gonia are sexualized as oocytes in first
meiotic prophase, with some in the earliest perinucleolar stage (cf.
Yamamoto, 1956). Germ cell counts in the fish at 26°C conform to a
curve with an upward flexure of 90 degrees, at about 4.5 mm SL and
30 days postsyngamy. All 26°C fish larger than 5.3 mm or older than
30 days had sexualized (female) gonia. Germ cell counts in fish incu-
bated at 19°C conform to a curve with an upward flexure at about the
same size and age as the 26°C fish but with an angle of about 45 de-
grees. Germ cell counts of 19°C fish approached the maximal counts
attained by the 26°C fish at larger sizes and greater ages than the
26°C fish. The 11 primary males used for germ cell counts (Fig. 4)
were on the threshold of'sexual differentiation. Only fish with un-
mistakeable primary spermatocytes and adjacent first meiotic meta-
phases, or with minimal numbers of spermatids but no meiotic phases
between primary spermatocytes and spermatids, were used for the termi-
nal germ cell counts. This procedure allowed an accurate germ cell count
at the threshold of sexual differentiation, because each tetrad of
spermatids could be counted as the last cell of a mitotic line. Sexu-
ally undifferentiated fish represented by open circles could be ex-
pected to differentiate as hermaphrodites, but 33% or less of the
undifferentiated fish represented by solid circles could also be ex-
pected to differentiate as hermaphrodites, despite incubation at 19°C
(Table 1). The two fish represented by the solid circles nearest the
uppermost two hermaphrodite symbols may have been future hermaphro-
dites, in spite of their incubation at low temperature. The 11 males
on the threshold of sex differentiation were 6.5 - 8.0 mm SL and 76 -
96 days old. The smallest male at sexual differentiation was 2 mm
larger and 48 days older than the youngest and smallest hermaphrodite
to reach sexual differentiation. The largest male on the threshold of
sexual differentiation was 3.5 mm larger and 66 days older than the
255
tardiest hermaphrodite to have sexualized female gonia. Several fish

incubated at 19 0 e and within the same range of sizes as the males of
Fig. 4 were still undifferentiated sexually.
Nr-----------------------------------------------------~
800
f i'
t
i'
i"
600 Incubation i"
o 26°C i'
• 19°C i"
i'
t J'
i'
400 i'
.
o. ..
! •
200 8
: DO:.;
a
..
~
• 'r i, .II:!.-
• i '"
~
i . .I I I I I I I I
Stage 27 28 29 30 31A 31 B 5 6 7 MMSL
Fig. 4. Germ cell counts on homozygous isogenic embryos and hatchlin~s

of Rivulus marmoratus . Incubation at moderate temperatures, e.g. 26 ± 1 e
(open circles), yields 100% hermaphrodites, the usual sex phenotype.
At or below 19 ± 0.08°e (solid circles) at least 67% primary males are
obtained. Plain circles refer to sexually undifferentiated fish. Male
or hermaphrodite symbols refer to fish at earliest sexual differentia-
tion. Standard length at eclosion: 4.0 - 5.4 (mean, 4.8) rnrn. Embryonic
stages: 27, rounded pectoral fins; 28, peritoneum pigmented; 29, cir-
culation in pectorals; 30, rays in caudal; 31a, swimbladder; 31b,
neural and hemal arches
These germ cell counts on presumptive and differentiated hermaphrodites
and primary males are the first to be reported on a hermaphroditic fish,
so that they can only be compared with counts on presumptive females
and .males of gonochoristic fishes. Hardisty (1967) cites evidence for
several teleost fishes of either or both an earlier onset and faster
rate of germ cell mitosis in future females than in future males.
Recently, germ cell counts were made on Oryz ias latipes (Satoh and Egami,
1972; Quirk and Hamilton, 1973) before, during and after hatching,
but apparently these counts were not continued long enough to include
counts on males at the threshold of sex differentiation. Neither pair
of authors reported an upward flexure in the curve of germ cell counts
in putative and differentiating males, such as found in R. marmor at us .
Both pairs of authors found an earlier and more rapid rate of germ
cell mitosis in future and differentiating females, comparable to that
in future and differentiating hermaphrodites of R. marmoratus. Onitake
256
(1972) also affirms earlier germ cell mitoses in females of Oryzias

latipes, but gives no counts. Satoh and Egami recorded a long interval
of gradually increasing germ cell counts before hatching, but Quirk
and Hamilton found instead, a prolonged rather constant germ cell
count, similar to that in R. marmoratus during the greater part of the
prehatching period.
F. Gonadal Morphogenesis
The differentiation of primary testes (testes of primary males) of

R. marmoratus resembles that of testes of gonochoristic teleosts. Even
the much discussed testis ova (Atz, 1964; Yamamoto, 1969) were repre-
sented by only one testis ovUm, in one section of one differentiated
male found after a search th~oughout the complete serial sections of
the 148 fish incubated at 19 C. Long before its gonia are sexualized,
a future primary testis can be recognized by the conspicuous enlarge-
ment of its hilar stroma (Fig. 5, Column 1) as compared with its dis-
tal portion, which contai~s the germ cells. Starting at about 4.6 rom
SL, the size of the hilar strbma began increasing in a majority of
the gonads 9f Clone-DS fish and in all gonads of Clone-M fish incubated
at 19°C. This recalls conditions before germ cells are sexualized in
the gonads of all-male broods of the platyfish (Chavin and Gordon,
1951). The developmental series of primary testes of R. marmoratus
(Column 1 of Fig. 5) belonged to fish of the following standard lengths
(rom) and ages (days postsyngamy), top to bottom: 4.6 rom (28 days),
5.1 (35), 6.4 (61), 6.9 (134) and 18.3 (206). Only the last two had
sexualized germ cells. The smaller of these two fish was the youngest
and smallest male with complete spermatogenetic series.
It would be impossible to estimate the soma-cell:germ-cell ratio in
differentiating primary testes of stages later than those whose germ
cell numbers are plotted in Fig. 4. In any case, the sexualizing in-
fluence of somatic tissue on germ cells presumably begins some time
before the earliest sexualization of germ cells. Soon after this it
becomes impracticable to measure and summate the cross-sectional areas
of the testes occupied by hilar stroma versus germ cells, because by
the time a testis reaches the stage shown at the bottom of Column 1
of Fig. 5, the stromatic tissue has arborized from the hilar region
into the zone of differentiated male germ cells (shaded in Fig. 5).
Hardisty (1965) could recognize testes of Lampetraplaneri also by so-
matic features, e.g. a broader mesogonial stalk and a higher proportion
of somatic to germinal tissue, with greater development particularly
in the hilar region.
The gonads in Columns 1 - 4 of Fig. 5 are right gonads viewed from

the tail end toward the head end, so that the midline of the fish in
each case is to the left of the gonad. The distal ends of the primor-
dial gonads of R. marmoratus (not shown in Fig. 5) are first directed
ventrad, next ventrolaterad, then laterad, and finally the gonads com-
plete a rotation of almost 180 degrees, as exemplified by the lower-
most testis and by all ovotestes of R. marmoratus illustrated in Fig. 5.
The rotations of ovotestes are often completed before the first germ
cells are sexualized, as exemplified by the uppermost gonad of Column
2. The successive stages of development of the testes and ovotestes of
R. marmoratus were drawn from microprojections of histological sections.
All but the lowermost testis and the lowermost two ovotestes were drawn
to the same scale. The germ cell areas of the smallest and youngest
primary male (6.9 rom SL, 134 days old) and hermaphrodite (4.5 rom SL,
28 days old) with sexualized germ cells are virtually the same, which
257
1 3 4 5 6 7
orr
--
~
)a 0ir6
0:0
0j0 ~
OC)
t:;~ ~ ~
>< QO
....
9
Fig. 5. Comparison of gonadogenesis in Rivulus marmoratus, another her-
maphroditic fish, and two gonochoristic fishes. Black: Mesogonadium,
hilar stroma, and its outgrowth, the lateral ramus. Shaded: d' gono-
cytes. Stippled: ~ gonocytes. Clear: Undifferentiated germ cells.
Rivulus marmoratus: (1) primary testis; (2) ovotestis. Sparus auratus:
(.3) ovotestis, based on photographs of Pasquali (1941). Opsanus tau:
(4) ovary, based on photographs of Odum (1936). Poecilia reticulata:
(5) ovary; (6) testis; (7) testis changed into ovary by estrogen, after
Miyamori (1964) and Anteunis (1959)
calls attention to the massiveness of the hilar stroma of the testis

in contrast to the scarcely perceptible stroma in the same region of
the ovotestis.
There is uncertainty as to the exact point at which a germ cell is

sexualized (Reinboth, 1972; Harrington, 1974), but Vitagliano (1950)
considers a germ cell to be undifferentiated up to the pachytene stage.
Most of the germ cells of the hermaphrodite just mentioned were in
late pachytene, with two oocytes in the early perinucleolus stage.
The primary male just mentioned, was the first with spermatogenetic
series complete through to spermatids. The cross-sectional area of the
zone of differentiated germ cells in each of these two fish was 1/14
that of the zone of differentiated male germ cells in the lowermost
ovotestis of R. marmoratus in Fig. 5. When projected at the same magni-
fication, the entire cross-sectional areas of the lowermost testis
and ovotestis of R. marmoratus (Fig. 5) were nearly equal. The primary
male (18.2 rom SL, 206 days old) was closer to functional maturity
(cf. Table 2, Clone OS) than the hermaphrodite (15.2 rom SL, 122 days
old) .
258
Hilar stroma first becomes conspicuous in developing ovotestes long

after the oocytes have reached the perinucleolus stage (at about 10 mm
SL), but is already conspicuous in primary testes of fish as small as
4.5 mm SL (uppermost testis of Fig. 5). Soon after its proliferation
in ovotestes, it gives rise to a fast growing lateral ramus (black in
the lowermost four ovotestes of Fig. 5, Column 2), to use the terminol-
ogy of Pasquali (1940/41). The distinction between the mesogonadium
(Okada, 1965) or mesogonial stalk (Hardisty, 1965) and the hilar stroma
and its outgrowth, the lateral ramus, is best seen in the fifth ovo-
testis from the top of Column 2 of Fig. 5. All have a common origin
in the most proximal and inconspicuous part of the hermaphrodite gonad
of R. marmoI'atus, as may be seen from a comparison of the upper three
ovotestes with the lower four ovotestes of Column 2, Fig. 5.
With the outgrowth of the lateral ramus an entovarial sulcus is formed

between lateral and medial rami. The male territory of the ovotestis
of R. maI'mOI'atus appears with the onset of spermatogenesis in the hilar
region at the base of the lateral ramus (sometimes also at its distal
end) just before or after the lateral and medial rami fuse at their
tips to enclose the entovarial canal or ovisac. No male territory was
found differentiated before 13.3 mm SL and 76 days, but the female
territory, which coincides with the medial ramus, differentiated as
early as 4.5 rom SL and 28 days. Despite this difference, male and fe-
male territories are coordinated from outset of functional maturity
and the first eggs emitted usually contain developing embryos.
Whether a fish differentiates as a primary male or hermaphrodite de-

pends on the incubation temperature (see above) and genotype (Table 1),
but the age and size of hermaphrodites at which the ovotestes reach
successive stages of development depend chiefly on the post-hatching
rearing temperature, as well as on the genotype (Table 2). The last
factor in male platyfish was elegantly identified with a pair of al-
leles that time onset of functional maturity by timing the appearance
of gonadotrops (Kallman et al., 1973). The developmental stages of
the ovotestes of R. marmoI'atus shown in Column 2 (Fig. 5) were those
of hermaphrodites of the following standard lengths and ages, top to
bottom: 4.2 mm (age unknown), 4.5 (28 days), 9.0 (28), 10.2 (37),
11.9 (44), 17.9 (108) and 15.2 (122).
G. Discussion
Allowing for the rotation of ovotestes in R. marmoI'atus, it is evident

from the gonadogenesis of protogynoid R. maI'moI'atus, protandric SpaI'us
aUI'atus, and gonochoristic Opsanus tau that the medial ramus comprises
the female territory of the gonad (Fig. 5, Columns 2 - 4) . The male
territory is associated with the most proximal somatic tissue of the
gonad, viz. the hilar stroma and its outgrowth, the lateral ramus. In
R. marmoI'atus, and presumably in protogynous fishes, the female terri-
tory (medial ramus) is the first to be differentiated sexually, while
in SpaI'us aUI'atus it is the male territory (hilar stroma and lateral
ramus). In Opsanus tau, the medial ramus becomes the female territory
and the lateral ramus becomes the sterile wall of the ovary, as in
other gonochoristic fishes. In the viviparous gonochorist Poecilia
I'eticulata (Fig. 5, Columns 5 - 7), the ovaries do not develop as ento-
varial canals but as parovarial canals, which fuse into an unpaired
ovary. Potentially male territory can be sought only in the hilar
stroma of the primordial gonads. Hilar stroma is prominent in testes
(Column 6) but is obliterated during ovarian differentiation (Column 5)
259
and in estrogen-induced conversions of testes into ovaries (Column 7) .

Whether the affinities in Fig. 5 can be matched by conditions in her-
maphroditic gonads that seem to lack preformed heterologous territories
(cf.Reinboth, 1970) remains to be determined by studies of closely
graded, critical, early stages of gonadogenesis. This question is re-
lated to the problem of the embryonic derivation of the, as yet, un-
identified foci from which senescent or teratological masculinization
of ovaries spreads in gonochoristic fishes. The similarity in the mor-
phogenesis of piscine ovaries and ovotestes was noted by'Pasquali (1940/
41) and suggests the potential amphisexuality of ovaries as contrasted
with primary testes.
In the primary sex determination (hermaphrodite versus primary male)

of homozygous, homogametic R. maY'l7loY'atus, low temperature usurps the
role usually ascribed to genic or chromosomal factors, and in the
theory of Mittwoch (1973, and references therein) assigned to the odd
chromosome of the heterogametic sex. In both the theory and the case
of R. maY'moY'atus, it is carried out at the morphogenetic level by the
control of mitotic rate. In R. maY'l7loY'atus incubated at low temperature,
whether over-all growth and development are slowed down with reference
to mitosis in the hilar stroma, or the reverse, is less important than
the fact that in presumptive primary males the hilar stroma is increas-
ingly prominent in proportion to the germ cell area of the gonad, while
in presumptive hermaphrodites it is hard to recognize at all (Fig. 5,
Columns 1 - 2). As already noted (Harrington, 1974), the comment of
Quirk and Hamilton (1973) that the theory of Mittwoch does not apply
to germ cells of Oryzias because of their precocious mitosis in the
homogametic sex is irrelevant, to the extent that it conflicts with
the widely accepted inductive interpretation of sex differentiation,
which stresses the sexualization of germ cells by the somatic substrate.
In the case of primary males of R. maY'l7loY'atus, if one accepts the primacy
of the somatic substrate in the sexualization of germ cells, the hilar
stroma would seem to be implicated in retarding the sexualization of
germ cells as male cells and perhaps, before that, in slowing the rate
of germ cell mitosis in presumptive males (Fig. 4). This, of course,
could imply an autonomous mitosis and sexualization of germ cells in
hermaphrodites, but that is beyond the scope of this report. The im-
portance of the hilar stroma is evidenced further by the fact that it
later becomes conspicuous in the ovotestes, soon after which it sends
out an offshoot, the lateral ramus, at the base of which the male ter-
ritory becomes manifest with the sexualization of male gonocytes (Fig.5,
Column 2, lowermost three gonads) .
These considerations reopen the question of a unitary primordium versus

dual primordia in the development of teleost gonads. D'Ancona (1940/41)
denied the existence in teleost gonads of a cortex and medulla such as
described in higher vertebrates (Witschi, 1957, and references therein)
but had difficulty in rationalizing the presence of discrete male and
female areas in ovotestes (D'Ancona, 1949), and finally denied their
occurrence in phylogenetically primitive families, specifically in the
one to which R. marmoY'atus belongs (D' Ancona, 1955). Gropp and Ohno
(1966) claim that cattle have a unitary gonadal primordium, and Ohno
(1967) considers the fate of the indifferent gonad to be decided by
the direction taken by a unitary somatic blastema, instead of by compe-
tition between hypothetically antagonistic cortical and medullary pri-
mordia. Hardisty (1965) takes the view that the distinction between
dual and single primordia is blurred by the close spatial relationship
between interrenal and gonadal elements and by the succession of neph-
ric, adrenocortical and gonadal elements, derived from successive dif-
ferentiations of the peritoneum. D'Ancona seems to have been justified
in denying explicitly cortical and medullary areas in teleost gonads,
260
but ironically, perhaps their counterparts are the medial and lateral
rami described by Pasquali and D'Ancona.
The identification of the medial ramus with female territory and the
lateral ramus, or perhaps more basically, the hilar stroma, with male
territory does not explain why in protogynous (and protogynoid) herm-
aphrodites the medial ramus differentiates first, while the reverse
is true of protandrous hermaphrodites. Perhaps the speculation (Har-
rington, 1971) that protandry may derive from male homogamety and pro-
togyny from female homogamety can be harmonized with the theory of
Mittwoch, which emphasizes in a new way the relation of homogamety
and heterogarnety to sexual differentiation.
References
Anteunis, A.: Recherches sur la structure et le developpement de

l' ova ire et de l' oviducte chez Lebistes reticul-atus (TeH~osteen).
Arch. de Biol. 70, 783-809 (1959).
Billingham, R.E., Silvers, W.K.: Inbred animals and tissue transplan-
tation immunity. Transplant. Bull. 6, 399-405 (1959).
Chavin, W., Gordon, M.: Sex determination in Pl-atypoeciZ-us macul-atus.
I. Differentiation of the gonads in members of all-male broods.
Zoologica (N.Y.) 36 (2), 135-145 (1951).
D'Ancona, U.: Ulteriori osservazioni e considerazioni sull'ermafro-
ditismo ed il differenziamento sessuale dell'orata (Sparus auratus L.) •
Pubbl. Staz. Zool. Napoli 18, 313-336 (1940/41).
D'Ancona, U.: Ermafroditismo ed intersessualita nei Teleostei. Expe-
rientia 5 (10), 381-389 (1949).
D' Ancona, U-:-: Osservazioni sulle gonadi giovanili di Amia cal-va. Arch.
Ital. Anat. Embriol. 60, 184-200 (1955).
Gropp, A., Ohno, S.: The:presence of a common embryonic blastema for
ovarian and testicular parenchymal (follicular, interstitial and
tubular) cells in cattle, Bos taurus. Z. Zellforsch. 74, 505-528
(1966). -
Hackmann, E., Reinboth, R.: Delimitation of the critical stage of
hormone-influenced sex differentiation in Hemihapl-ochromis muUicol-or
(Hilgendorf) (Cichlidae). Gen. Compo Endocrinol. 22, 42-53 (1974).
Hardisty, M.W.: Sex differentiation and gonadogenesiS:in lampreys
(Parts I and II). J. Zool. 146, 305-387 (1965).
Hardisty, M.W.: The numbers of -vertebrate primordial germ cells. Biol.
Rev. 42, 265-287 (1967).
Harrington, R.W., Jr.: Oviparous hermaphroditic fish with internal
self-fertilization. Science 134 (3492), 1749-1750 (1961).
Harrington, R.W., Jr.: Twenty-four-hour rhythms of internal self-fer-
tilization and of oviposition by hermaphrodites of Rivul-us marmoratus.
Physiol. Zoal. 36 (4), 325-341 (1963).
Harrington, R.W., Jr.: Environmentally controlled induction of primary
male gonochorists from eggs of the self-fertilizing hermaphroditic
fish, Rivul-us marmoratus Poey. Biol. Bull. 132 (2), 174-199 (1967).
Harrington, R.W., Jr.: Delimitation of the thermolabile phenocritical
period of sex determination and differentiation in the ontogeny of
the normally hermaphroditic fish, Rivul-us marmoratus Poey. Physiol.
Zoal. 41 (4), 447-460 (1968).
determine when self-fertilizing hermaphrodites of Rivul-us marmoratus
change into functional secondary males, with a reappraisal of the
modes of intersexuality among fishes. Copeia (3), 389-432 (1971).
261
Harrington, R.W., Jr.: Sex determination and differentiation in fishes.

In: Control of Sex in Fishes (ed. C.B. Schreck), pp. 4-12. Blacks-
burg, Virginia: Sea Grant and Virginia Polytechnic Institute and
State University Press (V.P.I.-SG-74-01) (1974).
Harrington, R.W., Jr., Kallman, K.D.: The homozygosity of clones of
the self-fertilizing hermaphroditic fish, Rivulus marmoratus Poey
(Cyprinodontidae, Atheriniformes). Amer. Nat. 102 (926), 337-343
(1968). -
Harrington, R.W., Jr., Rivas, L.R.: The discovery in Florida of the
cyprinodont fish, Rivulus marmoratus, with a redescription and ecolog-
ical notes. Copeia (2), 125-130 (1958).
Hoedeman, J.J.: Rivulid fishes of the Antilles. Stud. Fauna Cura9ao
Carib. Isl. ,!! (32), 112-126 (1958).
Kallman, K.D., Harrington, R.W., Jr.: Evidence for the existence of
homozygous clones in the self-fertilizing hermaphroditic fish,
Rivulus marmoratus (Poey). Biol. Bull. 126 (1), 101-114 (1964).
Kallman, K.D., Schreibman, M.P., Borkoski, V.: Genetic control of
gonadotrop differentiation in the platyfish, Xiphophorus maculatus
(Poeciliidae). Science 181 (4100), 678-680 (1973).
Kristensen, I.: Competition in three cyprinodont fish species in the
Netherlands Antilles. Stud. Fauna Cura9ao Carib. Isl. ~ (119),
82-101 (1970).
Lindsey, C.C., Harrington, R.W., Jr.: Extreme vertebral variation
induced by temperature in a homozygous clone of the self-fertiliz-
ing cyprinodontid fish Rivulus marmoratus. Can. J. Zool. 50 (6),
733-744 (1972). --
Massaro, E.J., Massaro, J.C., Harrington, R.W., Jr.: Biochemical com-
parison of genetically-different homozygous clones (isogenic, uni-
parental lines) of the self-fertilizing fish Rivulus marmoratus Poey
(In prep.).
Mittwoch, U.: Genetics of sex differentiation. New York-London: Aca-
demic Press 1973.
Miyamori, H.: Study on the morphogenic process of the estrogen-induced
feminization of male reproductive organs of Lebistes reticulatus.
J. Biol. Osaka City Univ. 15, 1-22 (1964).
Odurn, E.P.: Notes on the history of the germ cells in the toadfish
(Ops.anus tau). J. Elisha Mitchell Sci. Soc. 52, 235-246 (1936).
Ohno, S.: Sex chromosomes and sex-linked gene~ Monographs on Endo-
crinology. Berlin-Heidelberg-New York: Springer 1967.
Okada, Y.K.: Bisexuality in sparid fishes. I. Origin of bisexual gonads
in MyUo macrocephalus. Proc. Jap. Acad. 41, 294-299 (1965).
Onitake, K.: Morphological studies of normal sex-differentiation and
induced sex-reversal process of gonads in the medaka, Oryzias latipes.
Annot. Zool. Japon. 45 (3), 159-169 (1972).
Pasquali, A.: Contributo-allo studio dell'ermafroditismo e del differ-
. enziamento della gonade nell' orata (Sparus auratus L.). Pubbl. Staz.
Zool. Napoli 18, 282-312 (1940/41).
Quirk, J.G., Hamilton, J.B.: Number of germ cells in known male and
known female genotypes of vertebrate embryos (Oryzias latipes).
Science 180 (4089), 963-964 (1973).
Reinboth, R~Morphologische und funktionelle Zweigeschlechtlichkeit
bei marinen Teleostiern (Serranidae,. Sparidae, Centracanthidae,
Labridae). Zool. Jb. Abt. Zool. Physiol. Tiere 69, 405-480 (1962).
Reinboth, R.: Zum Problem der amphisexuellen Fische. Verh. Dtsch.
Zool. Ges. Heidelberg, 316-325 (1967).
Reinboth, R.: Protogynie bei Papageifischen (Scaridae). Z. Naturforsch.
23b (6), 852-855 (1968).
Reinboth, R.: Intersexuality in fishes. Mem. Soc. Endocrinol. 18,
515-543 (1970). --
Reinboth, R.: Hormonal control of the teleost ovary. Amer. Zool. Il,
307-324 (1972).
262
Satoh, N., Egami, N.: Sex differentiation in germ cells in the teleost
Oryzias Zatipes, dur ing normal ernbryoni c development. J. Ernbryol.
expo Morph. 28 (2), 385-395 (1972).
Vitagliano, G.:-osservazioni sul comportamento delle cellule follicu-
lari nella gametogenesi di CavoZinia tridentata Forskal (Moll. Pterop.)
Pubbl. Staz. Zool. Napoli 22, 367-377 (1950).
Hokkaido Univ., Ser. VI, Zool. (Prof. T. Uchida Jubilee Vol.) 11,
428-439 (1957).
Yamamoto, K.: Studies on the formation of fish eggs. I. Annual cycle
in the development of ovarian eggs in the flounder, Liopsetta obscUl'a.
J. Fac. Sci. Hokkaido Univ., Ser. VI, Zool. 12 (3), 362-373 (1956).
Yamamoto, T.: Artificially induced sex-reversal-rn genotypic males
of the medaka (Oryzias Zatipes). J .. Exp. Z 001. 123, 571-594 (1953).
Yamamoto, T.: Artificial induction of functional sex-reversal in geno-
typic females of the medaka (Oryzias Zatipes). J. EXp. Zool. llZ,
227-262 (1958).
Yamamoto, T.: Permanency (sic) of hormone-induced sex-differentiation
in the medaka, Oryzias Zatipes. Annot. Zool. Japon. !l (4), 172-179
(1968).
and D.J. Randall), Vol. 3, pp. 117-175. New York: Academic Press
1969.
Protogynous Hermaphroditism In Fishes of the Family
Scaridae
J. H. Choat and D. R. Robertson
A. Introduction
The demonstration of sequential hermaphroditism in shallow water marine

fishes is no longer a novelty. Initial studies of sequential hermaphro-
ditism served to focus attention on problems relating to sex determi-
nation, differentiation and the physiological mechanisms involved in
the change of sexual identities. Attempts to clarify the selective
advantage of sequential hermaphroditism, which is now recognized as
a widespread population phenomenon (Atz, 1964; Reinboth, 1970), have
not always kept pace with histological and physiological studies.
This paper is an amplification of arguments previously put forward

by Robertson and Choat (1973), concerning the selective advantages
ass8ciated with a protogynous hermaphroditic sequence in labrid fishes.
These arguments were based on the following assumptions. The opportu-
nity to breed is distributed unequally amongst males in a given popu-
lation, being restricted to larger individuals. Hales are promiscuous
and capable of fertilizing a large number of females. Under these cir-
cumstances there will be a selective advantage for males to be larger
than females and protogynous hermaphroditism may evolve. These argu-
ments are largely an extension of the size advantage model of Ghiselin
(1969). Implicit in such assumptions is the idea that competition for
gametes between individuals will be a powerful moulding force in the
evolution of hermaphroditism. This argument accomodates selection at
the individual level only. We believe that attempts to interpret se-
quential hermaphroditism in the context of zygote production, or ad-
justment of numbers of entire populations (Fishelson, 1970; Hoe, 1969;
Smith, 1967), will run into difficulties, especially where the orga-
nism concerned have highly dispersive larvae.
The above assumptions accommodated the situation in which a single

large male individual would dominate a group of smaller females, within
a site defended against other males. All male recruitment was assumed
to be via the sex-inversion of females from within the group. However,
as the studies of Reinboth (1962, 1970) have shown, this is simply one
aspect of a complex and variable pattern of sexual ontogeny. Therefore,
we also attempted to account for the phenomenon of dual patterns of
male identity, monandry and diandry (Reinboth, 1967), which is charac-
teristic of many species of labrids.
Similar patterns of sex-inversion and dual male identity occur in the

Scaridae, which are closely related to the Labridae, and are highly
characteristic of coral reef environments (Reinboth, 1968). live will
present evidence that the scarids also conform to the model developed
to account for protogynous sex-inversion in labrids. In doing so, we
emphasize that the expression of dual male identities is complex and
not really amenable to a straightforward classification of sexual
behavior and ontogeny.
264
In teleost fishes a protogynous sequence of sexual succession is the

dominant mode of hermaphroditism. However, any attempt to interpret
the significance of protogyny in fishes must deal with two sorts of
biases; (1) taxomonic biases, as protogynous fishes seem to be re-
stricted to particular groups, and (2) spatial biases, in that there
are more species of protogynous fishes (in terms of both relative and
absolute estimates) in coral reef waters than anywhere else. The prob-
lem as to why such biases exist is beyond the scope of this paper,
which is confined to a review of hermaphroditism in the Scaridae of
Heron Island, Great Barrier Reef. Our approach has been strictly cor-
relative. To resolve the questions considered, two sorts of experimen-
tal studies are required. Those in which monitoring and manipulation
of individuals under natural conditions playa major part (Robertson,
1972; Reinboth, 1973; Buckman and Ogden, 1973; Ogden and Buckman,
1973), and those in which the influence of environmental factors on
the ontogeny of sexual development may be assessed (Harrington, 1971)
B. Methods
The data presented here were collected mainly at two localities on a

patch reef at Heron Island (Lat. 23° 26'S, Long. 152 o E), a part of
the Capricorn group of reefs at the southern extremity of the Great
Barrier Reef. Collections were made in January and February, 1967, and
in the same months during 1972 and 1973.
To obtain data on the relative abundance of individuals of different

species, counts of individuals were made in 5 x 50 m and 10 x 100 m
transects. Measured areas of substrate were also observed for fixed
periods of time. Grazing intensity was assessed as the number of bites
taken by each individual from the substrate during this time. Individ-
ual scarids were followed for a fixed period of time, usually 10 min,
and all feeding activities and interactions were recorded. Estimates
of the relative abundance of the different grazing substrates were
obtained by running 10 metre line transects across the various areas
in which scarid observations were carried out. Large scale distribu-
tion of substrates was estimated from maps of the outer reef-slope,
covering approximately 4000 square metres.
For gonad analysis scarids were collected individually and from poison
stations. Attempts were made to cover the entire range of the species
observed when collecting. Gonads and gonoducts were either immediately
dissected out and fixed in Bouin's fluid, or the entire specimen was
fixed in 10% formalin by injection into the body cavity. Size was
measured in mm and expressed as standard length.
C. Results
The following species of scarids were recorded from Heron Island:

Bolbometapon bieolor, Calotomus spinidens, Searus rubroviolaseeus¥, S. gibbus*,
S. venosus*, S. forsteri¥, S. bleekeri, S. sordidus¥, S. flavipeetoralis¥, S. lunula*,
S. formosus¥, S. sehultzi*, S. ovieeps*, S. sexvittatus*, S. niger~ S. seaber,
S. ghobban*, S. ehlorodon*, S. lepidus, S. faseiatus*, S. globieeps* and one un-
descr ibed species of Searus.
Although it has been assumed that all of the above species are proto-
gynous, it has only been confirmed by histological examination in those
265
marked with an asterisk. The present investigation concentrated on

nine of the most abundant species in the study area. These are listed
as follows in order of local numerical abundance: S. sordidus, S. faseia-
tus, S. globieeps, S. niger, S. forsteri, S. venosus, S. formosus, S. lunula and
S. sexvittatus. It was not possible to distinguish between small indi-
viduals of S. faseiatus and S. globieeps under water. They are identified
as S. glob/fase.
All these species were diurnally active herbivores grazing on calcar-

eous substrates, and the vast majority were distributed within the
upper 12 m of the reef slope. The populations sampled at Heron Island
were members of species widely distributed in the tropical Pacific and
Indian Oceans and were thus near the southern limits of their range.
I. The Relationship between Color Phase and Size
The majority of scarids from the study area were dichromatic, and un-
like many species of labrid fishes, the dichromatism in the different
phases was sharply discontinuous.
Of the 22 species listed above, there were only 3 which did not fit
the picture of discontinuous dichromatism. Searus niger showed no dif-
ferentiation of color over the adult size range. In the case of Searus
sehultzi, larger individuals were more brightly colored but the dichro-
matism did not involve sharply distinct color patterns. Searus gibbus
did not fit the pattern of usual scarid dichromatism. Two color vari-
ants, red and green were observed each occupying the full size range,
and it is thought that these may represent a color polymorphism.
In the remaining 20 species, large individuals were more brightly

colored than small individuals. The former were usually characterized
by a green body coloration with intricate and distinctive patterns of
red, orange or blue markings on the head and opercula. The latter were
either grey, brown or reddish overall and usually lacked any distinc-
tive anterior markings. For the purposes of this study these phases
were called "gaudy" and "drab" respectively.
The distribution of color phase as a function of size in pine of the

most abundant species, is shown in Fig. 1. The lower size limits for
each color phase represent the size at which sexually mature individ-
uals were first detected by histological examination and spawning
observations. All nine species, with the exception of S. niger, are
dichromatic with sharply discontinuous color phases. In all instances
where it was possible to collect large numbers of individuals there
was considerable size overlap in the color phases, although gaudy phase
individuals always had a greater mean size than the drab phase ones.
Color transitional specimens were observed in most species, but were
never abundant. The data in Fig. 1 reflect this. In nearly all species,
distinctive juvenile color phases were observed.
There appear to be no predictable trends in the degree of overlap be-

tween colo~ phases. Field observations confirmed that the degree of
overlap in the color phases of S. formosus, S. lunula and S. forsteri would
increase if larger collections were made. The large range and variance
shown by the different color phases suggest that a considerable degree
of flexibility is involved in the process of transformation from one
color phase to another.
266
G nolO
S. formosus
I
o no18
I
T
I G nolO
S.lunula o n 012 I
G n, 42
S. globiceps
Dn,54
_ -_ _~oiiiG-.:.n ' 16
S. forsle ri
o n ,7
Tn,6
G n 099
S. sordidus
o n 097 I
T no 2
G n' 61
S. fas-ciatus
o n =84 I
T noS
S.sexviltatus G no 53
I G n ,13
S. v e nos U 5
o no 39
n = 69
S. n i 9 e r
o 50 100 150 200 250 300

Size in mm (Standard Length)
Fig. 1. The size distribution of color phases in nine species of sca-

rids. Lower size limits are estimates of the size at which sexual ma-
turity occurs. The vertical line represents the mean, the black bar
one standard deviation each side of the mean. (D) drab phase, (G) gaudy
phase, (T) transitional phase, (n) sample sizes in each phase
II. Sex Structure of Scarid Populations
The same specimens used to compile Fig. 1 were also examined to deter-
mine their sexual identities. Slight discrepancies in numbers between
Fig. 1 and Table 1 indicate that we were unable to examine all indi-
viduals histologically. The data in Table 1 show that scarids are very
close to the labrids in their patterns of sexual identity. Following
Reinboth (1962, 1970), four sexual identities were recognized: females,
primary males, secondary males and individuals undergoing sexual tran-
sition. Primary males were presu~ed to be true male gonochorists.
Secondary males comprised false male gonochorists, that is secondary
males from juvenile females (Harrington, 1971), and males derived from
mature females.
Primary males had large testes with centrally located sperm ducts.
The gonoducts of primary males were massive, with numerous connective
tissue septae. The testes of secondary males clearly indicated their
ovarian origin by being lobate, with a central lumen and peripheral
sperm ducts. The gonoducts were similar to those described for second-
ary male labrids (Reinboth, 1962). Arbitrary criteria were used for
defining transitional gonads, as functional females may have sperm
crypts and functional males, atretic eggs. However, no instances of
simultaneous maturity of male and female products in the same gonad
were found.
267
Table 1. Color and sex distribution in nine species of scar ids collected
at Heron Island. 1°0"0" refers to primary males, 2°0"d' to secondary males
and t to sexually transitional individuals. Sizes are grouped in 20 rom
intervals of standard length
(A) S. formosus
Color phase Drab Gaudy
Sexual identity 1 0 0"0"
61 - 80
81 - 100
1 01 - 120 2
1 21 - 140 8
1 41 - 160 6 3
1 61 - 180 5
1 81 - 200
Sexually mature 105 rom ?
(B) S. lunula
Color phase Drab Transitional Gaudy
Sexual identity t 2 0 ci'd'
61 - 80
81 - 100
101 - 120 2
121 - 140
141 - 160
1 61 - 180 5
181 - 200 5
201 - 220 2 2
221 - 240 3
Sexually mature 115 rom ?
(C) S. forsteri
Color phase Drab Gaudy
Sexual identity '('( 1°0"d' t 1 0 d'd'
61 - 80 12 4
81 - 100 4 4
101 - 120 3
1 21 - 140 3
141 - 160 2
161 - 180
181 - 200 3
201 - 220 2 2
221 - 240 3 3
Sexual maturity 120 rom ?

268
(D) S. gZobiceps
Sexual identity 'i' 'i' °
1 d'd' t 1 ° t °
1 d'd' °
2 d'd'
61 - 80 27 5
81 - 100 17 4
101 - 120 18 9
121 - 140 15 6 2
141 - 160 9 5 7 4
161 - 180 7
181 - 200 2 3 4
201 - 220 4 6
221 - 240
Sexual maturity 115 103 rnrn
(E) S. sordidus
Sexual identity 'i?'i' 1 od'd' °
1 d'd' t °
1 d'd' °
2 d'd'
61 - 80 10
81 - 100 15 2
101 - 120 5
1 21 - 140 12 3
141 - 160 20 3
161 - 180 18 3 3
1 81 - 200 19 15
201 - 220 10 24
221 - 240 7 17
241 - 260 3 22
261 - 280 11
Sexual maturity 135 rnrn 98 rnrn
(F) S. fasciatus
Sexual identity 'i''i' 1°d'd' t °
1 d'd' t 1° d'd' 2° d' d'
81 - 100 9 2
101 - 120 10
1 21 - 140 6 2
141 - 160 9 6
1 61 - 180 10 4
181 - 200 12 2
201 - 220 7 3 3
221 - 240 23 8
269
(F) S. fasciatus (cont. )

Sexual identity 'i'Q 1 00"0" t 1 00"0" t 1 ° 0"0" 2 0 0"0"
241 - 260 21 2 3 6
261 - 280 10 4 4 4
281 - 300 2 2 12
301 - 320 3 7
321 - 340 8
3 41 - 360
Sexual maturity 183 rom 167 rom
(G) S. sexvittatuB
Sexual identity 'i'Q 1 0 0"0" t t t 2 0 0"0"
61 - 80 10
81 - 100 8
1 01 - 120 7
1 21 - 140 7
1 41 - 160 4
1 61 - 180 5
1 81 - 200 5
201 - 220 14 2 4 5
221 - 240 7 7
241 - 260 18 3 7
261 - 280 6 7
281 - 300 10
301 - 320 6 7
3 21 - 340 4
Sexual maturity 194 rom ?
(H) S. venosus
Sexual identity 'i''i' 10 2° t t 2 0 0"0"
61 - 80
81 - 1 00 2
1 01 - 1 20 2
1 21 - 140 5
1 41 - 160 2
1 61 - 180
1 81 - 200 6
201 - 220 11
270
(H) S. venosus (cont. )

Sexual identity ~~ 1° 2° t t 2°d'd'
221 - 240 12
241 - 260 8 2
261 - 280 2 4
281 - 300 6
301 - 320
Sexual maturity 184 mm ? ?
(I) S. niger
Sexual identity ~~ t 2° d'd'
61 - 80 3
81 - 100
101 - 120 4
121 - 140 3
141 - 160 5
161 - 180 9 1
181 - 200 7 3
201 - 220 13 2
221 - 240 5 5
241 - 260 11
261 - 280 8
281 - 300
Sexual maturity 163 mm
In the majority of species there was a statistical relationship be-

tween size, color phase and sexual identity (Table 1). The probability
that a gaudy phase individual would be a male was overwhelming. No
female gaudy phase individuals were collected. In the case of S. sex-
vittatus, 3 gaudy phase individuals were transitional. Drab phase in-
dividuals were strongly biased toward the female identity. The pro-
portion of drab males ranged from 27.9% in S. globiceps to 1.02% for
S. sexvittatus. The few transitional specimens collected displayed a
variety of sexual identities (Table 1).
Sub-division of the primary and secondary male identities revealed

very little in the way of a predictable relationship with color phase
and sex, with the exception that the majority of functional drab phase
males were primary in derivation. A minority were transitional or sec-
ondary. The distribution of male identities in the gaudy phase tended
to reflect the distribution in the drab phase.
Color transitional specimens were either primary males or sexually

transitional. In the case of S.niger(dichromatism absent), the sexual
identities were size distributed, with secondary males having a greater
mean size than females. Patterns of sexual ontogeny inferred from the
data appear to conform to the dual male scheme of classification (mon-
271
andry and diandry) of Reinboth (1967). However, the condition of dian-

dry is highly variable, ranging from the situation in which primary
males make up only 1% of the drab population to that in which they
approach 30%. True monandry seems to be a restricted condition, and
given the flexibility of labroid sexual ontogeny the term should be
used with caution.
Histological examination of gonads, coupled with observation and col-

lection from spawning assemblages, enabled an estimate of the lower
limit of sexual maturity to be made. In the diandric species, S. sor-
didus, S. gZobiaeps and S. fasaiatus , drab phase males were substantially
smaller than females, when first mature. In the case of S. sexvittatus,
a 157 rom gaudy secondary male with mature sperm in the gonoduct was
smaller than any mature female examined. This suggests that primary
males will come to sexual maturity at an earlier age than females,
and also that females (by undergoing prenuptial transition) may enter
into the reproductive population at an earlier age than if they had
remained females.
Also, in diandric species, both primary and secondary males contribute

to the gaudy male phase. In the case of S. sordidus, the proportions of
primary males to females in the drab phase is essentially similar to
that of primary to secondary males in the gaudy phase. Inspection of
the size frequency distribution in the drab phase suggests that primary
males leave this phase via color transition at a smaller mean size than
do females via sexual transition.
However, it is difficult to make inferences concerning the demography

of the different identities on the basis of this sort of information.
If the distribution of primary males in S. gZobiaeps and S. fasaiatus
is examined, an anomaly appears. In each instance, the relative pro-
portion of primary males to secondary males increased with transition
to the gaudy phase. There are a number of mechanisms which might ac-
count for this. There may be a reduced probability of survival of
secondary males. Different schedules of growth and transition may
exist which convert drab primary males at a faster rate than females,
or we may be seeing a reflection of past events in which drab popula-
tions comprised mainly primary males. However, given the present meth-
ods of sampling and our state of knowledge of social organisation and
dispersion in scarids, the argument that the varying proportions of
primary and secondary males reflect a sampling bias cannot be dis-
missed.
If the gonad weights of mature individuals and the different color

and sexual identities are compared, it is obvious that two different
scales of gonad weight to body length relationship are present (Fig.2).
Although gaudy phase individuals have a greater mean length than drab
phase i.dividuals, their gonads are consistently smaller. This is true
for both male and female drab phase individuals. Sample sizes are too
small and too variable to assess whether there were any consistent
trends within the different male identities in each color phase. How-
ever, on entering the gaudy phase, regardless of whether this is ac-
complished by sex and color succession (female to secondary male) or
by color change alone, the gonad shows a sharp reduction in weight
(Table 2). This reduction can be clearly seen in specimens in transi-
tional coloration and may account for some of the variability seen in
the gonad weights of mature drab phase individuals. A similar conclu-
sion was reached by Roede (1972) concerning Caribbean labrids.
272
o
o
8 o Colou r : 0 Drab
o 6. Gaudy
o Trans i tion al
'" o
c 6 o o Sex
o o " cI cI
.::: z· d'd'
..'" o00o0 + Trans i tional
~ o
."
"c 4
0
<..1
(D
o <II
100 150 200 250
St an dard Len gth In mm
Fig. 2. The relationship of gonad weight to standard length in sexu-

ally mature specimens of S. sOlo didus collected during January and Feb-
ruary 1972 and 1973
III. Ecology
Scar ids gave the impression of being an ecologically uniform group

with striking similarities in feeding behavior and distribution. How-
ever, a closer examination of the various species revealed a number
of subtle differences.
The study sites were two adjacent, but topographically distinct regions
of the reef, as shown in Fig. 3. Area A was in a region sheltered from
prevailing southerly winds by nearby Wistari Reef. The outer slope of
this was covered with luxuriant growth of Acropora, the open interstices
of which provided a great deal of shelter for reef organisms. The upper
and lower boundaries of the reef slope were characterized by a decrease
in coral cover and an increase in dead calcareous substrates, when com-
pared to the mid reef slope. The area was subdivided into 4 zones and
these corresponded to obvious structural features of the reef (Fig. 3).
Area B was 1.4 kilometres from area A, at the north-west corner of

Heron reef. Beyond the reef crest was a shallow shelving area of car-
bonate rock. The amount of bottom covered in live coral was much less
than in area A.
273
Table 2. Mean gonad weight in grams of drab and gaudy individuals of

eight species of scarids. All individuals were mature and collected
during January - February 1972 - 1973
Species Color phase and sex

Drab <j><j> Gaudy 0"0"
S. formosus n 15 , x 3.094 ± 1 . 1 2 n 5, x 0.739 ± 0.647

S. lunula n 5, x 1.572 ± 0.839 n 6, x 0.252 ± 0.082
S. forsteri n 7, x 2.056 ± 1.047 n 4, x 0.792 ± 0.663
S. venosus n 5, x 3.709 ± 2.81 n 5, x 0.499 ± 0.275
S. sexvittatus n 5, x 7.539 ± 4.98 n 9, x 0.958 ± 0.914
S. fasciatus n 11 , x 8.374 ± 5.370 n 5, x 3.264 ± 1 .785
<j><j> 0"0"
S. niger n = 20, x 3.930 ± 2.185 n = 4, x

- 1 .452 ± 0.802
Drab <j><j> Drab o"d
S. globiceps n = 1 9, x 2.41 ± 1.09 n = 12,x 3.968 ± 6.018
Gaudy dd
n = 1 7 , x- = 0.558 ± 0.839
Table 3. Relative proportions of different substrates estimated by

10 meter line transects in Area A and Area B. Area A has been subdi-
vided as shown in Fig. 3. All substrates on which scarids were observed
to graze are identified as G.S. The remainder are live coral (L.C.)
macroscopic algae (Alg.), sand, sponge and soft coral (S.C.). Values
are mean percentages
G.S. L.C. Alg. Sand Sponge S.C.
Area A
Reef Crest n 8 73.8 21 .9 3.7 0.2 0.2 0.2
Reef Slope (1) n 4 26.7 40.3 1 .9 31 . 1
Reef Slope (2) n 4 8.4 86.9 1 .2 3.3 0.07
Patch zone n 5 60.7 6.5 2.9 26.5 3.4
Area B
Reef Slope n = 5 87.3 2.2 0.1 10.1 0.27
In Area A, counts of scar ids in 50 x 5 meter transects running parallel

to the reef crest, revealed that the different species were not distrib-
uted evenly within and between the different zones (Table 4). S. sordidus
was a possible exception. The richest scarid fauna, in terms of both
individuals and species, occurred on the reef crest where the greatest
amount of grazing substrate was also found (Table 3). However, the
amount of grazing substrate does not alone determine the abundance of
scarids. Counts made on the outer slope of area B revealed a relatively
impoverished scarid fauna, although the amount of grazing substrate
was greater (Table 3). When patterns of scarid abundance on a local
scale were examined, high variances were observed in some species
counts (Table 4). S. fasciatus on the crest and patch zone is a case
in point. This species displayed a considerable range of movement,
274
with transient groups of spawning individuals being observed in the

patch zone and large mobile feeding schools on the reef crest at high
water.
----N ~
b----r;:; - : - - - - - - - - - - - - - MHW - - - - - - - - - - "

A
metres
, 0 20 J0 4
' 0 50 60 70 80 90
Fig. 3. (I) Western tip of Heron Island reef showing the positions of
study areas (A) and (B). (1) Sand Cay. ( 2 ) Reef Flat, emergent at high
water. (3) Reef Crest. (4) Approximate position of the 12 meter depth
contour. (II) Profile of the reef at area (A) showing the subdivisions
of the reef used in Table 4. (A) Reef Crest. (E) Outer Slope (1 ).
(e) Outer Slope (2). (D) Patch Zone
Numerous observations made on feeding scarids showed that all of them

were feeding on plant material, although a certain amount of animal
material may have been included. Most grazing occurred on calcareous
substrates derived from coral skeletons in various stages of decay.
These supported a short turf of small algae, fragments of which were
found to be the major food item in parrot fish stomachs. In Table 5,
the major categories of substrates grazed by scarids are listed. From
the data shown, it can be seen equal areas of the different grazing
substrates were not browsed upon uniformly. In terms of the numbers
of species involved, and also the amount of grazing carried out, sca-
rids showed preferences for certain substrates. Two substrates were
275
grazed more intensively than any others by a great number of scarids.

These were dead patches on Porites growths and dead Acropora hyacinthus
plates.
Table 4. Relative abundance of scarids at area A (4 localities) and

in area B
Species Area A Area B

Crest Slope (I) Slope (2) Patch Area B
n = 6 n =7 n = 3 n = 6 n =5
s. sordidus 29.2 ± 22.6 10.7 ± 7.4 4.0 ± 3.6 6.5 ± 6.9 2.8 ± 3.3
S. gZob/fasc 60.5 ± 68.3 7.4 ± 9.4 1.0 ± 1.0 4.1 ± 8.4 2.6 ± 2.4
S. gZobiceps 0.8 ± 0.9 1.8 ± 1.6 0.5 ± 0.8
S. fasciatus 2.2 ± 4.3 2.8 ± 2.6 1.7±1.1 4.0 ± 6.4
s. niger 2.8 ± 2.4 1.4 ± 1.2 3.7 ± 2.1 0.2 ± 0.4
s. formosus 4. I ± 3.9 1.6 ± 1.5 0.3 ± 0.6
s. forsteri 1.0 ± 2.3 2.8 ± 2.3 5.2 ± 3.7
s. venosus 0.7 ± 1.2 1.0 ± 1.0 0.7 ± 0.8 4.6 ± 3.8
s. ZunuZa 2.8 ± 6.4 0.1 ± 0.4 1.7 ± 2.1 2.3 ± 1.7
s. gibbus 1.0 ± 1.5 0.4 ± 0.8 0.7 ± 0.6 1.3 ± 1.4
s. chZorodon 0.4 ± 0.5 49.4 ± 50.2
s. sexvittatus 1.8 ± 1.7 0.1 ± 0.4 0.7±0.6 0.2 ± 0.4
s. oviceps 0.83 ± 1. 2 0.3 ± 0.5
s. schuZtzi 0.3 ± 0.6 0.6 ± 0.9
s. ghobban 0.3 ± 0.5 0.2 ± 0.4
s. rubrovioZaceus 0.2 ± 0.4
s. scaber 0.2 ± 0.4 0.3 ± 0.5
S. fZavipectoraZis 3.3 ± 4.2
Scarus sp. 0.3 ± 0.8
If the grazing substrates are considered in terms of unbroken surface

area and permanence, the massive Porites structures must be ranked first.
The grazed areas, which were devoid of polyps, offered an extensive and
flat surface which was covered with a turf of filamentous algae. This
was raised above the reef sediments and sand pockets. Large Porites
growths on the outer slope were also the most permanent reef structures
present. Twelve Porites growths observed in 1967 were present in 1973,
despite the passage of two cyclones, which extensively damaged other
corals in the intervening years (Connell, 1973). Dead Acropora hyacinthus
plates were more readily disturbed by being broken up and buried in
sediment, but they still offered large uninterrupted grazing surfaces.
The fragments of Acropora rubble and the thickets of dead stalks and
holdfasts were prone to disturbance by wave action and abrasion, and
were covered by shifting reef sediments.
From this outline of feeding behavior and distribution, we conclude

that the smaller, less mobile, scarids are usually associated with
cover in the form of Acropora growths. Open environments, with a high
proportion of grazing substrate, harbour larger species which form
Table 5. Major grazing substrates on reef slope, area A. Approximately 5 square meters of substrate
Ol
'"....,
were observed for 10 minute periods
Substrate Dead Porites Dead Acropora Acropora Impacted Dissected Dead Bases of
surface hyacinthus rubble carbonate carbonate Acropora live
plate in sand rock rock thickets Acropora
Observations n 14 n 13 n = 22 n 5 n 5 n 3 n 9
Numbers
Total scar ids 46 62 63 20 9 15 8
x number 3.3 4.7 2.8 4.0 1 .8 5.0 0.9
Most abundant S. niger B. glob/fasc s. sordidus S. sOY'didus S. flavi- S. sordidus S. flavi-
species % (30.4) (50.0) (30.1) (35.0) pectoralis (53.3) pectoralis
(33.3) (25.0)
Grazing intensity
Total bites 3514 5475 1360 351 162 433 135
x bites 251 .2 421 .0 61.8 70.2 32.4 144.3 15.0
Most intensive S. niger s. glob/fasc S. venosus S. forsteri S. flavi- S. sordidus S. flavi-
grazer % bites (27.7) (65.0) (31 .2) (38.7) pectoralis (60.3) pectoralis
(54.9) (37.0)
277
schools. Both types of species tend to graze selectively on large and

permanent calcareous substrates.
These tendencies are shown clearly by the two species, S. niger and
S. faeiatus. S. niger is of interest, in that it is the only species pre-
sent in the study area which lacks dichromatism in the adult stages,
it is the only species which could be classified as monandric, and it
is the only one which was found concentrated about the large permanent
grazing substrates. It was never seen in schooling associations, and
showed little change in numbers or distribution with the tidal cycle.
S. niger achieved its greatest abundance on the outer reef slope in
areas covered with branching Acropora. Here S. niger was associated with
massive Porites growths. Its dependence on shelter is suggested by its
absence in counts made at the N.W. corner, where numerous Porites growths
occurred. Maps covering 1040 and 3025 square meters respectively, showed
that large Porites growths in area A were minority components of the
substrate. They were estimated at 0.7% and 1.1% of bottomcover respec-
tively, and were isolated islands of clean grazing substrate used
mainly by S. niger, surrounded by abundant cover in the form of Acropora
growths.
S. fasciatus showed considerable variability in its pattern of distri-

bution on the reef. This could be correlated with the tidal cycle.
At low water, large schools congregated on the reef slope. Table 6
shows the variation in numbers of S. fasciatus on the reef slope in
relation to tidal level. Far less variability was seen in S. niger.
Observations on the reef flats at high water revealed that S. fasciatus
moved onto the flats, grazing across extensive areas and returned to
the outer slope with the falling tide. Few other scarids exploited
these grazing resources.
Table 6. Counts of scarids in 100 x 10 meter transects on the reef

front (reef slope (1)) area A at high and low water
Species High water n = 5 Low water n 5

x-
-
x Number Number
S. faciatus 3.6 177.5
S. sexvittatus 8.6 6.6
S. sordidus 48.4 23.0
S. niger 4.8 7.8
S. venosus 3.6 1.0
S. globiceps 7.4 5.0
S. glob/fasc 12.4 17.2
S. formosus 0.4 2.0
S. lunula 0.2
s. fOl'steri 0.8 3.0
s. gibbus 1 .2
S. ghobban 4.4 34.8
S. oviceps 0.2 0.4
278
IV. Reproductive Behavior and Social Organisation
Both pair and group spawning, as described by Randall and Randall

(1963), were observed in Heron Island scarids. Group spawning was ob-
served most frequently in S. globiceps. It always occurred on the outer
slope of the reef, in areas of high current activity. Spawning took
place at, and immediately after, high water. Gaudy males present at
the site of group spawning reacted in three ways. They ignored the
spawning sequence entirely, attacked the drab males participating or,
in a minority of cases, actually joined the spawning run.
Group spawning also occurred in S. fasciatus but much less frequently.

In these instances only drab phase individuals participated. Pair
spawning by S. fasciatus occurred on the reef front in water deeper than
that of their normal feeding range.
No other instances of group spawning were observed in scarids. Pair

spawnings were frequently observed in most other species and invariably
occurred at the reef crest or reef slope at peak or falling tide. In
all dichromatic species in which pair spawning was observed, the number
of sexually active gaudy ~ales was always a minority of the total num-
ber of gaudy males present at anyone time.
As S. niger was not dichromatic, males were identified on the basis of

their size and behavioral displays. Active males tended to swim with
the tail and caudal peduncle in a characteristic raised position.
Intra-specific aggression was observed in S. niger becoming more fre-
quent with increase in size. This was not confined to periods of sexual
activity. A number of instances of inter-specific aggression, directed
almost exclusively ag.ainst other grazing scarids, were also observed.
Intra-specific aggression took two forms. Disputes involving a smaller
and a larger individual were transient and resolved by the retreat of
the smaller. Disputes involving individuals of similar size were more
prolonged and carried out at specific sites. These appeared to be
boundary disputes. Inter-specific aggression was directed toward other
species of scarids grazing in close proximity. The greatest focus of
inter-specific aggression occurred around massive areas of Porites and
probably explains the abundance of S. niger grazing on these substrates.
Only S. fasciatus and S. globiceps grazed to any significant degree on
these substrates and in nearly all instances did so in schools. Anal-
ysis of observations reveal that S. niger engaged in one intra-specific
interaction for every 13 min of observation and one inter-specific
interaction every 19 min from a total of 163 min.
In comparison, S. fasciatus showed only one intra-specific interaction

per 29 min of observation, and no inter-specific interaction in 117 min
of observation. The intra-specific interactions occurred only when
gaudy males actually touched in feeding schools. This species browsed
in dense schools over wide areas on a variety of substrates. A number
of these schools showed a mixed species composition and were charac-
terized by intensive grazing on different substrates. Aggressive inter-
actions within and between species were at a minimum during these feed-
ing episodes.
D. Discussion
In summarizing the relationship between dichromatism, sexual succession

and habitat selection in the Scaridae, the following assumptions have
been made:
279
1. Distinct color phases in scarids represent a compromise between

the necessity of concealment and rapid sexual and species-specific
recognition.
2. The sexual composition and dynamics of a particular population is

a result of selection pressures operating on individuals to produce
characteristic mating systems.
3. Coral reef environments are subject to local and unpredictable

disturbances even though they do not experience pronounced seasonal
regimes.
The numbers of predators present on reefs are enormous, and their

numbers must increase almost exponentially with decreasing size of
prey organisms. In the scarids examined here, the drab phases of the
smaller species (which showed a limited range of movement) were usu-
ally brown or reddish, overlaid by light bars or blotches (S. sordidus,
S. venosus, S. formosus). We contend this is a disruptive coloration,
being effective in species frequenting coral interstices and overhangs.
Species which formed feeding schools on the reef crest and flats (S.
faseiatus, S. gZobieeps, S. forsteri) were a uniform pale grey in the drab
phase. This is presumably a g60d compromise pattern, enhancing con-
cealment ag~inst a variety of backgrounds in which pale hues predomi-
nate. It is difficult to distinguish between small individuals of
these species in life which attests to strong uniform selection for
such a pattern.
Gaudy coloration is associated with pair spawning and larger size.

Rapid recognition by the female with respect to species and sex, is
obviously to the male's advantage. Scarids are remarkably similar in
morphological and also in meristic characters (Smith, 1956, 1959;
Schultz, 1958), and most species overlap in their local distribution.
Under these circumstances the convergent color patterns of the drab
phase are a liability to a male, which must be quickly and positively
identified by the females. Thus, gaudy, sexual and species-specific
coloration appears, with the adoption of pair spawning. Group spawning
involves different behavioral premises (Randall and Randall, 1963;
Reinboth, 1973; Robertson and Choat, 1973). Here the male's close re-
semblance to the female, inhibits attacks by larger gaudy males and
enhances the probability of a successful spawning. In most instances
there would be a strong selective advantage in resembling females as
closely as possible. Resemblances between drab males and females, and
the relative proportions of these to gaudy phase individuals, could
be interpreted in the context of automimicry (Brower et al., 1967).
Selection is subjecting the two male color identities to very different

demands, as in the Labridae (Robertson and Choat, 1973; Reinboth, 1962,
1973). Within groupspawnings, competition for female gametes must be
intense. The consequence of this competition, is the development of
testes that will deliver a maximum amount of sperm, as rapidly as
possible, through a massive gonoduct strengthened with connective tis-
sue septae. Gaudy males have small, sometimes minute testes (Randall
and Randall, 1963). Selection should be for greater body size and
distinct releaser components in the gaudy pattern. Roede (1972) has
argued that a reduction in testes weight in Caribbean labrids, should
be interpreted in the context of senescence and a decline in the func-
tional significance of the gaudy (terminal) male phase. The testes of
all gaudy phase scarids collected here displayed active spermatogene-
sis. The only exceptions were occasional sexually transitional individ-
uals. With development of the gaudy phase a reduction in gonad weight
occurred, in both primary and secondary males (Fig. 2). Development
of the gaudy color phase is therefore associated with access to female
280
gametes under conditions of reduced inter-male competition during each

spawning event. Reduction in testes size reflects this.
Monandry and diandry, with their associated pair and group spawning,
are at opposite ends of an ecological spectrum. The contrasting behav-
ior of S. niger and S. fasciatus represent different solutions to the
problems confronting grazing herbivores on a coral reef. These "prob-
lems" are the selective forces with which each species must contend.
Scarids and, indeed all herbivores, must gain continued access to
grazing substrates in an environment which may be subject to consider-
able disturbance. Predation is a selective force which can also strongly
influence the dispersion and distribution of different species (Randall,
1963; Talbot, 1965). Finally, selection will favor those individuals
which maximize their contribution to subsequent generations, and this
will be reflected in their particular mating system.
S. niger shows some degree of site attachment and is apparently defend-

ing two categories of resources. The first of these is a permanent
grazing substrate adjacent to cover, from which it will drive away
other herbivores (mainly scarids). Other herbivores reef fishes, for
example, Pomacentrusflavicauda, also act in this fashion (Low, 1971).
Secondly, the sexual structure of S. niger populations, and the nature
of the interactions within and between groups of individuals, suggest
that this species has a similar pattern of social organization to that
of Labroides dimidiatus. Here, most mating occurs within the confines of
a size-ordered group of females, dominated by a single secondary male
which has been recruited from within the group. The male has priority
of access to the entire female gamete crop. Under those circumstances
there will be prolonged bonds between males and females, and bright,
sexually-specific coloration will not be selected for.
S. fasciatus does not defend any particular site, with the exception of
a transient pair spawning territory. In no observed case were groups
of females dominated by a single male. The diandry of this species is
seen as a consequence of a loose social organization, and lack of site
attachment. Selection for diandry is postulated for circumstances sim-
ilar to those in Thalassoma lunare (Robertson and Choat, 1973).
Interpretation of the mechanisms determining the numerical proportions

of the sexual identities in a given population, is a more difficult
and compelling problem. Our statement regarding the relative propor-
tions of females, drab males and gaudy males in diandric populations,
implies that adjustment is obtained via some form of balancing selec-
tion. Similar arguments have been advanced by Gadgil (1972) to explain
the occurrence of male dimorphism in ungulates and insects. In scarids
with a loose social organization, the ability of males to fertilize
females becomes more uniformly distributed with size, and in these
circumstances smaller, behaviorally differentiated males may partici-
pate.
Such arguments imply that the frequency of occurrence of different

sexual identities in various mating systems is genetically determined.
However, there are no data as yet, which enables us to predict the
types of progeny that would be produced by the mating of a primary
male and a protogynous female. In addition, there is unequivocal evi-
dence that at least four sexual phenotypes may be expressed in genet-
ically identical populations of the cyprinodontid fish, Rivulus marmo-
ratus (Harrington, 1967, 1971). This demonstrates that different sex
phenotypes need not have a genetic basis. Moreover, laboratory experi-
ments have clearly dembnstrated that relatively short term environ-
mental influences early in life determine an individual's sexual ontog-
eny and subsequent reproductive activities. In these species, sexual
281
succession is a gradual process in which seasonal events play an im-

portant part. As Harrington (1971) points out, this remains the only
model available with which to interpret the genetics of hermaphrodi-
tism in fishes.
However, if we adopt such a model for patterns of sexuality in coral

reef fishes, some difficulties emerge. Coral reefs harbour diverse
assemblages of protogynous fishes exhibiting a variety of mating sys-
tems. It is difficult to envisage environmental factors, such as tem-
perature and day length, being the main determinants of sexual identity
in these species. Rather, monandric species, such as L. dirrridiatus,
should be subject to strong selective pressures to produce only proto-
gynous females. In the context of the social system described for this
speCies, the production of primary males would lead to a rapid loss
of fitness in individuals that did so.
Reports describing rapid sexual succession in two species of reef

fishes, Anthias squa:mipinnis (Fishelson, 1970) and LcibY'oides dimidiatus
(Robertson, 1972), are contrary to the age-distributed, environment-
ally-sensi tive schedule of succession characteristic of R. marmoY'atus
(Harrington, 1971). Admittedly, as these are marine fishes with plank-
tonic larvae, detailed environmental histories have not been compiled.
However, a general argument for the rapidity of sexual succession in
a monandric, site-attached species is compelling. The attainment of
male function provides priority of access to numerous female gametes.
A rapid development of the male identity and pair-spawning mode en-
hances this priority. Further experimental work clarifying the influ-
ence of age structure, and also environmental factors, on the schedule
of sexual succession in such species is awaited.
Finally, we note that unpredictable disturbances result in a great

deal of local temporal and spatial heterogeneity on reefs (Glynn, 1968;
Connell, 1973; Grassle, 1973). The argument that reef fishes display
flexibility of habitat selection in response to a localized heteroge-
neity has been anticipated by Sale (1972), in an experimental analysis
of habitat selection in reef dwelling pomacentrids. Scarids, expecially
the widely distributed and abundant s. sOY'didus, appear to possess con-
siderable flexibility within the context of their diandric population
structure. Primary males are precociously mature when compared with
females. They may leave the drab primary male mode at a small size,
via color transition. Females undergo succession at a variety of sizes.
Field observations on the spawning activities in both labrids and sca-
rids, strongly suggest that pair and group spawning are opposite ends
of a spectrum and not mutually exclusive. A further report of flexi-
bility of social organization and behavior in scarids is provided by
Ogden and Buckman (1973), in their study of S. cY'oicensis.
With respect to the possible mechanisms which might mediate this flexi-
bility, the suggestions of Harrington (1971) and Reinboth (1973) must
be noted. Their suggestions that the density of primary males may have
a suppressive influence on the testicular activity of hermaphroditic
R. maY'moY'atus, and that the densi ty of the gaudy male Thalassoma bifasciatum
influences the schedule of color transition in this species, point the
way to the types of experimental approaches required to answer these
questions.
Acknowledments.We wish to thank E.L. Nelson and A.V. Choat for histolo-
gical preparation of the gonad samples, the Heron Island Research Board
for the use of facilities at the Heron Island Research Station, Dr.
F.H. Talbot for providing accommodation and facilities on One Tree Is-
land and at the Australian Museum and J. Goedon and Dr. J. Connell for
data on coral distribution. We have benefited from discussions with
J.F. Grassle, W. Starck II, J.E. Randall, and P. Sale.
282
References

including Man (eds. C.N. Armstrong and A.J. Marshall), pp. 145-232,
Buckman, N.S., Ogden, J.C.: Territorial behaviour of the striped par-
rotfish Scarus croicensis Bloch (Scaridae). Ecology 2i (6), 1377-
1382 (1973).
Brower, L.P., Brower, J.V., Corvino, J.M.: Plant pOisons in a terres-
trial food chain. Proc. natl. Acad. Sci. (Wash.) 22, 893-898
(1967).
Connell, J.H.: Population ecology of reef building corals. In: Biology
and Geology of Coral Reefs, Vol. II (eds. O.A. Jones and R. Endean),
pp. 205-244. New York-London: Academic Press 1973.
Fishelson, L.: Protogynous sex reversal in the fish Anthias squa;nipinnis
(Teleosti, Anthiidae) regulated by the presence or absence of a
male fish. Nature (Lond.) 227, 90-91 (1970).
Gadgil~ M.: Male dimorphism asa consequence of sexual selection.
Amer. Natur. 106 (951), 574-580 (1972).
Ghiselin, M.T.: The evolution of hermaphroditism among animals. Quart.
Rev. BioI. 44 (2), 189-208 (1969).
Glynn, P.W.; Mass mortalities of echinoids and other reef flat orga-
nisms coincident with midday, low water exposures in Puerto Rico.
Marine Biology 1, 226-243 (1968).
Grassle, J.F.: variety in coral reef communities. In: Biology and Geol-
ogy of Coral Reefs, Vol. II (eds. O.A. Jones and R. Endean), pp.
247-270. New York-London: Academic Press 1973.
male gonochorists from eggs of the self fertilizing hermaphroditic
fish Rivulus marmoratus, Poey. BioI. Bull. 132 (2), 174-199 (1967).
determine when self-fertilizing hermaphrodites of Rivulus marmoratus
change into functional secondary males with a reappraisal of the
modes of intersexuality among fishes. Copeia 3, 389-432 (1971).
Low, R.M.: Interspecific territoriality in a pomacentrid reef fish,
PomacentY'Us flavicauda, Whitley. Ecology 22, 648-654 (1971).
Moe, M.A.: Biology of the red grouper Epinephelus morio (Valenciennes)
from the eastern Gulf of Mexico. Fla. Dept. Natur. Res. Lab. Prof.
Papers No. 10, 1-95 (1969).
Ogden, J.C., Buckman, N.S.: Movements, foraging groups, and diurnal
migrations of the striped parrot fish Scarus croicensis, Bloch (Sca-
ridae). Ecology .2.i (3), 589-596 (1973).
Randall, J.E.: Analysis of fish populations on artificial and natural
reefs in the Virgin Islands. Carib. J. Sci. 3, 1-16 (1963).
Randall, J.E., Randall, H.A.: The spawning and early development of
the Atlantic parrot fish, Sparisoma rUbripinne, with notes on other
scarid and labrid fishes. Zoologica, N.Y. 48; 49-60 (1963).
Reinboth, R.: Morphologische und funktionelle-Zweigeschlechtlichkeit
Labridae). Zool. Jb. (Allg. Zool. Physiol. Tiere) 69, 405-480 (1962)
Reinboth, R.: Biandric teleost species. Gen. compo Endocrinol. 2
(Abstr.), 146 (1967). -
23, 852-855 (1968).
Reinboth, R.: Intersexuality in fishes. In: Hormones and the Environ-
ment. Mem. Soc. Endocrinol. 18, 516-543 (1970).
Reinboth, R.: Dualistic behaviour in the protogynous wrasse Thalassoma
bifasciatum and some observations on its day-night changeover. Hel-
golander wiss. Meeresunters. 24, 174-191 (1973).
Robertson, D.R.: Social control of sex reversal in a coral reef fish.
Science l1l, 1007-1009 (1972).
283
Robertson, D.R., Choat, J.H.: Protogynous hermaphroditism and social

systems in labrid fish. Proceedings Second International Symposium
on Coral Reefs, Vol. I (1973).
Roede, M.J.: Color as related to size, sex, and behaviour in seven
Caribbean labrid fish species (genera Thalassoma, Halichoeres, and
Hemipterinotus). Studies on the fauna of Curacao and other Caribbean
Islands 138, 1-264 (1972).
Sale, P.: Niche overlap and instability in reef fish communities.
Crown-of-Thorns Starfish Seminar. Queensland 72-81 (1972).
Schultz, L.P.: Review of the parrotfishes family Scaridae. Bull. U.S.
natl. Mus. 214, 1-143 (1958).
Smith, C.L.: Contribution to the theory of hermaphroditism. J. Theoret.
13iol. 11 (1),76-90 (1967).
Smith, J.L.B.: The parrot fishes of the family Callyodontidae of the
Western Indian Ocean. Ichthyol. Bull. Rhodes Univ. 1, 1-23 (1956).
Smith, J .L.B.: The identity of Scarus gibbus Ruppell, 1828, and other
parrot fishes of the family Callyodontidae from the Red Sea and
the Western Indian Ocean. Ichthyol. Bull. Rhodes Univ. 1.§..I 263-282
(1959) .
Talbot, F.B.: A description of the coral structure of Tutia Reef and
its fish fauna. Proc. Zool. Soc. Lond. 145, 431-470 (1965).
Ecology and Physiology of Sex Reversal in
Anthias squamipinnis (Peters), (Teleostei: Anthiidae)*
L. Fishelson
A. Introduction
In the strategy of propagation in teleost fishes, intersexuality and

sex-reversal seem to be no less important than "normal", genetically
fixed bi-sexual reproduction. The more we know about these animals,
the more taxa appear exposing innate abilities for sex changes at
various stages of their life cycles (Reinboth, 1961, 1962; Atz, 1964,
1965; Hoar, 1969; Roede, t972). This process of sex-reversal is very
common among the serranoid fishes, typical being the protogynous cycle
in the genus Anthias and related species (Reinboth, 1964; Fishelson,
1970). Numerous papers have been published about these interesting
fishes, but only a few of them were performed on extensive material
in the natural habitats. In our region, Anthias squamipinnis (Peters) is
one of the most common fishes found along the western shore of the
Gulf of Aqaba, along the Sinai Peninsula (Fishelson, 1970j Popper and
Fishelson, 1973). Their typical habitats are shallow-water coral for-
mations, especiaJly isolated coral knolls and pinnacles, usually found
in front of the descending coral walls (Fishelson, 1971). Here, A.
squwmpinnis form stationary groups, active during daytime but hiding
within the coral crevices during nighttime. Occurring in large numbers
and together with the group-forming pomacentrid fishes Abudefduf azysron.
ChY'omis dimidiatus and often also Dascyllus t:r>imacuZatus, they form swarms
hovering around their rocky retreats and reacting in unison to various
stimuli from the surrounding (Fishel son et al., 1974). This paper
analyses the behavior of Anthias, especially the population ecology
and factors regulating the sex-reversal. A part of the data presented
were taken from the unpublished work of Miss Nurit Gunderman.
B. Methods
Monthly observations of A. squwmpinnis in its natural environment, were

carried out over a period of several years, by snorkeling and scuba
diving at various times of the day. Simultaneously, samples of fishes
were collected with a special net, described by Popper and Fishelson
(1973). For the purpose of histological examination of the gonads,
fishes were usually collected at 0.900 - 11.00 hr and 16.00 - 17.00 hr
on the same day. During February - March, the time of their peak re-
production, additional samples were taken at 12.00 and 15.00 hr. The
gonads of 360 fish were sectioned and an additional 200 fish were ex-
ternally examined. Collected gonads were measured and fixed in Bouin
or 10% neutral formaline. Histological sec.tions were cut at 6 \lm, 8 \lm
and 10 \lm thickness and stained in Delafield Haematoxylin-Eosin, Heiden-
hain Haematoxylin and Masson-trichrome. All fish were measured and their
¥partly sponsored by the Smithsonian Institution Grant, No. SFC'-7-0074.

285
gonadal index calculated in percent of weight of gonad to the body

weight. For determination of the developmental stages of oocytes, the
oocytes were divided into 8 stages, 4 of which (stage 1 to stage 4)
corresponding to the vitellogenic process of growth and ripening, and
the next 4 stages (stage 5 to stage 8) showed their maturation and
ovulation (Fig. 1). The ripe oocytes are 400 ~m to 500 ~m in diameter
and are provided with a large oil droplet that helps to keep the eggs
afloat after spawning. The same method was employed to investigate
the testes and ovotestes of sex-reversing individuals. A part of the
observations were performed in aquaria of the marine section, Depart-
ment of Zoology, Tel Aviv University, where tests with gonadotropic
hormones were also conducted.
Fig. 1a-d. Developmental stages of oocytes in A. squamipinnis:

(a) Stages 1, 2, 3; (b) stages 3 and 4; (c) stage 5; (d) stage 6.
(Black mark = 100 ~m)
c. Results
Field Studies
Usually A. squamipinnis form large groups, in which three forms of fishes

are observed according to their morphological pattern and behavior
(Popper and Fishelson, 1973). The most numerous are the immature and
reproducing females, that comprise 90% of the group. These are reddish-
orange fishes with a shining violet iris. Less numerous are the usually
larger males, characterized by a violet colored body, dark eyes, pro-
longed first rays of the dorsal fins and a special pattern of brown-
violet blotches on the pectoral fins. The third type is formed by sex-
286
reversing individuals, usually occurring sparsely within the group.

At various stages of sex-reversal these fishes display meristic and
behavioral characteristics intermediate between male and female. Ob-
servation of the gonads showed that from the smallest individuals up
to 60 - 65 mm TL (total length) only females were found, the largest
ones reaching ± 70 mm. The sex-reversing specimens were from 65 to
86 mm TL, whereas males were almost always larger, attaining 110 mm TL
(Fig. 2). Only occasionally were morphological males observed with a
size similar to that of females. They occurred in isolated groups or
singly among normal males.
... ..... . .... . .

..... ...............
... . ..... .... ........
~
..
a ..... ... .........

............. ..
..:
z
o
o d' ..
Fig. 2. Correlation be-
80 90 100
tween body length and
30 40 50 60 70
sex in A. squarrnp1-nnis.
BODY LENGTH (mm) (Each dot = 1 fish)
Such mixed groups of Anthias, occurring stationary on a given place,

feed on plankton transported by the .local water currents. Most promi-
nent are swarms numbering several thousand individuals, but smaller
groups are also found hovering beside coral blocks and i30lated rocks.
A strong correlation is observed between the number of males in a group
and the total number of females. In groups of less than 10 fish, males
were not usually found. These seem to be temporary units, for some
reason isolated from larger groups close by. Single males were found
in groups of 10 or more fish, and with the increasing number of female
fish the number of males increased accordingly.
In such groups the males form two castes: One caste is the dominant
fishes, holding separate territories over the upper part of the coral
block. These "superior" individuals defend their places against con-
specific males performing typical U-swims and threatening postures
(Popper and Fishelson, 1973). During reproduction these courting males
ascend toward the female group and pair with receptive females, fer-
tilizing the expelled eggs. While feeding, they mix among the females,
but remain alert and close to their territory. It seems that in the
case of A. squamipinnis that feed on plankton and produce pelagic eggs,
the territory of a male has a rudimentary meaning and seems to act
for "selbstbehauptung" of the sexually active male. The second caste
of males, usually more numerous than the first, are the bachelor males
that form a loose assembly, living separately and close to the lower
part of the habitat occupied by the territorial males. Each time that
such bachelors try to ascend and join the females, or take part in
courtship, they are prevented from doing so by the antagonistic action
of the territory-holding males.
287
D. Reproduction
Al though A. squamipinnis are tropical. fish, they show a typical annual

cycle of reproduction, visible in the courtship behavior of males and
spawnings observed in nature. Monthly samples of gonads confirmed the
ethological observations - a peak of reproductive activity from Novem-
ber till March, that then decreases until July. From August to October,
when no reproduction occurs, the gonads are characterized by ripening
of oocytes and increase of weight (Fig. 3). During reproduction a
diurnal cycle of behavior is observed, characterized by a peak of
courtship in the late afternoon. Between 16.00 to 17.00 hr the actual
mass-spawning occurs, during which numerous females ascend toward
the surface accompanied by males who fertilize the eggs. This behav-
ioral regularity is well illustrated in the ripening of oocytes during
the day (Table 1) and the occurrence of ovulated eggs in the ovaria
(Table 2) .
OVARIUM 6
TESTIS 0
12-
6.6-
x
w
0 5.7-
~
Vl
;::: 4.8-
(J)
w
>-
3.9-
0<1
>-
<r
« 3.0-
>
0
x 2.1-
«
:2
1.2-
0.6-
6 7 8 9 10 11 12
2 3 4 5 Fig. 3. Indices of ovaries
MONTHS and testes in A. squamipinnis
Table 1. Ripening of oocytes during the day (as observed in time of

reproduction)
Time No of Stage of oocytes

<i'<i' 4 5 6 7 8
0.900 5 4 0 0 0
12.00 5 0 0 5 0 0
15.00 5 0 0 0 2 3
17.00 5 0 0 0 4
The number of eggs found in the ovary of a single female is very vari-
able, depending on the size of the animal and the season (Fig. 4).
Occurrence of ovulated eggs in ovaria of 85% of the females caught
288
daily, supports the view that each fish spawns numerous times during
the reproductive season. Artificial fertilization, performed in the
laboratory, showed that the hatching occurs 36 hours after fertiliza-
tion. On the third day the larvae of 1.5 rom length absorb their yolk
remnants and begin to feed on unicellular organisms.
Table 2. Ovulated eggs in ovaria of females collected in the morning

and afternoon
Month Morning Afternoon

Total Ovulated Without Total ovulated Without
n eggs eggs
January 20 0 20 14 12 2
February 30 0 30 30 26 4
March 30 0 30 30 27 3
April 10 0 10 10 4 6
18
4000
3000 8
:2: 20
:2
a<
~
0
11
~ 2000
0-
m
c.:>
c.:>
w
0
I
w 7
>-
j 1000
:::J
>
0
I Fig. 4. Number of ovulated

3 4 5 6 eggs in ovarial lumen of
MONTHS A. squamipinnis
E. Sex-reversal
In the protogynous A. squamipinnis sex-reversal normally begins after

the fish first functioning c'-s a female, attains a length of ± 65 rom TL.
Yamamoto (1969) states that in Sparidae and Serranidae, the juvenile
gonad has an ambisexual structure, with ovarian and testicular pri-
mordia. This does not apply to A. squamipinnis, in which the development
289
Fig. Sa-g. The gonads of sex-reversing fishes in various stages of

development: (a) Developed testis in spermatogenesis. (b) Beginning of
spermatogenesis (8) in an ovary. (c) Pockets of spermatogenic cells
(8) in a resting ovary. (d) Atretic (a) development in an ovary.
(e) Central part of ovary with spermatogonia on the lamellae.
(f) Atretic oocytes and spermatogonia. (g) The fresh developed testis
with luteinized oocytes stage 1 and 2. (Black mark = 100 ~m)
290
of male gonads resembles descriptions for various serranid and labrid

fishes (Reinboth, 1967, 1970). As the histological picture shows
(Fig. 5), the first signs of testicular tissue appears on the isthmus
between the ovarian lobes, close to the genital pore, and as islets
below the various ovarian tissue on the lamellae. As the testicular
tissue gradually develops it pushes the oogenic part toward the cen-
ter of the ovary lumen. The beginning of masculinization is accompanied
by atretic developments within the female part, especially the resorp-
tion of ripe eggs (Fig. 5) and the disappearance of stage 1 to 4 00-
cytes. These changes in the gonads are usually synchronized with mor-
phological and behavioral changes in the fish. The first markers are
the agonistic motions of the sex-reversing fish toward the individuals,
and especially toward aggressive males. Soon these fish start to per-
form the U-swims typical for the male pattern of behavior (Popper and
Fishelson, 1973). This behavior is also accompanied by the development
of pigment blotches on the pectoral fins and a slight violet coloration
on the trunk. All these markers increase gradually, and in more ad-
vanced stages such fishes begin to compete for a place with territory-
'holding males and also for the chance to fertilize females. Observa-
tions in nature and the laboratory showed 'that such "new" males are
very active and "milking" them results in much more sperm than from
older males. Measurements showed that, in some cases, the testes of
a new male are two or three times larger than those of an older male
taken from the same group. In the laboratory such new males attain
their entire male pattern in 10 to 12 days. It seems that such males
establish their own territory or compete for a territory, pushing the
older males down into the bachelor group. This process of sex-reversal
was seldom observed during the peak months of reproduction. As Fig. 6
demonstrates, the highest number of inters exes occurs during the summer
months of May to August, which are marked by their low in reproduction.
OV ST 4 - 0
OVST 1-6
100
15 0
'"
>
'*- '"'"
(f)
(f) III
'"«
<.:l
o
...
(f)
1O~
Z 50
...0
«
13x
...J
w
(f)
::J 5
>
0
'"...w
Z
0 Fig. 6. Correlation be-

2 3 4 5 6 7 8 9 10 II 12 tween ovulation and sex-
MONTHS reversal in A. sqamipinnis
A similar phenomenon was found in SacUJ"a margaritacea by Yamamoto (1969)

The length of time a freshly "evolved" male remains on the top layer
of the habitat is unknown. Sectioning the gonads of bachelor males
always shows filamentous degenerative testes, with a filamentous struc-
ture and very few spermatogonia. Such senescent male organs were re-
cently described by Roede (1972) in males of Thalassoma bifasciatwn. In
291
the laboratory, sex-reversal of A. sqummp~nnis was successfully achieved

only if other males were not present in the same aquarium. By separat-
ing males from a group of females in the aquarium it was possible to
induce sex-inversion in all of them (Fishelson, 1970). Additional
experiments showed that in several cases sex-reversal was prevented
for a time if a group of females could only see males beyond a glass
partition. Experiments with separation of males and the construction
of a water flow from their aquarium into a container with females, did
not prevent sex-reversal. Thus, it seems that visual (but not chemical)
stimuli are important for sex-reversal.
Artificial Sex-reversal
Induction of sex-change and behavior by the use of gonadotropic hor-

mones is widely discussed by Liley (1969), Yamamoto (1969) and Roede
(1972). Experiments on A. squcurripinnis were performed with the gonado-
.tropic hormones, testosterone and estradiol, to learn about the factors
regulating sex-reversal. In one set of the tests, three groups of
fishes each consisting of individuals of 40 to 60 mm TL, were kept in
separate aquaria. TestostEfrone was added twice to the water in varying
doses of 1.5 mg/l; 3.0 mg/l and 5.0 mg/l respectively. As the solubil-
ity of testosterone in water is poor, obviously only some of it affected
the fishes. In an additional experiment, silicone capillaries filled
with testosterone, were implanted in the trunk muscles of the fish.
The results of some of the experiments are summarized in Table 3. The
first sign of reaction to testosterone (one day after the treatment)
was the more frequent agonistic behavior among fishes in aquaria with
5.0 mg/l. In the days following these behavioral acts in the aquaria
containing various concentrations of testosterone, became more promi-
nent, along with the development of fin pigmentation typical of inter-
sex-fishes.
Table 3. Changes in morphology and behavior of females of A. squcurripinnis

(40 to 60 mm TL) treated with various concentrations of testosterone
(days after treatment)
Concentra- Begin- Display Fin pigment Body pigment U-swinnning Loop swimming
tion of ning of
hormone agonism Lateral Mouth Firs·t Final First Final Partly Full
1.5mg/lw 4. 4. o. 6. o. 5. o. o. o. o.
3.0 mg/lw 1.5 2. 6. 3. 7. 5. 8. 8. o. O.
5.0 mg/lw 1. 2. 4. 2. 5. 4. 7. 6. 9. II.
~he clim~x ~f.thi~ development w~s the occurrence of male-courting

loop-sw~ms 1n ~1shes treated w1th 5.0 mg/l testosterone (Table 3).
Gonads or such f1sh (55 - 60 mm TL) were similar to the ovotestes of
la:ger sex-reversing specimens collected in nature. Here also, degener-
at10n of oocytes and spermatogenesis were visible. Contrary to this,
the gonads of smaller specimens (35 - 40 mm) from the same medium
7
showed d generative oogenic tissue but no signs of spermatogenic tissue.
Gonads d1ssected from fish from 1.5 mg/l testosterone concentration had
degenerated oogenic tissue, but no development of male cells.
After 12 days in a medium containing testosterone, the fishes were trans-

ferred to a large aquarium with normal sea water, corals and rocks _
292
imitating a small natural habitat. Three days after transfer fishes,

previously from the 1.5 mg/l testosterone medium, began to lose their
poorly developed male coloration and dark blotches on the pectoral fins.
Two weeks later they were again in female dress, mixing with females in
the aquarium and behaving like them. Contrary to this, fishes from the
5.0 mg/l concentration preserved their male characteristics and ag-
onistic behavior, establishing their separate places on the rocky sur-
face and various spots of the aquarium. Their gonads showed typical
spermatogenic tissue. From this group of 8 fish, four large ones re-
mained male and four smaller ones reverted to the female color after
three weeks. The gonads of these specimens showed a mixture of dis-
persed spermatogenic islets, connective tissue and regenerating oogenic
layer.
In another experiment, sex-reversing individuals of 75 mm TL, the sperm

of which was used for artificial insemination, were put into isolated
aquaria, in which doses of 1.0 mg/l and 2.0 mg/l of estradiol were
added for five successive days. The first signs of estradiol influence
appeared after six days. All the females lost their primordial dark
blotches on the pectoral fins and the violet color on their trunks.
After ten days, the gonads of some of these females were dissected and
histologically investigated. Other fish were put into a community tank.
These mixed with the group of intact females and behaved like them for
10 - 12 days. Then their behavior changed, beginning to become more
agonistic and displaying swimming patterns belonging to a part of the
male display. After 18 - 20 days these females again showed U-swims
and color development typical for sex-reversing individuals.
The gonads of the dissected specimens showed atretic oocytes and dense
groups of testicular cells, but no normal spermatogonia or ripe sperm.
F. Discussion
Like various rodent mammals and weaver birds, groups of Anthias squami-
pinnis, form monospecific communities. In Anthias, the process of sex-
reversal seems to be the factor that regulates the eco-social structure
and the position of each individual in the community. The main pathway
of this development is the behavioral one, that in a wide sense belongs
to a very general FAP (Fixed Action Pattern) and the morphological
one, that may be marked as a part of a FMP (Fixed Morphological Pat-
tern). These patterns, regulated by eNS and sex-hormones, are innate
and intermingle with each other. Both these sets of characters mould
the group of Anthias in which the division in females, active males and
bachelor males, produce a community with a specific hierarchical struc-
ture: The females form stationary shoals hovering around a coral pin-
nacle; the active males occupy individual territories along the upper
part of the rock; and the bachelor males form a group hiding close to
the bottom layer of the habitat. All the young are females, and, if
they are lucky, end up as males in the bachelor group. There is a
constant flow of individuals from one social structure to the other.
This movement seems to be induced by hormones of gonadal or pituitary
origin. As demonstrated, administration of testosterone will induce
U-swims that are a part of the FAP of a male Anthias. This is prominent
even before testicular development. As such behavior is observed also
in juvenile fish with undeveloped gonads, it seems that the target
organ is the CNS. From here the stimulation of new behavior occurs.
In natural habitats, such developments evolve gradually during which

the changes in FAP and FI~ are in unison with development of the gonads.
293
As sex-reversal occurs seasonally, the shift of a specific fish through-

out the social structure is a seasonal one.
Observations on breeding populations of Anthias showed that the terri-

tory size and spatial pattern of distribution are of high stability.
As the number of territories is space-dependent, this will lead to a
formation of male groups without territories. During summer, as the
spawning decreases and sex-reversal increases, a stimulus develops,
along this line, during which new males seek territories. Such individ-
uals will squeeze themselves into the existing colony, chasing an old
male toward the bachelor group, or may split away with a group of fe-
males forming a small colony. Processes like this could be produced
only partly by artificial treatments with androgens or estrogens. In
Anthias, as in other fish, administration of testosterone produces males
even if the size of the fish is smaller than the normal males observed.
The occurrence of such mini-males in nature, pOints to the possibility
t~at the normal f~male-male sequence of hormonal development may be
dlsturbed and a flSh may reach the male stage without passing through
the female stage. In the laboratory, such males were produced from fish
of more than 40 mm TL. Below this size, testicularization of the Dri-
mary ovaries was not observed. It seems that at a certain stage of
development, an "X-receptor" evolves, that enables the utilization of
testosterone, and induces male development.
The action of estradiol on sex-reversing individuals of Anthias, also

exposes the role of this hormone on the FAP and FMP. Such fishes show
degenerative developments in their male morphology and behavior and
their spermatogonic tissue. It seems that in Anthias, as in mammals
(Neumann et al., 1969) the estrogens have an anti-androgenic effect,
inhibiting the hormonal function of the testis. In this way, the fe-
male system forms a block over the primordial male. In Anthias, and
possibly also in other fish, the decrease of ovarian activity caused
by seasonal changes, minimizes this block and leads toward testicular-
ization and development of male characteristics. Experimental addition
of androgens acts as a factor in decreasing the blocking ability of
female hormones - and so encourages the development of spermatogonia.
Such process also enables the production of small males. When normal
androgenic development starts, the process is irreversible and hormonal
treatment can only slow it up.
The problem of sex-reversal is also interesting from the point of view

of the sex-ratio in specific populations of animals. As mentioned by
Wenner (1972), the deviation from the normally expected 1:1 sex-ratio
is more conspicuous than expected. This also seems to be the case for
Anthias and other protogynous fishes, in which all the small fishes are
female and all the large ones are males. The question arises that if
all females change to males, then the sex-ratio at a given stage should
be close to 1:1. But this is not the case. As demonstrated before, the
number of males to females is 1:15 to 1 :20. It seems that two factors
may mould such population structures: One is predation acting on larger,
more exposed fishes, rather than on smaller ones living closer to the
substrate. The other factor (corresponding to the first one) is the
differential ability to utilize the retreats within the coral habitat.
Observations show that larger holes are occupied by larger fishes,
especially predators such as pteroids (Fishelson, in press), moray
eels and locus fish. Anthias males trying to enter such holes (a) will
be captured or, (b) must interact with larger animals. This is a dis-
advantage for their survival. Contrary to this, minute holes and crev-
ices are excellent places for the smaller females that cannot be at-
tacked here by larger predators. Thus, each of these factors seems to
act constantly against larger specimens, especially the second factor
that is typically density-dependent.
294
It seems that in Anthias squo.mipinnis, the survival curve is an opposite

one compared to curves of other animals, in which the ability to sur-
vive increases with age and size.
References
Atz, J.~.: Intersexuality in fishes. In: Intersexuality in Vertebrates

Atz, J.W.: Hermaphroditic fish. Science 150, 789-797 (1965).
(Teleostei, Anthiidae) regulated by presence or absence of male
fish. Nature (Lond.) 227, 90-91 (1970).
Fishelson, L.: Ecology and distribution of the benthic fauna in the
shallow waters of the Red Sea. Mar. Biol. 10, 113-133 (1971).
Fishelson, L.: Ethology and reproduction of the pteroid fishes found
in the Gulf of Aqaba (Red Sea) especially Dendrochirus bloachypterus
(Cuvier) (Pteroidae, Teleostei) (in press) .
Fishelson, L., Popper, D., Avidor, A.: Biosociology and ecology of
pomacentrid fishes around the Sinai Peninsula (Northern Red Sea).
J. Fish Biol. 6, 119-133 (1974).
Hoar, W.S.: Reproduction. In: Fish Physiology (eds. W.S. Hoar and
D.J. Randall), Vol. 3, pp. 1-72. London: Academic Press 1969.
Liley, N.R.: Hormones and reproductive behavior in fishes. In: Fish
Physiology (eds. W.S. Hoar and D.J. Randall), Vol. 3, pp. 73-116.
Neumann, F., Elger, W., Steinbeck, H.: Drug induced intersexuality
in mammals. J. Reprod. Fert., Suppl. 7, 9-24 (1969).
Popper, D., Fishelson, L.: Ecology and behavior of Anthias squamipinnis
(Peters, 1855) (Anthiidae, Teleostei) in the coral habitat of Eilat
(Red Sea). J. expo Zool. 184,409-424 (1973).
Labridae). Zool. Jb. (Physiol.) 69, 405-480 (1961).
Reinboth, R.: The effect of testosterone on female Coris julis (L.) a
wrasse with spontaneous .sex-inversion. Gen. Compo Endocrinol. 2,
39 (1962). -
Reinboth, R.: Natlirlicher Geschlechtswechsel bei Sacura marga:ritacea
(Hilgendorf) (Serranidae). Annot. Zool. Jap. 40, 181-186 (1963).
Reinboth, R.: Inversion du sexe chez Anthias anthias (L.) (Serranidae).
Vie Milieu (Suppl.) 17, 499-503 (1964).
Reinboth, R.: Sex reversal in the Black Sea bass Centropristes striatus.
Anat. Rec. 151, 403 (1965).
Reinboth, R.: Protogynie bei Chelidoperca hirudinacea (Cuv. et Val.)
(Serranidae). Ein Diskussionsbeitrag zur Stammesgeschichte ambi-
sexueller Fische. Annot. Zool. Jap. , 181-186 (1967).
Reinboth, R.: Intersexuality in fishes. Mem. Soc. Endocrinol. ~, 515
- 543 (1970).
Roede, 11. J.: Color as related to size, sex and behavior in seven Carib-
bean labrid fish species (genera Thalassoma, Halichoeres and Hemiptero-
notus). The Hague: Publ. Martinus Nijhoff 1972.
Short, R.V.: An introduction to some of the problems of intersexuality.
J. Reprod. Fert. Suppl. 7, 1-8 (1968).
Smith, C.L.: Hermaphroditism in some serranid fishes from Bermuda. Pap.
Michigan Acad. Sci. 44, 111-119 (1959).
Smith, C.L.: The patterns of sexuality and the classification of ser-
ranid fishes. Amer. Mus. Novit. 2207, 1-20 (1965).
and D.H. Randall), Vol. 3, pp. 117-175. London: Academic Press 1969.
Wenner, A.M.: Sex ratio as a function of size in marine crustacea.
Amer. Nat. 106, 321-350 (1972).
The Evolution of Hermaphroditism In Fishes
C. L. Smith
A. Introduction
Recently, a renewed interest in the philosophy and methodology of

systematic biology has resulted in some dramatic changes in our con-
cepts of the phylogenetic relationships of the vertebrates. These
changes are perhaps most obvious among the fishes, primarily because
the fishes are the least well-known of all of the vertebrates. There
are several reasons for this. ~irst of all, many aquatic habitats are
difficult to sample so that our inventory of recent species is still
far from complete. Second, there are some 21,000 species of fishes,
as many as all of the other vertebrate groups put together and third,
the demands of moving through the relatively dense water medium have
limited the general external form of fishes so that detailed anatomi-
cal studies have been required to reveal their true diversity. Finally,
it is generally difficult to maintain and rear fishes, especially ma-
rine species, in the laboratory and this has severely limited the
studies of genetics and physiology that could be carried out. Even
today, with modern aquarium techniques, there are still only a rela-
tively few species that can be maintained in the laboratory for gener-
ation after generation.
In 1966 a major paper by Greenwood et al. spurred a revival of inter-

est in the higher classification of fishes. The new classification
proposed by those authors has subsequently been modified by Nelson
(1969), Rosen and Patterson (1969), Rosen and Greenwood (1970), Nelson
(1972), Gosline (1969) and others and most recently, by a series of
authors in a Symposium of the Linnean Society, edited by Greenwood
et al (1973). The overall success of these studies is largely due to
rigorous attempts to base assessments of relationships on precise
phylogenetic concepts, particularly those so lucidly defined by Hennig
(1966), Brundin (1966) and Crowson (1970).
It must be emphasized, however, that these papers are only a beginning
and there are still vast groups of fishes (particularly among the
perciforms) whose limits are undefined and whose relationships remain
unclear. Nevertheless, it is appropriate at this time to re-examine
the distribution of hermaphroditism among fishes in order to (1) eval-
uate hermaphroditism as a derived character that can aid in our under-
standing of fish phylogeny and (2) gain a better understanding of the
phenomenon of hermaphroditism from an examination of the fish groups
in which it occurs.
The occurrence of normal and teratological hermaphroditism and other

types of intersexuality was summarized by Atz (1964). More recently,
Reinboth (1970) has reviewed what is known of the occurrence and func-
tioning of normal hermaphroditism. The present paper is an attempt to
supplement their reviews with further observations on the independent
derivation of hermaphroditism in different groups of fishes and to
postulate the environmental conditions under which hermaphroditism is
selectively advantageous.
296
This discussion will be limited to those fish groups in which hermaph-

roditism is the usual mode of reproduction, in other words, those
fishes for whom hermaphroditism has proved to be a superior means of
assuring that hereditary materials are passed from one generation to
the next. We will not be c.oncerned with the physiology of hermaphro-
ditism, except indirectly, nor will we consider the groups such as
the agnathans and anguilliforms that have only juvenile hermaphroditism,
in which the undifferentiated gonads of both sexes contain cells that
stain-like oocytes. These so-called auxocytes may ultimately prove to
be of great importance to our understanding of the nature of sex dif-
ferentiation but in the light of present knowledge they are too diffi-
cult to interpret to be considered here.
B. Characteristics of Hermaphroditic Fishes
From a functional viewpoint normal hermaphrodites must be either syn-

chronous, capable of producing both eggs and sperms at the same time,
or successive, in which case they function as one sex first, then
undergo sexual inversion and conclude their reproductive life as the
other sex. petailed anatomical study however, has revealed that there
is a considerable variation in the morphology of the gonads and that
both synchronous and successive hermaphroditism can be effected with
a number of different arrangements of the sexual tissues. While it is
by no means proved, I believe that the evidence from hermaphroditic
fishes indicates that spermatogonia and oogonia arise from separate
tissues and therefore have different ontogenetic origins.
The observed differences in gonad structure strongly suggest that her-

maphroditism has originated independently several times among the
fishes and this is borne out by the fact that hermaphroditism occurs in
groups which are not particularly closely related. If the phylogenies,
as formulated on the basis of other evidence are correct, the only al-
ternative to independent origin of hermaphroditism would be multiple
independent origin of gonochorism. This is highly improbable, although
there are some groups for which secondary reversion to hermaphroditism
seems to have occurred.
Some of the structural features of gonad anatomy that seem to indicate

independent convergent origins of hermaphroditism are:
1. The degree of separation (territorialization) of the male and fe-

male sections of the gonad. In the synchronous Aulopiformes, testes
are completely separated from the ovaries and each is enclosed in its
own separate connective tissue capsule. In the sparids and emmelich-
thy ids the male and female zones are separated by connective tissue
but the sperm ducts pass within the ovarian wall. In Serranus, the male
and female tissues are enclosed in a common capsule and in Epinephelus
there is a complete mixture of male and female elements scattered
throughout the germinal epithelium.
2. Whether the male tissue is dorsal or ventral to the ovarian part

of the gonad. In Aulopiformes the testes are dorsal to the ovaries
but in the Sparidae the testicular region is on the ventral wall of
the gonad. In those serranids with territorial gonads the male tissue
is confined to the margin of the ventral non-germinal wall of the ovi-
duct. Other arrangements include linear alternation of male and female
regions in Gonostoma and having the testicular lobe lying within the
oviduct in Pseudogramna.
297
3. The arrangement of the gonoducts and other non-germinal parts of

the reproductive system. In Aulopiformes each testis has a tortuous
sperm duct that either leaves the body separately or joins the other
just before leaving the body. Serranus and Pseudogramma have a common
sperm duct that runs within the posterior (dorsal) wall of the common
oviduct. In the transforming EpinepheZus and protogynous labrids, the
sperms flow through a series of lacunae that develop in the outer wall
as the gonad transforms from an ovary to a testis. In some species of
platycephalids the oviducts form by fusion of the gonad wall only af-
ter the organ has finished functioning as a testis.
There is also variation in the arrangement of the muscle layer of the

capsule and other supporting tissues.
4. Type of testes. In general, gonochorists tend to have tubular testes

and tranforming hermaphrodites tend to have acinar testes, but this is
not always true nor is the distinction between the two types always
clear. Intermediates, in which the tubule boundaries are indistinct
and the spermatogonia develop in clumps, are common.
5. Variations in sexual structures within a single species. Liem (1.968)

has reported variation in patterns of sex inversion in geographically
separated populations and in the labrids and scarids it is common for
there to be two types of males (Reinboth, 1970).
The general lack of correlations among these features is itself further

evidence that they have originated independently in different groups.
C. The Occurrence of Normal Hermaphroditism in Fishes
Fig. 1 is a summary of the interrelationships of the higher groups of

fishes as they are presently understood. In the following discussion
we will use the term "primitive" to indicate a lineage that diverged
early and "advanced" to indicate a line that diverged later. We re-
cognize, of course, that a primitive group can have many unique spe-
cialized features that .arose after the separation occurred. Hence,
they are not shared by the more advanced lineages.
The following notes are supplementary to the reviews by Atz (1964)

and Reinboth (1970).
I. Ostariophysi
The most primitive teleost group for which normal hermaphroditism

has been reported is the division Ostariophysi. With few exceptions,
Ostariophysi are confined to fresh waters where they dominate the fish
faunas of all continents except Antarctica and Australia. Many of the
8,000 or more species have been studied thoroughly but only one,
Cobi tis taenia Linnaeus, has been reported to be hermaphroditic (Lodi,
1967a,b). From the description and figure given by Lodi, the gonad
seems to be a mixture of male and female elements without any separa-
tion into zones or territories. The oocytes appear to be embedded in
a matrix of semeniferous crypts. I have seen this same condition in
an abnormally hermaphroditic sturgeon (Atz and Smith, in prep.) and
it appears that in some populations, at least of Cobitis taenia, this
"abnormal" condition has become genetically fixed. Lodi reports "herm-
aphrodite gonads ranging from almost exclusively male territories to
almost exclusively female territories". "The existence of pure males
is not demonstrated as yet." Whether or not other species of Cobitidae
will be found to be hermaphroditic, as Lodi suggests, remains to be seen.
298
B
OSTARIOPHVSI STOMIATIFORMES Myctophiformes ATHERINOMORPHA
Porocanthopterwii 4 PERCOMORPHA
Euteleostei
Percomorpho
Fig. 1 A-C. Phylogenetic relationships of hermaphroditic fishes. Com-

piled from various sources. (A) Lower fishes. Agnatha to Euteleostei.
(B) Ostariophysi to Percomorpha. (C) Percomorpha. Groups with hermaph-
roditic representatives are listed in capitals. Numbers indicate lin-
eages in which hermaphroditism has arisen independently
299
II. Stomiatiiormes
The only member of the order Stomiatiformes that has been recorded to
be hermaphroditic is the protandrous Gonostoma graci le Gunther (Kawa-
guchi and Marumo, 1967). In this species the testis develops as a se-
ries of swellings in the cord-like germinal ridge. After functioning
as a male, ovarian tissue develops between the- testicular parts and
eventually the entire gonad becomes an ovary. Such serial alternation
of male and female zones is commonly found as an abnormal condition
in gonochoristic species. Other Stomiatiformes such as Malacosteus,
Idiacanthus and Maurolicus are gonochoristic. As in the Cobi tidae, the
rare occurrence and distinctive anatomy of the single hermaphroditic
species suggest that hermaphroditism is a recent evolutionary experi-
ment that has not yet permitted its possessor to enter a new adaptive
zone and give rise to other species.
All Stomiatiform fishes are meso- or bathypelagic.
III. Aulopiformes
Rosen (1973) recognizes two suborders of aulopiform fishes, the Aulo-

poidei and the Alepisauroidei. All of the aulopoids that have been
studied, including representatives of the families Ipnopidae, Bathyp-
teroidae and Chlorophthalmidae, are synchronous hermaphrodites (Merrett
et al., 1973). The Alepisauroidei is divided by Rosen into two super-
families, the Synodontoidea and the Alepisauroidea. All of the six
families of Alepisauroidea are synchronous hermaphrodites (Mead, 1960;
Mead et al., 1964; Maul, 1971; Smith and Atz, 1973; Merrett et al.,
1973). The Synodontoidea however, includes two families that are gono-
choristic, the Synodontidae and the Harpadontidae, and one, the Gigan-
turidae, which is synchronously hermaphroditic (Smith and Atz, in prep.)
The critical point here is that the only two families of aulopiform
fishes that inhabit shallow waters are gonochoristic. All of the rest
are deep-sea forms and are synchronous hermaphrodites. This strongly
suggests that the synodontids and harpadontids are secondary gonocho-
rists.
The hermaphroditic gonads of aulopiform fishes are, in essence, com-

pletely separate testes and ovaries. The testes are solid organs of
the tubular type and they lie dorsal or dorsal-mesial to the ovaries.
In some species the sperm ducts are separate through their entire
length with separate openings to the exterior on either side of the
urinary papilla, which lies within a shallow cloaca-like depression
(Smith and Atz, in prep.). Although we suspect that gonochoristic go-
nads of synodontids and harpadontids are derived from hermaphroditic
structure, it may never be possible to determine this conclusively
because gonochorism could be achieved simply by loss of one set of
reproductive organs.
IV. Atheriniformes
Harrington (1967, 1971) has reported a remarkable case of synchronous

self-fertili zing hermaphrodites, Rivulus marmoratus Poey. Harrington has
done extensive experimental work on this species and has found that by
manipulating the environmental conditions, especially temperature,
during the phenocritical embryonic period he can produce "four sex
phenotypes. (1) Self-fertilizing hermaphrodites, (2) secondary males
from adult hermaphrodites, (3) false male gonochorists (secondary
males from juvenile hermaphrodites), (4) true male gonochorists (pri-
300
mary males or males at primary sex differentiation.)" Transformation

occurs with proliferation of testicular tissue that eventually fills
and obliterates the gonocoel of the ovotestis. In functional hermaph-
rodites the testicular portion or portions of the gonad remain rela-
tively small and variously located.
The Atheriniformes include many of the best known aquarium fishes and
some of these, of the families Poeciliidae and Cyprinodontidae, have
been studied extensively by fish breeders and geneticists. They ex-
hibit a wide spectrum of reproductive specializations including sexual
parasitism (Schultz, 1971) but so far, only Rivulus marmoratus has been
shown to be normally hermaphroditic.
V. Synbranchiformes
The synbranchiform fishes are elongate eel-like fishes that live in

swampy areas of the tropics and subtropics. Their phylogenetic affin-
ities are uncertain but they are generally considered to be a separate
order of the Acanthopterygii. Greenwood et al. (1966) divided the group
into two suborders, the Alabetoidei and the Synbranchoidei. Liem (1968)
concurred with this subdivision and expressed the opinion that "based
on an examination of its gonadal structure, Alabes is considered to have
no phylogenetic affinities with the Synbranchiformes." Recently Rosen
(pers. corom.) reports that Alabes is not a synbranchiform.
Within the Synbranchoidei Amphipnous euehia is gonochoristic with a single

ovary and paired, solid testes. Amphipnous is generally placed in its
own monotypic family.
The other three genera examined by Liem, Monopterus, Synbranehus and Typhlo-
synbranehus were found to be diandric or at least to have some diandric
populations. As in the Perciformes, primary males have solid testes,
secondary males have the old ovarian cavities remaining although not
functional. Chan and Phillips (1967) and Okada (1966) have described
the ovarian lumen as divided into three spaces which they term pseudo-
coels. From their illustrations it appears that the ovarian lamellae
are joined to the capsule wall along both dorsal and ventral edges
rather than just along the dorsal edge as is the usual situation in
cryptovarian fishes.
VI. Scorpaeniformes
The scorpaeniform fishes are also of uncertain taxonomic position,

some members sharing features of the Paracanthopterygii, while in most
respects they appear to be typical Acanthopterygian fishes. The only
group for which normal hermaphroditism has been reported is the family
Platycephalidae. Okada (1966, 1968), following the work of Aoyama et
al. (1963), has described sex reversal in Inegoeia meerdevoorti. According
to Okada the juvenile bisexual gonad passes through an ovarian phase
(presumably non-functional) as the testicular region degenerates.
Later, a testis develops from a different primordium and becomes func-
tional. Finally, the testicular tissue ceases to function and the gonad
becomes an ovary.
Fujii (1971) recognized three types of hermaphroditism among the Platy-

cephalidae: type A, in which the gonad develops first as a hermaphro-
ditic organ with only the testis functional and then transforms into
a female as the testicular portion ceases to function. A second type
(Ai) is essentially similar but the ovarian cavity is undeveloped until
301
the female phase. In the third type (B) the gonad develops first as
a testis with no indication of female elements. Only later does the
ovarian tissue appear and as the testis regresses the gonad appears
bisexual, finally becoming completely feminized. In type B the gonad
is U-shaped in cross-section during the testicular phase and only
closes to form an ovarian cavity during transformation. Thus, type Al
is intermediate between A and B.
The basic arrangement of the sexual tissues appears unique. In the

developing gonad the ovarian tissue is lateral to the testicular por-
tion but not separated by any well-defined structure. As the gonad
closes and becomes tubular, the ovarian epithelium comes to occupy
the region immediately surrounding the lumen and is in turn, surrounded
by the testis.
VII. Perciformes
The order Perciforrnes is very likely polyphyleti~. It is a "wastebasket"

taxon in that it encompasses a vast group of acanthopterygian fishes
that are not readily assigned elsewhere. It is large, including per-
haps 8,000 species in 20 suborders and more than 140 families. The
group is obviously in need of revision but because of the large number
of species it is difficult to assess the interrelationships of various
taxa or to define lineages.
Normal hermaphroditism occurs in some families in the suborders Percoi-

dei, Polynemoidei and Labroidei. In view of the tentative nature of
the classification one cannot attach too much importance to this sepa-
ration. There sean, however, to be several different lines of hermaph-
roditic patterns.
1. The Sparid Pattern
Sparid fishes cover the complete range of hermaphroditic function -

protandry, protogyny and synchronous hermaphroditism, although the
synchronous condition is usually considered transitory. Reinboth (in
Atz, 1965) has pointed out that in general, the protandrous species
lay pelagic eggs and lack sexual dimorphism and this process of sex
reversal requires several months, whereas the protogynous species tend
to produce adhesive eggs, exhibit sexual dimorphism and change from
female to male in as little as six weeks. Typically, the sparid gonad
is strongly territorial with connective tissue separating the male
and female zones. The male tissue is ventrolateral to the ovary al-
though sometimes it forms a groove in which the ovary lies. Emmelich-
thyid fishes have a similar gonad anatomy and three species have been
shown to be protogynous.
2. Pseudograrnrnidae
Hermaphroditism in Pseudogramma has been studied in Smith and Atz (1969).

In this species the testis forms a dorsal lobe that projects into the
common oviduct. Smith and Atz did not find any specimens with sexually
active testicular tissue although they examined specimens of various
sizes and from different times of the year. Apparently, in the oldest
i~dividuals there is progressive masculinization of the ovary immedi-
ately adjacent to the testicular lobe. This species also has a curious
invagination of the ventral wall of the common oviduct, the function
of which remains unknown.
302
3. Serranid Patterns
If the order Perciformes is a "wastebasket" the family Serranidae is

even more so. l<1embers of the genera Serranus, DipZ-ectrum and HypopZ-ectrus
are synchronous hermaphrodites with well defined ovarian and testicular
zones, although both tissues are contained within a common gonadal cap-
sule. EpinepheZ-us and Mycteroperca are protogynous hermaphrodites in which
the male and female elements are completely mixed. Rypticus, Sacura and
CheZ-idOperca have a similar intermediate pattern of gonad anatomy, in
which the male tissue is confined to the same region of the gonad as
it is in Serranus, but within that region male and female elements lie
side by side without an intervening connective tissue wall. Ryptieus,
Sacura and eke UdOperca are protogynous hermaphrodites.
Finally, several species appear to be secondary gonochorists. Smith

and Young (1966) discussed the finding that the testis of ParaZabrax
is a hollow structure exactly like that of a secondary male, but there
is no evidence of transformation nor were any males with solid testes
found. Liopropoma spp. also have hollow testes and may be descended
from hermaphroditic ancestors (Smith, 1971).
4. Labrid Patterns
Reinboth (1973) has discussed the special situation in the wrasses.

The species that he has studied have two kinds of males, primary males
that differentiate as males and have solid testes and secondary males
which begin life as females and later transform into males. Secondary
males are recognized by hollow testes, the cavity being the remains
of the ovarian lumen which is no longer functional. Sexual inversion
appears to be the same as in EPinepheZ-us; the germinal epithelium is
either a mixture of male and female elements or else there is only a
single phenolabile tissue that can produce either sperms or ova.
Members of the closely related family, Scaridae, also appear to be

diandric (Reinboth, 1968) and have similar breeding habits.
5. Cepolidae
Vives et al. (1959) have reported that the red bandfish CepoZa T'Ubescens
is diandric.
6. Polynemidae
PoZydactyZus indicus (Shaw), P. heptadactyZus and EZeutheronema tetradactyZum

have been shown to be hermaphroditic (Kagwade, 1970; Nyak, 1959;
Patnaik, 1967). According to the description and figure given by Kag-
wade, there are well-defined male and female zones separated by con-
nective tissue. The testicular tissue is in the dorsal part of the
gonad near the main blood vessels. Hida (1967) and Longhurst (1965)
have shown that some species of polynemids are protandrous. Apparently
the low proportion of recognizable hermaphrodites, as observed by
Kagwade and others, are transitional individuals that may function for
a short time as synchronous hermaphrodites.
The distribution of hermaphroditism among the various groups of fishes

certainly indicates a mUltiple origin. To suggest otherwise, would be
to ignore all anatomical evidence and to postulate that all gonocho-
ristic species above the ostariophysans are secondary gonochorists
303
and have become so independently. Neither of these alternatives is

acceptable.
Further evidence of the independent origin of hermaphroditism is the

vast difference in gonad structure in different groups. The gonads of
GOrlOstoma, Cobitis and RivuZus are all different and each is unique. Wi thin
the Perciformes the picture is less clear. In the Serranidae there
seems to be a logical progression from the patterns of Serranus through
Ryptiaus-Saaura to EpinepheZus. The gonad of Pseudogramma is unique and is
not obviously derivable from Serranus or any subsequent stage in the
serranid line. The structure of the secondary males of labrids and
scarids suggests a close alliance with the serranid line.
The Sparidae and Emmelichthyidae are closely related and have a similar
gonad structure, which is different enough from the serranid patterns
to suggest that they evolved hermaphroditism independently. The ventral
position of the testicular region seems to characterize this lineage.
The polynemids also have well separated male and female territories,
but according to Kagwade, the testicular region is near the main blood
vessels, i.e. dorsal to tQe ovarian region. Hence, despite their func-
tional similarity to the protandrous sparids, I believe that they also
represent a~ independent derivation of hermaphroditism.
Within the Synbranchiformes as it is now constituted, only the Syn-

branchidae are hermaphroditic. Although the anatomy of the single gonad
is unique in having the ovarian cavity divided into three chambers,
this might merely be a reflection of the extreme body elongation.
Otherwise, the gonad seems to be a mixture of male and female elements
and transformation is similar to that of EpinepheZus-like serranids. On
structural grounds it does not seem likely that the synbranchids are
closely related to the serranids and I would be inclined to regard
the gonadal similarities as the result of convergence. The evidence
is not strong, however.
The Platycephalidae are the only known hermaphrodites in the Scorpaeni-

formes. Their gonad structure is so different from that of any other
fishes that it is safe to assume that it originated independently.
In summary, it appears that hermaphroditism has originated among fishes

at least ten and possibly twelve times.
D. Discussion
The foregoing records make no claim to being a complete list of hermaph-

roditic fishes and there are certainly many more species to be found.
Transforming hermaphroditism in particular often can only be detected
through examination of large series of specimens of varying life his-
tory stages taken from different times of the year.
There can be little question that hermaphroditism in any form (includ-

ing secondary gonochorism) is a derived, specialized condition (apo-
morphic in Hennig's terms). As such, it is a valid indication of re-
lationship among groups that have it in the same form. The observed
difference in the anatomy of the gonad in various groups of fishes
provides ample evidence that hermaphroditism has arisen repeatedly in
unrelated groups of fishes.
304
I. The Ecology of Hermaphroditism
Liem (1968) has pointed out that hermaphroditism occurs in the fresh-
water swmap-dwelling synbranchids, in coral reef and other continental
shelf environments and in meso- and bathypelagic fishes and that each
of these environments is shared with many species that are gonocho-
ristic. Liem has called this chance evolution. Apparently he believes
that the selective forces are different in each of these environments
and thus hermaphroditism has been established in response to different
selective pressures in each setting.
One can also reason that each of these habitats, different though they
are, has some common aspect that makes hermaphroditism advantageous.
The overall habitat is unimportant; it is the precise aspect of the
habitat that is critical. One must also consider the possibility that
a particular group evolved in a different habitat from the one it now
occupies.
Arguments presenting selective advantage of hermaphroditism center on

two phenomena: (1) Increasing the number of zygotes produced; and (2)
increasing the chance that mating and fertilization will occur. These
aspects are not the same. In my 1967 paper I attempted to compare the
advantages of synchronous hermaphroditism with those of transforming
hermaphroditism and gonochorism using the argument that it was the
number of zygotes produced that was important. I now believe that this
argument is incomplete in that it neglects the question of selecting
for an optimum level of reproduction. Moreover, fecundity can be ad-
justed in much simpler ways - by increasing size of ovary, decreasing
the size of the individual eggs, and increasing turnover rates (i.e.
number of spawnings per season) all of which avoid the very consider-
able physiological problems of completely reorganizing the gonads.
1. Synchronous Hermaphroditism
The chief advantage of synchronous hermaphrodites is that every en-

counter between sexually active individuals is potentially a mating
encounter and this would seem to be of critical importance for low
density species that only rarely encounter another individual of the
same species. This would appear to be the situation in mesopelagic
and pelagic environments.
2. Successive Hermaphroditism
Protogynous and protandrous hermaphroditism Cas Liem correctly pointed

out) would seem to have the greatest advantage in situations where the
populations are periodically reduced to a very few individuals, as in
the case of synbranchids living in swamps. In such a situation, one
can postulate that protogyny would be favored using this reasoning:
Under the extreme stress of a drying swamp the advantage rests with
young adults. Small juveniles are subject to intense predation and
large old individuals require more space and are perhaps past their
prime and thus, have less tolerance for extremely adverse conditions.
A gonochoristic species with a 1 : 1 sex ratio would have in the most
extreme condition of only two individuals remaining, a 1/4 chance of
their both being males, a 1/4 chance of their both being females and
a 1/2 chance of their being 1 male and 1 female. A protogynous species
would have close to a 100% chance of having only females present but
if the switch to males could be moderated by social factors [as has
been demonstrated for Anthia$ (Fishelson, 1970) and LMroides (Robertson,
1972)J then reproduction would be assured. Even normal variation in
305
time of inversion would result in both sexes, hence social control is

not necessary.
Although it may be less obvious, the same arguments can be applied to

coral reef situations. Coral reef habitats are extremely patchy. The
large reef areas of the Indopacific consist of islands and atolls se-
parated by many kilometers of open water that does not offer a habitat
for coral reef fishes. Even within extensive reef areas such as the
Great Barrier Reef of Australia, individual coral patches and "bommies"
have very sedentary populations of the smaller fishes. Here the situa-
tion is very much like the swamp habitat - a series of limited popula-
tions consisting of only a few individuals. Social triggering of trans-
formation from female to male has already been demonstrated for two
such species and it will undoubtedly be found in many more, when ap-
propriate studies are carried out. Interestingly enough, the behavior
of the synchronous hermaphrodite Ser:mnus subUga:f'ius (Cope) supports
this idea. Although this species is functionally synchronous, in that
an individual will contain mature sperm and ripe eggs at the same time,
it undergoes courtship and behavior like a gonochorist and each indi-
vidual functions either as male or female during courtship and mating.
In the course of interact~ons between a single pair, however, the sex
roles are frequently reversed (Clark, 1959). Behaviorally at least,
this species is a transforming hermaphrodite, even though its gonad
structure is that of a synchronous hermaphrodite.
The "double gonad" structure of the spar ids offers a phenetic plas-
ticity that is not available to the mixed tissue gonads of EpinepheZus
and its relatives. The recent study by Mehl (1973) of the steenbras
Lithognathus lithognathus, seems to suggest that a single individual can
function only as a male or as a female during its life, although some
individuals may possibly also function as synchronous hermaphrodites.
Unfortunately, the phylogeny of the sparids has not been worked out,
but the observations of Reinboth (in Atz, 1965) that protandrous spe-
cies tend to have pelagic eggs and transform slowly, whereas the pro-
togynous species tend to lay demersal eggs and change quickly in a
period of a few weeks, certainly bears on the problem. Here again, in
the isolated but stable populations that produce demersal eggs, pre-
sumably with some parental care, the ability of some individuals to
transform to males would be advantageous - as it would always assure
the occurrence of large males in the population. If, as in certain
pomacentrids, large males guard the eggs there might be a strong ad-
vantage. Why protandry would be advantageous to a pelagic spawning
fish is less apparent, unless it maximizes egg production by larger
females while assuring an adequate supply of males. As Smith (1967)
pOinted out, the advantage of one or the other types of transforming
would depend to a large extent on mating patterns of polyandry or
polygyny.
3. Diandry
The situation that is perhaps the most puzzling is that of diandric

fishes, in which there are some males that differentiate as males
and others that pass through a functional female stage before trans-
forming into males. This situation is best known in the West Indian
wrasse ThaZassoma bifasaiatum, as a result of the thorough and ingenious
studies of Reinboth (1973), but it is also known in other wrasses.
In the blueheaded wrasse ThaZassoma bifasaiatum, Reinboth has found that

there are three color phases - plain yellow with a dark lateral stripe,
bright male (the blueheaded "super male") and transitional, an inter-
mediate color phase. Juveniles, females, primary and secondary males
are all found in the plain dress. Fish in the transitional and bright
color phases are either primary or secondary males.
Reinboth found that bright males are territorial and, although a dom-
inant bright male will permit other bright males in the territory,
only the dominant territorial male engages in pair spawning with in-
dividual females. Plain males participate in group spawning in which
large numbers of fish congregate in a local area and then spawn in
small temporary groups.
To this may be added the observations that TnaZassoma is a facultative

cleaner and that it is the most ubiquitous species of fish in the West
Indies, occurring in almost all reef and peripheral environments.
At first glance one might suggest that the diandric fishes represent
a transitional phase between protogyny and secondary gonochorism, as
postulated by Smith and Young (1966). In the Smith and Young model,
however, the reverpion to gonochorism would have resulted from part
of the population transforming to males earlier and earlier, until the
female phase was eliminated altogether. As a result, all of the males
would have the anatomical make-up of secondary males. Thus, the model
does not account for the presence of primary males and is not appli-
cable. Furthermore, this does not answer the question of the adaptive
significance of diandry, for diandry surely is highly adaptive. Not
only is it widespread in the labrids and scar ids which are extraordi-
narily successful coral reef fishes, but in the cepolids it apparently
represents a plesiomorphic character state rather than secondary or
independent development.
In the cleanep wrasses uwroides Robertson (1972) has demonstrated

social control of protogynous sex inversion. One result is to assure
that a territory is always occupied by a defending male. Presumably,
this is important to an obligate cleaner which depends on a steady
supply of fish to be cleaned. If the station were to be periodically
unoccupied, owing to the loss of a dominant male, fish would have to
go elsewhere to be cleaned and this would be disadvantageous to the
cleaner. It will be of great interest to know whether uwroides is dian-
dric or merely protogynous. We might predict the latter - the advan-
tage of primary males having disappeared with the development of the
socially controlled switch mechanism.
It appears then, that the situation with the bluehead and presumably
other wrasses and scarids as well, is a kind of balanced polymorphism.
Individuals that remain on a cleaning station have a better chance of
survival if they can switch. At the same time sexual inversion is
equivalent to early mortality of the first sex and it would be advan-
tageous for fishes that do not occupy territories to avoid switching.
There are, however, still many questions to he answered. For example,
are there two kinds of females, one capable of undergoing sex inver-
sion to produce males and the other not? What are the sex ratios at
the time of sexual differentiation? And perhaps most important, are
the behavioral patterns fixed so that some individuals are always
territorial and others always wanderers, or do they change back and
forth during their lifetime?
II. Evolutionary Opportunism
Interpretation of any feature, be it structure or mechanism, in an

evolutionary context is predicated on two assumptions: (1) That the
condition, or at least the capacity to develop the condition, is genet-
ically controlled, and (2) that its occurrence is evidence that it
307
confers some advantage on its possessor, or that it has been advan-

tageous at some time in the past. To these fundamental assumptions
must be added further observations. First of all, it is important to
bear in mind that most problems have several equally acceptable solu-
tions. In an ecological context, different ways of accomplishing the
same end are often advantageous solely because they are different,
and the difference itself becomes a mechanism for reducing direct
competition.
If synchronous hermaphroditism is a mechanism for increasing the num-

ber of zygotes two individuals could produce, it is only one means for
doing so. The question of why all organisms are not synchronous her-
maphrodi tes- (Atz, 1965) is not meaningful.
Attempts to explain hermaphroditism, or any other mechanism, must be

limited to the questions:
1. Under what conditions is this feature advantageous?
2. What has been the role of this adaptation in shaping the evolution-
ary trends of its possessor?
In conjunction with both questions it must be recognized that many

organisms have evolved in one habitat and later shifted to a new
habitat.
Mathematical models can often be invoked to clarify the interactions

between various adaptations. It is important, however, to keep in mind
that reproductive success cannot be measured by zygote production alone.
Evolutionary changes in adaptive features, as illustrated by the de-

velopment of new sexual mechanisms, are basically changes in the pro-
portion of individuals in the population bearing a particular genotype.
Each species is in dynamic equilibrium with the other species with
which it lives. A new adaptation that tremendously increased its re-
productive success would not necessarily be advantageous because com-
plex interactions with its predators, its prey, and the physical lim-
itations of the environment might actually work to its detriment. The
crucial aspect is not the absolute number of zygotes bearing the new
genotype but proportion of the successful zygotes of that species that
bear the new genotype.
Attempts to answer the question, under what circumstances is the con-

dition selectively advantageous, may sometimes lead to reasoning that
has the appearance of circularity. We argue, for example, that proto-
gyny maximizes zygote production because all members of the population
spawn first as females and then also argue that protandry maximizes
egg production because larger and more fecund individuals function
as females. Both reasonings, I think, are valid but until the critical
data, such as the patterns of fecundity and mortality for each age
g+oup are available, they will remain speculation. Like the mathemat-
ical models, the value of such speculation is that it helps us to see
what kinds of information are required for a real understanding of
hermaphroditism or any other adaptation.
Dr. James W. Atz has generously provided bibliographic

Aaknow~edgments.
and editorial advice and patiently listened to ideas as they developed.
Original investigations and gonad structure of aulopiform fishes were
funded by National Science Foundation Grant GB-7730. Drs. Donn E. Rosen
and Gareth Nelson have guided my concepts of phylogenetic relationships
in many ways. Mrs. Alice Lawson typed the manuscript and checked the
references. Dr. Rudolf Reinboth has been a constant inspiration and
all too seldom a stimulating field companion.
308
References
Aoyama, T., Kitajima, T., Mizue, K.: Study of the sex reversal of ine-
gochi, CocieUa cY'ocodila (Tilesius). Bull. Seikai Region Fish. Res.
Lab. 29, 1 3 - 33 ( 1 963) .
Atz, J.~ Intersexuality in fishes. In: Intersexuality in Vertebrates
Atz, J.W.: Hermaphroditic fish. Science 150, 789-797 (1965).
Brundin, L.: Transantarctic relationships and their significance, as
evidenced by chironomid midges with a monograph of the families
Podonominae and Aphroteniinae and the austral Heptagyiae. K. Svenska
Vetenskakad Handl., ser. 4, vol. 11, no. 1, 472 pp. (1966).
Chan, S.T.H.: Natural sex reversal in vertebrates. Phil. Trans. Roy.
Soc. London, B. 259, 59-71 (1970).
Chan, S.T.H., Phillips, J.G.: The structure of the gonad during natural
sex reversal in MonopteY'us albus (Pisces: Teleostei). J. Zool., London
151,129-141 (1967).
Clark, E.: Functional hermaphroditism and self-fertilization in a ser-
ranid fish. Science, N.Y. 129, 215-216 (1959).
Crowson, R.A.: Classification-alld Biology. XI + 350 pp. London: Heine-
man Education Books 1970.
(Teleostei, Anthiidae) regulated by the presence or absence of a
male fish. Nature (Lond.) 224, 90-91 (1970).
Fujii, T.: Hermaphroditism and sex reversal in fishes of the Platy-
cephalidae. II. Kwnococius detY'usus and Inegocia japonica. Japan. J.
Ichthyol. UL 109-117 (1971).
Ghiselin, M.T.: The evolution of hermaphroditism among animals. Quart.
Rev. Biol. 44,189-208 (1969).
Gosline, W.A.:~he morphology and systematic position of the alepo-
cephaloid fishes. Bull. Brit. Mus. Nat. Hist. (Zool.) ~, 183-218
(1969) .
Greenwood, P.H., Rosen, D.E., Weitzman, S.H., Myers, G.S.: Phyletic
studies of teleostean fishes, with a provisional classification of
living forms. Bull. Amer. Mus. Nat. Hist. 131, 339-456 (1966).
Greenwood, P.H., Miles, R.S., Patterson, C. (eds.): Interrelationships
of fishes. Suppl. No. 1 to the Zoological Journal of the Linnean
Society, vol. 53. London-New York: Academic Press 1973.
male gonochorists from eggs of the self-fertilizing hermaphroditic
fish, Rivulus maY'7110Y'atus Poey. Biol. Bull. ill, 174-199 (1967).
determine when self-fertilizing hermaphrodites of Rivulus maY'moY'atus
change into functional secondary males, with a reappraisal of the
modes of intersexuality among fishes. Copeia 1971, 389-432 (1971)
Hennig, W.: Phylogenetic systematics. Urbana: Univ. Illinois Press
1966.
Hida, T.S.: The distribution and biology of polynemids caught by
bottom trawling in Indian seas by the R/V Anton Bruun, 1963. J. Mar.
Biol. Ass. India 9, 281-299 (1967).
Kagwade, P.V.: Hermaphroditism in Polydactylus indicus (Shaw). J. Mar.
Biol. Ass. India 10, 399-401 (1970).
Kawaguchi, K., Marum~ R.: Biology of Gonostoma gY'acile (Gonostomatidae)
I. Morphology, life history and sex reversal. Informat. Bull. on
Plantol. Japan. Cornrn. No. Dr. Y. Matsue, pp. 53-67, pls. 1 and 2
(1967). (Included in Coll. Repr. The Oceanic Res. Inst. Univ. Tokyo.
Vol. VI, pp. 253-269, 1968).
Liem, K.F.: Geographical and taxonomic variation in the pattern of
natural sex reversal in the teleost fish order Synbranchiformes.
J. Zool., London 12£, 225-238 (1968).
309
Lodi, E.: Sex reversal of Cobitis taenia L. (Osteichthyes, fam. Cobi-

tidae). Experientia 23, 446-447 (1967a).
Lodi, E.: Inversione sessuali in Cobitis taenia L. (Cobi tidae, Osteich-
thyes). Archivo Zoologico Italiano 52, 129-135 (1967b).
Longhurst, A.R.: The biology of West African polynemid fishes. J. Con-
seil 30, 58-74 (1965).
Maul, G.~: Report on the fishes taken in Madeiran and Canarian waters
during the summer-autumn cruises of the "Discovery II" 1959 and
1961. III. Order Iniomi I. One toothless, sexually mature Anotopterus.
Bocagiana 28,1015 (1971).
Mead, G.W.: Hermaphroditism in archibenthic and pelagic fishes of the
order Iniomi. Deep-sea Research 6 (3), 234-235 (1960).
Mead, G.W., Bertelsen, E., Cohn, D.M.: Reproduction among deep-sea
fishes. Deep-sea Research 11, 569-596 (1964).
Mehl, J.A.P.: Ecology, osmoregulation and reproductive biology of the
whi te steenbras, Lithognathus lithognathus (Teleostei: Sparidae). Zool.
Africana 8 (2), 157-230 (1973).
Merrett, N.R~, Badcock, J., Herring, P.J.: The status of BenthalbeUa
infans (Pisces: Myctophoidei), its development, bioluminescence,
general biology and distribution in the eastern North Atlantic.
J. Zool., London 170, 1-48 (1973).
Nelson, G.J.: Gill arches and the phylogeny of fishes, with notes on
the classification of vertebrates. Bull. Amer. Mus. Nat. Hist. !il,
475-552 (1969).
Nelson, G.J.: Cephalic sensory canals, pitlines, and the classifica-
tion of esocoid fishes with notes on galaxiids and other teleosts.
Amer. Mus. Novit. 2492, 1-49 (1972).
Nyak, P. D.: Occurrencesof hermaphroditism in Polynemus heptadactylus
Cuv. & Val. J. Mar. Biol. Ass. India 1 (2), 257-259 (1959).
Okada, Y.K.: Observations on the sex reversal in the synbranchoid eel,
Pluta alba (Cuvier). Proc. Japan. Acad. 42, 491-496 (1966).
Okada, Y. K.: Sex reversal in Inegocia meerdevoorti with special reference
to repetition of hermaphroditic state. Proc. Japan. Acad. 42, 497
-502(1966). -
Okada, Y. K.: A further note on sex reversal in Inegocia meerdevoorti.
Proc. Japan. Acad. ii, 374-378 (1968).
Patnaik, S.: Hermaphroditism in the Indian Salmon Eleutheronema tetradac-
tylum (Shaw). CurrentSciencel§. (19),525 (1967).
Reinboth, R.: Diandric teleost species. Gen. Compo Endocrinol. 9, 486
(1967). -
li, 852-855 (1968).
Reinboth, R.: Intersexuality in fishes. In: Hormones and the Environ-
ment (ed. G.K. Benson and J.G. Phillips). Mem. Soc~ Endocrinol. 18,
pp. 515-543. London-New York: Cambridge Univ. Press 1970. --
Reinboth, R.: Dualistic reproductive behavior in the protogynous
wrasse Thalassoma bifasciatum and some observations on its day-night
changeover. Helgolander wiss. Meeresunters. 24, 174-191 (1973).
Robertson, D.R.: Social control of sex reversal:Ln a coral reef fish.
Science 177, 1007-1009 (1972).
Rosen, D.E.: Interrelationships of higher euteleostean fishes. In:
Interrelationships of Fishes (eds. P.H. Greenwood, R.S. Miles,
C. Patterson), Suppl. No. 1 to the Zoological Journal of the Lin-
nean Society, vol. 53, pp. 397-513. London-New York: Academic Press
1973.
Rosen, D.E., Greenwood, P.H.: Origin of the Weberian apparatus and
the relationships of the ostariophysan and gonorynchiform fishes.
Amer. 11us. Novit. 2428, 1-25 (1970).
Rosen, D.E., Patterson, C.: The structure and relationships of the
paracanthopterygian fishes. Bull. Amer. Mus. Nat. Hist. lil, 357-
474 (1969).
310
Schultz, R.J.: Special adaptive problems associated with unisexuality

in fish. Amer. Zool. 11, 351-360 (1971).
Smith, C.L.: Contribution-to a theory of hermaphroditism. J. Theoret.
Biol. 17, 76-90 (1967).
Smi th, C. r:;-:-: Secondary gonochorism in the serranid genus Lioproporna.
Copeia 1971,316-319 (1971).
Smith, C.L., Atz, E.H.: The sexual mechanism of the reef bass, Pseudo-
gramma bermudensis and its implications in the classification of the
Pseudogrammidae (Pisces: Perciformes). Z. Morph. Tiere 65, 315-326
(1969).
Smith, C.L., Atz, E.H.: Hermaphroditism in the mesopelagic fishes
Omosudis lowei and Alepisaurusferox. Copeia 1973 (1),41-44 (1973).
Smith, C.L., Young, P.: Gonad structure and the reproductive cycle of
the kelp bass, Paralabr= clathratus (Girard), with comments on the
relationships of the serranid genus Paralabr=. Calif. Fish Game
52, 283-292 (1966).
Vives, F., Suau, P., Planas, A.: Sobre la biologia de la cinta. In-
vestigacion pesq. li, 3-23 (1959).
Warner, R.R.: The adaptive significance of sequential hermaphroditism
in animals (in press) .
Unisexual Female Offsprings in the Salamander,
Pleurode/es waltlii Michah
A. Collenot
A. Introduction
The early stages of the organogenesis of ovaries and testes are diffi-
cult to study in bisexual species, when the genotypic sex of the ex-
perimental embryos or larvae is unknown. However, such studies are
relatively simple in mammals, owing to the precocious appearance of
sex chromatin before the sex differentiation stage of the embryo
(Picon, 1965; Issa et al., 1969); and in birds, through the prepara-
tion of the karyotype from embryonic cells (Gasc, 1973). In amphibians
and especially in the Salamander Pleurodeles waUlii, these characteris-
tics are lacking. Neverthe~ess, the genotypic sexes can be determined
by using the unisexual offsprings of Xenopus laevis (Breuer et al.,
1966). In Pleurodeles, Gallien (1954) obtained the first "neo-females"
(they were functionally phenotypic females but genotypic ZZ males}
which gave rise to unisexual male offsprings when mated with standard
ZZ males. In the same species, I have obtained females of the sexual
genotype Wfl, which give rise only to female individuals when mated
with standard males.
There are three types of females presently known in Pleurodeles, each

type being distinguished by the sex-ratio of the offspring after mating
with standard ZZ males:
(1) The arrhenogenous females (or "neo-females", according to Gallien),

their sexual genotype being ZZ and the sex-ratio 100% i f ; (2) the thely-
genous females, their sexual genotype being WW and the sex-ratio 100%~
and (3) the amphogenous females (or standard females), their sexual
genotype being ZW and the sex-ratio 50% if and 50% ~ .
I have previously obtained the first thelygenous female in Pleurodeles

(Collenot, 1973), and in this report I will present the experimental
procedure by which several thelygenous females have been obtained.
The embryos used in the experiments belonged to a strain of Pleurodeles

waltlii, which has been introduced and bred in the laboratory by Gallien
(1952). In this species the gonads are differentiated at metamorphosis.
The larvae from standard mates had a sex-ratio very close to one.
Spontaneous intersexuality has never been observed through the usual
breeding conditions at the laboratory.
312
I. Experimental Procedure
I have used a method which differs slightly from that followed success-
fully for the first time in Amhystoma, sp. by Humphrey (1942). Essentially
it consists of masculinizing the embryonic ovary of a genotypic female
into functional testis. In Pleurodeles (as in Ambystoma) the female sex
is the heterogametic one (ZW). Therefore, such a testis would produce
two types of spermatozoa and, in cross-breeding with an amphogenous
ZW female, would produce offspring consisting of ZZ males (25%), ZW
females (50%) and WW animals (25%) (provided that the latter are viable,
as in Amhystoma sp., Humphrey, 1945) and in Xenopus laevis (Mikamo and
Witschi, 1963). Here, they are functional females giving rise to uni-
sexual female offspring, when mated with standard ZZ males.
The experiment comprises five steps:
1. Grafting of the Embryonic Gonad. The right intermediate mesoderm (Fig. 1)

is orthotopically grafted at stage 24 (normal table of Pleurodeles,
Gallien and Durocher, 1957) according to the method previously described
in that species (Lacroix and Capuron, 1966). At that stage of develop-
ment, the primordial germ cells are located in the grafted tissues
(Maufroid and Capuron, 1972). The sex of the donor embryos (which are
derived from an amphogenous female) is unknown at that time, but all
the receptor embryos are males derived from an arrhenogenous female.
An incomplete extirpation of the right gonad primordium from the re-
ceptor would lead to the mixing of different kinds of germ cells (ZZ
and ZW) on that side. To avoid such a risk, the host embryos were pre-
viously triploidized by a cooling shock administered at the time of
egg deposition, according to the method described by Beetschen (1960).
The remnant triploid germ cells could then develop into spermatozoa.
These spermatozoa, by fertilizing the eggs from diploid females, would
produce lethal, or exceptionally viable, aneuploid embryos (Beetschen,
1960; Lacroix, 1967; Guillemin, 1972). Thus, when crossing a triploid
receptor male with a standard diploid female, any viable larva can be
considered as produced from the grafted diploid germ cells.
THE HOST : c1(2A+ZZ) ~ o--(3A+ZZZ)

(early cooling)
AFT : ~ (2A+ZW)
c1 [(3A+ZZZ) or (2A+ZW)] x
t~ (2A+ZW)
aneuploid offspring diploid offspring

(exceptionally viable) 1r-
25-%-0"-(-2-
A-+-
ZZ-)-7-5o/c- 'o' ~ [SO%(2A+ZW)+25% (2A+WW)]
Fig. 1. Scheme of the experimental procedure

313
2. Recognition of the Sex of the Grafted Gonads. The experimental pairs are
numbered, and donor and host embryos are bred separately, until they
metamorphose. Dissection of the donor then shows what kind of gonad
(testis or ovary) has been grafted, and also if the gonadal primordium
has been entirely extirpated.
3. Control of the Hosts. At puberty (about 3 months after metamorphosis)
when nuptial pads appear on the forelegs, the hosts corresponding to
female donors were laparotomized on the right side, in order to de-
termine the degree of development and the external features of the
gonad which had developed from the graft. The testis was removed from
the left side and used to make squashes of testicular cells, in order
to determine the host's ploidy.
4. Mating of the Triploid Hosts. When triploid hosts, carrying a testis on
the right side, are mature they are mated with unfertilized standard
diploid females. The sex-ratio of the consecutive offsprings are estab-
lished two months after metamorphosis, through exploratory laparotomy.
5. Specification of the Thelygenous Females. Females from mature offsprings,
where the sex-ratio is about 25% if /75% <jl , are mated with standard
diploid males. Samples of their offsprings are bred until metamorpho-
sis, when diagnosis of the sex is a simple procedure. Then, any re-
sulting unisexual female offspring specifies its mother as a thely-
genous female.
II. Method for Squashing Testicular Cells
Owing to procedural difficulties, it is impossible to inject colchicine

into males before removal of the testis. Therefore, the left testis is
taken out and cut longitudinally into two equal parts which are then
immersed for about two hours in 0,5% colchicine (Calbiochem or Sigma)
dissolved in 10% Steinberg medium. The pieces of testis are then rinsed
in distilled water for 10 to 15 min and fixed for 15 min in 45% acetic
acid. The testicular tissue is dissociated with needles on a micro-
slide, the diasaggregated cysts are then covered with a cover slip and
squashed. After freezing on dry ice (from 5 to 10 min), the cover slip
is quickly removed and the squashed cells are stained in acetic orcein
for 30 min (1% orcein - G.T. Gurr - in 45% acetic acid). The microslide
is rinsed, dehydrated in absolute ethanol and mounted (Euparal). It is
advisable to make several squashes from the same testis; triploidy can
then be easily indentified (Fig. 2).
Fig. 2 a-c. Chromosomes of primary spermatocytes from testicular squashes;

(a, b) from the diplOid R25 (12 bivalents; in Pleurodeles waltlii, 2n = 24);
(c) from the triploid R41 (10 trivalents plus 2 bivalents and their cor-
responding univalents, arrow heads)
314
C. Results
Thirty-nine grafts have been made and 17 experimental pairs have been
bred until metamorphosis. The sex-ratio of the 17 donors was 8 d' : 9 '2 •
Ploidy could be specified from only 5 hosts, out of the 9 corresponding
to the 9 female donors: 2 were diploid, 2 were triploid and the last
one (R35) exhibited a 2n/3n mosaicism. Moreover, 8 hosts (the corres-
ponding donors of which died during the first larval stages) were also
bred and studied: 2 were diploid and 6 were triploid (Table 1).
Table 1. Summary of the experimental design
Donor embryos Ploidy of corresponding host embryos

(sex)
2n 3n 2n/3n unknown
9 '1' RI9 - R25 R4 - R32 R35 Rs - RIO
R37 - R3S
8d' RI6 R29 R2 - RI2
. RI 3 - R2I
R33 - R39
8 (unspecified) R42 - R45 R4I - R43
R44 - R46
R47 - R 4S
The exploratory laparotomy on the right side had shown, either a testis
(R 4 , R4 1, R43, R46, R 47 , R4S) in 6 animals, or a whitish and filamen-
tous gonad in 3 animals (R 32 , R"5' R 44 ).
The fourth step of the experiment was the mating of the triploid males.
At that time, a testis had developed on the right side in males R35
and R44i but the right gonad of R32 remained in a vestigial state until
the death of the animal. Male R43 , which died, could not be tested.
The results of this stage are shown in Table 2. Encouraged by the re-
sults of the examination of the sex-ratio of the offsprings from'4
males, I tested (during the fifth and last step of the experiment) ,
the females as soon as they matured. There was, therefore, a great
disparity between the different offsprings according to the number of
tested females (Table 3). At the present time, 4 females out of the
25 tested appeared to have been thelygenous.
Table 2. Breeding experiments with triploid males
Males Corresponding offsprings (number and sex-ratio)
R4 I L69 (1 9 d', 81 '1' ) i M71 (6 d', 1 4 '1' ) i BB 73 (6 d', 17 '1' )

R44 2 spawns, no fertilization
R4 6 °
N6 9 ( 1 d', 38 '1' )
R47 1 spawn, no fertilization
R4 S 0 69 (14 d', 32'1' )
R4 2 spawns, no fertilization
R35 R69 (6 d', 10'1')
315
Table 3. Tested females from the offsprings listed in Table 2
Number of Corresponding females

offspring
amphogenous thelygenous
2 0
2 2
2 0
10 0
5 2
D. Discussion
Unavoidable problems resulting from the long duration of the experi-

ment led me to modify the strict experimental procedure I had initially
established. According to that procedure only males R4 and R32 would
have been suitable for mating. However, male R32 did not carry any
right testis. When male R4 twice failed to fertilize eggs, I attempted
artificial fertilization. Simultaneously, I used eggs from 3 diploid
females (217, 301, 302) and spermatozoa from R4 ; from another triploid
male, R29 (receptor of a male diploid gonad); and from a standard dip-
loid male, T. The results of the experiment are shown in Table 4. They
clearly indicate, that the fertilizing spermatozoa from the right tes-
tis of R4 had developed from triploid germ cells and had given rise
to lethal aneuploid offspring. The difference in lethality between
the offspring from d' T and d'R2 9 could be due to the development of
spermatozoa in the R29 right testis, from a mixture of diploid and
triplOid germ cells.
Table 4. Breeding experiments with standard diploid male (T), triploid

males and diploid females by artificial fertilization
Artificial Number of Number of Lethal per-

fertilization embryos larvae centage at
batches (early blastula) (feeding stage) feeding stage
d'T x ~ 217 41 32 22
d' 29 x ~ 217 a 85 51 40
d' 29 x ~ 301 a 49 30 39
d' 29 x ~ 302 30 14 53,5
if 4 x ~ 302 b 30 1b 96,5
if 4 x ~ 217 40 0 100
a 50 larvae from these 2 batches have been bred; 42 reached metamor-
phosis, the sex-ratio was: 19d' /23 ~ .
b This larva died soon
after the feeding stage.
During his numerous experiments, Beetschen (1960) found only one animal
exhibiting 2n/3n mosaicism and, moreover, this animal was not viable.
It is interesting to note that Beetschen controlled ploidy on somatic
cells, while I did this on germ cells. Therefore, one cannot exclude
that in the host R35 , some diplOid primordial germ cells could have
316
crossed from the graft on the right side to the left side. Therefore,
the host could have been triploid, while the squashes of its left
testis showed simultaneously diploid and triploid meiotic germ cells.
Host R3S has given rise to 2 thelygenous females.
Among the triploid male hosts (the sex of the corresponding donors
being unknown), R41 has also given rise to 2 thelygenous females. An
abnormal, but relatively slight lethality exhibited by its first off-
spring after it had reached the feeding stage (L 69 ) could be assumed
as due to aneuploidy. Another host, R44 , first had a vestigial right
gonad which developed into a testis (similar to R 3S ) after the removal
of the left testis. Similar facts have been described in Anibystoma, sp.
by Humphrey (1945). In Pleurodeles waltlii, I have begun to study (by
con~rolled heterosexual associations), how a genotypic female gonad
can be masculinized into a functional testis. Preliminary results show
that the grafting of either a testis primordium on a female embryo,
or a testis from a metamorphosing animal into a female larva before
the sex differentiation stage, essentially induces a severe inhibition
of the development of the female gonads. The sex inversion is seldom
obtained and when the grafted testis is removed, mature ovaries can
develop from the previously inhibited female gonads. These observations
confirm those obtained in the same species with heterosexual parabiotiC
associations (Gallien, 1973). It appears that sex inversion of ovaries
into testes is more difficult to obtain in Pleurodeles waltlii than in
Ambystoma, sp. or Xenopus laevis.
The arrhenogenous and thelygenous females obtained in Pleurodeles waltlii

show (as previously demonstrated in other amphibians and in fishes)
that germ cells can develop into either type of gametes, regardless
of their genotyp~. Germ cells of birds and mammals behave differently,
as has been shown in the chicken (Haffen, 1969) and in the mouse (Tar-
kowski, 1969).
This discrepancy could be related to the fact that, in amphibians, it

is very difficult to detect a sex chromosome heteromorphism. Lacroix
(1970) has shown that such a heteromorphism exists in Pleurodeles poireti
lampbrush chromosomes but, as yet, he could not detect it in Pleurodeles
waltlii, even when comparing the lampbrush chromosomes from arrheno-
genous (ZZ) and thelygenous (WW) females to those from amphigenous
(ZW) females.
At it has been previously shown in Xenopus laevis (Witschi, 1971), one

is now able to freely experiment with the unisexual male or female
offspring of Pleurodeles waltlii, especially in histochemical and bio-
chemical studies of the early stages of sex differentiation.
References
Beetschen, J.C.: Recherches sur l'heteroploidie experimentale chez un

Amphibien Urodele, Pleurodeles waltlii 1-1ichah. Bull. Biol. France et
Belgique 93, 12-127 (1960).
Breuer, H., Dahm, K., Mikamo, B., Hitschi, E.: Differences in steroid
metabolism of male and female larvae of Xenopus laevis. In: Second
International Congress on Hormonal Steroids, International Congress
Series, no. 111, 215. Amsterdam: Excerpta Medica Foundation 1966.
Collenot, A.: Obtention, par la methode des greffes de gonades embry-
onnaires, d'une femelle a descendance unisexuee femelle, chez le
Triton Pleurodeles waltlii lUchah. Experientia~, 885-887 (1973).
317
Galli~n, L.: Elevage et comportement du Pleurodele au Laboratoire.

Bull. Soc. Zool. France 77, 456-461 (1952).
Gallien, L.: Inversion experimentale du sexe, sous l'action des hor-
mones sexuelles, chez Ie Triton PZeurodeZes waUUi Michah. Analyse
des consequences genetiques. Bull. BioI. France et Belgique 88,
1-51 (1954). -
Gallien, L.: Differenciation et organogenese sexuelle des Metazoaires.
Paris: Masson et Cie 1973.
Gallien, L., Durocher, M.: Table chronologique du developpement chez
neurodeZes waUUi. Bull. BioI. France et Belgique 91, 97-114 (1957).
Gasc, J.M.: Sur les resultats d'une technique d'identification precoce
du sexe genotypique des embryons d'oiseaux. C.R. Acad. Sci. 277,
1925-1928 (1973). -
Guillemin, C.: Sur un male trisomique fertile chez l'Amphibien Urodele
PZeurodeZes waUUi Michah. Analyse de sa descendance. C.R. Acad. Sci.
275, 1895-1898 (1972).
Haffen, K.: Incompatibilite entre determination chromosomique et sti-
mulation hormonale dans la differenciation sexuelle. Ann. Embryol.
Morph., suppl. I, 223-235 (1969).
Humphrey, R.R.: Sex of the offspring fathered by two AmbZystomafemales
experimentally converted into males. Anat. Rec. 82, 77 (1942).
Humphrey, R.R.: Sex determination in amblystomid salamanders, a study
of the progeny of females experimentally converted into males.
Amer. J. Ana t. 7 6, 3 3 - 6 6 ( 1 94 5) .
Issa, M., Blank, C~., Atherton, G.W.: The temporal appearance of sex
chromatin and of the late-replicating X chromosome in the domestic
rabbit. Cytogenetics 8, 219-237 (1969).
Lacroix, J.-C.: Obtention de femelles trisomiques fertiles chez l'Am-
phibien Urodele neurodeZes waUUi Michah. C.R. Acad. Sci. 264, 85-88
(1967). -
Lacroix, J.-C.: Mise en evidence sur les chromosomes en ecouvillon de
neurodeZes poireti Gervais, Amphibien Urodele, d' une structure liee
au sexe, identifiant Ie bivalent sexuel et marquant Ie chromosome W.
C.R. Acad. Sci. 271, 102-104 (1970).
Lacroix, J.-C., Capuron, A.: Localisation et greffe des cellules germi-
nales primordiales chez neurodeZes waUUi Michah. (Amphibien Urodele) •
Preuves cytogenetiques. C.R. Acad. Sci. 263, 1244-1247 (1966).
Maufroid, J.P., Capuron, A.: Migration des cellules germinales primor-
diales chez I' Amphibien Urodele PZeurodeZes waUUi Michah. Roux'
Arch. Entw. Mech. 170, 234-243 (1972).
Mikamo, K., Witschi, E.: Functional sex-reversal in genetic females
of Xenopus Zaevis, induced by implanted testes. Genetics 48, 1411-
1421 (1963).
Picon, R.: Chromatine sexuelle du foetus de rat et determination pre-
coce du sexe. Arch. Anat. micro Morph. expo 2&, 281-290 (1965).
Tarkowski, A.K.: Consequences of sex chromosome chimerism for sexual
differentiation in mammals. Ann. Embryol. Morph., suppl. I, 211-
222 (1969).
Witschi, E.: Mechanisms of sexual differentiation. Experiments with
Xenopus Zaevis. In: Hormones in Development (eds. M. Hamburgh and
E.J.W. Barrington), pp. 601-618. New York: Appleton Century Crofts
1971 •
Autoradiographic Study on the Mechanisms of
Testosterone-Induced Sex-Reversal in Rana Tadpoles *
E. Vannini, A. Stagni, and F. Zaccanti
A. Inhibition by Actinomycin D or Puromycin of the Testosterone-

Induced Sex Reversal
It is well known that testosterone masculinizes the ovaries of Rana

tadpoles, so that they become testes. However, the mechanisms by which
this phenomenon occurs ramain obscure. In a first series of our exper-
iments to clarify this situation, the effects of a simultaneous admin-
istration of testosterone and actinomycin D to Rana tadpoles have been
tested. Actinomycin D is an antibiotic which specifically inhibits any
DNA-dependent RNA-transcription.
The first experimen"ts were performed on tadpoles of a sexually undif-

ferentiated race of Rana daZmatina (Vannini and Stagni, 1967, 1971,
1972). Before the end of metamorphosis these gonads always develop as
ovaries, both in the male and female genotypes; and only later in the
male genotypes do they spontaneously become testes.
Plate I shows the results of a group of treatments precociously begun

when the tadpole gonads were very young ovaries - still showing a com-
pact medullary tissue and a cortex provided only with oogonia (Fig. 1).
In the control specimens these gonads normally developed into ovaries
within about two weeks, showing a more or less reduced medulla and a
cortex full of oogonia and very young oocytes at the first synaptic
stages of premeiosis (Fig. 2). The same result occurred in experimental
groups treated for two weeks with a sublethal dose of actinomycin D
(Fig. 3). However, in experimental groups treated with testosterone,
the androgenic hormone completely hindered the development of oocytes
in the ovarian cortex (Fig. 4) and stimulated a masculinizing activity
in the medullary tissue. This actively proliferated, attracting the
oogonia from the cortex and surrounding them (Fig. 5). After passing
into the medulla, the oogonia began to multiply as spermatogonia
¥This research was financed by the "Consiglio Nazionale delle Ricerche"

(Italy) .
Plate I. Rana daZmatina tadpoles; cross-sections of gonads

Fig. 1. Control animals; young ovary with oogonia. x 450
Fig. 2. Control animals; ovary with synaptic oocytes. x 450
Fig. 3. Actinomycin D treated animals (1,25 ~g/ml of tap-water); ovary
with synaptic oocytes. x 450
Fig. 4-6. Testosterone treated animals (0,1 ~g/ml of tap-water); 00-
cytes are not present, and the activated medulla attracts germ cells.
x 450
Fig. 7 and 8. Actinomycin D plus testosterone treated animals (same
doses as above); oocytes are not present, but the medullary tissue is
not active. x 450
319
Plate I. Fig. 1 - 8
320
(Fig. 6). Finally, in experimental groups simultaneously treated with

actinomycin 0 + testosterone, the gonads (Fig. 7, 8) after treatment,
were small ovaries containing only oogonia (and not oocytes) in their
cortex, and with an unactivated medulla never attracting germ cells.
Evidently, actinomycin 0 did not suppress the anti-oogenetic activity
of testosterone, but completely suppressed its medullo-stimulating
activity.
Plate II demonstrates the results of another treatment when, in pre-

metamorphic tadpoles, the gonads appeared as well-developed ovaries.
In the male genotypes, these ovaries showed a central medulla with
cavities and a cortex filled with oogonia and young oocytes, engaged
in synaptic stages of pre-meiosis. However, the ovarian cortex was
much more enlarged in the female genotypes, as most of the germ cells
reached the stage of large diplotenic developing oocytes. Consequently,
the central cavity lined by the residual medulla, was greatly reduced
in size by compression. These two different aspects of the male and
female genotype ovaries were not significantly modified (except for
scattered degeneration of some growing oocytes) by a three week treat-
ment with a sub-lethal dose of actinomycin 0 (Fig. 9, 10). In contrast,
an equally long period of treatment with testosterone constantly pro-
voked a sex reversal of the ovaries. In the male genotypes, this mas-
culinizingprocess precociously begins with a massive proliferation
of medu!lary tissue filling the ovarian cavity, and with a more or
less simultaneous degeneration of the synaptic cortical young oocytes
(Fig. 11). Shortly after, the preserved cortical oogonia actively
penetrate the new-formed medulla, multiplying here as spermatogonia
(Fig. 12). In the female genotypes the very same phenomena of medullary
activation, oocyte degeneration and spermatogonia multiplication oc-
curred more slowly. After about three weeks of testosterone treatment,
some remnants of the central cavities and several big diplotenic 00-
cytes still persisted in the largely masculinized ovaries (Fig. 13).
However, by associating the activities of actinomycin 0 and testos-
terone, both in the ovaries of the male (Fig. 14) and female (Fig. 15))
genotypes, only a severe oocyte degeneration resulted, while prolifera-
tion and activation of the medullary tissue was almost completely
lacking. Consequently, the medulla remained poorly developed and with
a large cavity in its central part, as is the case in normal ovaries.
Confirmation was also obtained showing that actinomycin 0 did not
suppress the anti-oogenetic activity, but did suppress the medullo-
stimulating activity of testosterone.
Other experiments likewise in Rana dalmatina tadpoles (Vannini and Stagni,

1968), have been directed to test the effects of a simultaneous admin-
istration of testosterone and puromycin. Puromycin is another anti-
Plate II. Rana dalmatina tadpoles; cross-sections of gonads. ~
Fig. 9. Actinomycin 0 treated animals (2,5 ~g/ml of tap-water); ovary

of male genotype with a cavity. x 270
Fig. 10. Actinomycin 0 treated animals (same dose as above); ovary of
female genotype with a cavity. x 150
Fig. 11 and 12. Testosterone treated animals (1 ~g/ml of tap-water);
induced sex reversal in male genotypes. x 270
Fig. 13. Tes~osterone treated animals (same dose as above); induced
sex reversal in female genotype. x 270
Fig. 14 and 15. Actinomycin 0 plus testosterone treated animals (same
doses as above); both in male and female genotypes the medullary-tissue
activation is lacking, and oocytes are degenerating. Fig. 14: x 270;
Fig. 15: x 180
321
Plate II. Fig. 9 - 15

322
bioti~ specifically inhibiting any messenger-RNA-dependent protein

synthesis at the level of molecular translation on cytoplasmic poly-
somes. For these experiments the treatment covered the period of two-
three weeks covering the completion of metamorphosis. At the end of
the experiments (Plate III), the newly-metamorphosed specimens of male
genotype had, in the control group, ovaries at the beginning of spon-
taneous sex-reversal (Fig. 16), with a cortex provided only with oogo-
nia and small synaptic oocytes (sometimes degenerating) and with a
new proliferating medulla, more or less completely filling their cen-
tral part. However, control female genotypes of the same age showed
more developed ovaries (Fig. 17), with a cortex crowded with many big
developing diplotenic oocytes, and with no evidence of new proliferat-
ing masculinizing medullary tissue in their central part. These well-
developed ovaries are never modified in female genotypes, even after
a two-three week treatment with a sublethal dose of puromycin. More-
over, the very same treatment in male genotypes almost completely in-
hibits the spontaneous start of sex reversal (Fig. 18). Here, the
poorly developed ovaries persist with cavities in their central part,
but without any significant new proliferation of masculinizing medul-
lary tissue. An equally long period of treatment with testosterone
definitely masculinizes the ovaries, both in male and female genotypes.
In male genotypes (Fig. 19), the gonads preserve several small synaptic
oocytes of t~eir old cortex, but in their central part, many oogonia
migrate to the very abundant newly proliferated medullary tissue -
there to multiply as spermatogonia. In female genotypes (Fig. 20) a
similar intersexual aspect of the gonad is detectable, but several
big developing diplotenic oocytes still persist. Finally, in the cases
of simultaneous administration of puromycin and testosterone, the for-
mer at least partially inhibits the medullo-stimulating masculinizing
activity of testosterone. In the gonads of male genotypes (which under
normal conditions would also have been subjected to a spontaneous sex
reversal) the testosterone-induced masculinization appears to be barely
repressed by the hindering activity of puromycin. In fact, in these
gonads only scattered spermatogonia are present (Fig. 21) within a
poorly developed medullary tissue containing many lacunae; while in
the cortex, many small synaptic oocytes persist. However, the testos-
terone-induced masculinizing process in the gonads of female genotypes
(Fig. 22) is much more significantly repressed. Here, the well-pre-
served cortex persists, with many growing oocytes, while the very
sparse newly proliferated medullary tissue merely lines the large re-
Plate III. Rana dalmatina tadpoles; cross-sections of gonads. ~

Fig. 16. Control animals; gonad of male genotype at the beginning of
spontaneous sex reversal. x 190
Fig. 17. Control animals; ovary of female genotype. x 85
Fig. 18. Puromycin treated animals (5 ~g/ml of tap-water); ovary of
male genotype lacking new proliferation of medullary tissue. x 190
Fig. 19. Testosterone treated animals (1 pg/ml of tap-water); induced
sex reversal in male genotype. x 190
Fig. 20. Testosterone treated animals (same dose as above); induced
sex reversal in female genotype. x 190
Fig. 21. Puromycin plus testosterone treated animals (same doses as
above); spontaneous and induced sex reversal is partially repressed
in male genotype. x 190
Fig. 22. Puromycin plus testosterone treated animals (same doses as
above); induced sex reversal is repressed almost completely in female
genotype. x 155
323
Plate III. Fig. 16 - 22

324
sidual ovarian cavity, but does not fill it. This sparse medulla seems
to be rather poorly activated towards a masculinizing tendency, as
shown by the few oogonia which had migrated there and were now multi-
plying as spermatogonia. Therefore, puromycin is able to inhibit the
masculinizing phenomena experimentally induced by testosterone, al-
though to a lesser degree than actinomycin D.
B. Working Hypothesis on the Mechanism of Masculinizing Activities of

Testosterone
On the basis of the previously described results, a tentati v e working

hypothesis is suggested, based on a scheme outlining the conditions
of Rana tadpole gonad (Fig. 23).
(I"slosl er one)
~----------
DNAc!' -',
.~',j
( acllnomycln/
m RNAc!'
pc!'
m"dullarin
Fig. 23. Scheme of the hypothet-

ical mechanism of testosterone
masculinizing action
In this scheme DNA if and DNA~, symbolize male-genes and female-genes

respectively. Normally, in the sexually undifferentiate d race s (as
have been used) the male-genes would persist, for the whole pre-meta-
morphic period, being always repressed in the male as well as in the
female genotypes. Moreover, in the same period the female-genes would
be spontaneously de-repressed, not only in the female but also in the
male genotypes. The expression of these female-genes, would mean an
activation of the nuclei of somatic cells in the gonadal cortex and
a transcription of a female messenger-RNA (m RNA ~ ). This would then
be r e sponsible for the translation of a female protein (P~ ; Witschi's
"corticin"?) endowed with a feminizing activity on the germ cells con-
tained in the gonadal cortex.
325
The sex reversal, experimentally induced by testosterone, depends

mainly on the fact that this steroid (or some of its metabolites;
probably bound to a cytoplasmic hormone-receptor) would de-repress
the normally latent male-genes in the nuclei of somatic target-cells,
thus activating them to proliferate as gonadal medullary tissue. The
expression of the de-repressed male-genes would then consist in the
transcription of a male messenger-RNA (m RNAif) responsible for the
translation of a male protein (Pif; Witschi's "medullarin"?) endowed
with a masculinizing activity on the germ cells, when migrating into
the gonadal medulla.
Our interpretation agrees with the modern view of the physiological

mechanisms of steroid hormone action (Karlson, 1965; Smellie, 1971),
and is supported by the above-described effects of simultaneous ad-
ministration of testosterone and actinomycin D or puromycin. In fact,
these antibiotics more or less completely suppress the medullo-stim-
ulating activity of testosterone, as the consequences of the male-gene
de-repression would be blocked by actinomycin D at the level of the
DNA-RNA transcription; and by puromycin at the level of the RNA-protein
translation.
As previously shown, another effect of testosterone is the anti-oogen-

etic one, which results in blocking the first appearance of synaptic
oocytes (in the cases of very precocious treatments), or in inducing
oocyte degeneration (in later treatments). As mentioned, this anti-
oogenBtic effect is not hindered by actinomycin D or puromycin. There-
fore, it would not be mediated by a de-repressing activity at the level
of gene-expression, but would probably depend on a direct disturbing
action of testosterone on the oogonia or oocytes developing in the
ovarian cortex (as symbolized by the hatched arrow in Fig. 23).
C. Autoradiographic Study on the RNA-synthesis during Testosterone-

induced Sex Reversals
According to our hypothesis, an increased synthesis of RNA would char-

acterize the testosterone target-cells, when activated by this steroid
hormone. This increased synthesis would also have to be easily detect-
able in histological preparations (by autoradiograph~c methods after
incorporation of radioisotopically labelled RNA-precursors) as it is
known that any steroid-hormone-induced synthesis of messenger-RNA is
normally linked to a simultaneously significant synthesis of ribosomal-
RNA (Tata, 1970; Williams-Ashman and Reddi, 1972).
Another series of experiments was therefore performed, in order to

detect the target-cells responsible for the testosterone-induced sex
reversals, by associating the testosterone administration with the
uptake of a radioisotopically labelled RNA-precursor, 3H-uridine.
These have been studied by autoradiographic methods.
The experiments were performed on very big tadpoles of a sexually dif-

ferentiated race of Rana escuZ-enta (Zaccanti, 1974a). The very large
size of these specimens was due to the fact that they were "hibernat-
ing" tadpoles, which had passed the winter in a persistent larval
state before approaching the end of metamorphosis at the beginning of
the second year of their life.
The treated animals were injected with a testosterone oil-solution,

and the controls with the same oil dose, without testosterone. Both
the treated and control specimens were killed for histological and
326
autoradiographic examinations, 5, 10, 15, or 20 days after the begin-

ning of the experiments (and a day after having been injected with
3H-uridine water-solution) .
At the beginning of the experimental period the female genotypes had

large, well-developed ovaries filled with diplotenic developing oocytes.
In the treated group, the first morphological evidence of a beginning
of sex reversal was noticed only after twenty days of testosterone
administration. In the control animals (Fig. 24), and as is also the
case under normal conditions, the ovarian central cavity was completely
devoid of somatic cells and at its proximal end, the mesovarium con-
sisted of only two closely adjacent coelothelial laminae - completely
devoid of any cellular content between them. However, in the ovaries
of treated specimens (Fig. 25), the mesovarium was greatly enlarged.
It was filled with a newly proliferated medullary tissue, extending
from an extra-gonadal territory (the inter-renal blastemic region,
adjacent to and partly intermingled with the medial face of the meso-
nephros), and expanding to completely fill the proximal part of the
ovarian cavity. Here, the oogonia are incorporated for their trans-
formation into spermatogonia.
The autoradiographic study (Fig. 26, 27) clearly demonstrated a strong

radioisotopic labelling of the RNA-containing ribosomes of the newly
proliferated somatic cells. The testosterone target-cells responsible
for the hormone-induced ovarian sex reversal (and whose activation
appears to be significantly linked with an RNA-neosynthesis) seem
therefore to be predominantly localized in the extra-gonadal inter-
renal territory, from which they descend towards the gonads. We pro-
pose to give the name " medulloblasts" to these extra-gonadal cells,
from which the masculinizing gonadal medulla originates.
It was noted that this new proliferation of medullary tissue, which

started the testosterone-induced sex reversal, was not preceded by
any significant oocyte degeneration. In the large ovaries of the Rana
escuZenta tadpoles used in our experiments, the anti-oogenetic activity
of testosterone resulted in a significant decrease in the RNA-labelling
in the loops of the lampbrush-chromosomes of the nuclei of diplotenic
oocytes. This is shown by the comparison of a normal (Fig. 28) with
a treated oocyte (Fig. 29).
Plate IV. Rana escuZenta tadpoles. Autoradiographic study; cross-sec- ~

tions of gonads
Fig. 24. Control animals; mesovarium consisting only of two closely
adjacent coelothelial laminae. x 570
Fig. 25. Testosterone treated animals (100 ~g/specimen); greatly en-
larged mesovarium, filled with a new proliferated medullary tissue.
x 270
Fig. 26. Testosterone treated animals (same dose as above). As in
Fig. 25. x 360
Fig. 27. Testosterone treated animals (same dose as above); detail of
Fig. 26. Autoradiographically labelled medulloblast in the mesovarium.
x 1110
Fig. 28. Control animals. 3H-uridine incorporation in a diplotenic
oocyte. x 760
Fig. 29. Testosterone treated animals (same dose als above). Signifi-
cant decrease of 3H-uridine incorporation in a diplotenic oocyte. x 760
327
Plate IV. Fig. 24 - 29

328
A quantitative evaluation of the RNA-labelling was made by calculating

the number of silver granules in the autoradiographic preparations.
It was clearly demonstrated that the decrease of RNA-synthesis (induced
by the testosterone in the nuclei of diplotenic oocytes) was very pre-
cocious (Fig. 30). It started after 5 days of testosterone administra-
tion, persisted at the same level after 10 - 15 days and then, abruptly
reached its greatest value after 20 days (concomitant with the begin-
ning of the medullary tissue proliferation in the ovaries) .
50 ----~r------------------------
1.0
'"o~
o
~30
eLI
::J
c:
ra Fig. 30. Scheme of
:;, 20 quantitative evalu-
ation of ru~A-label
01 ling in diplotenic
oq
oocytes of the con-
trol (e) and testos-
10 terone treated (T)
tadpoles of Rema
esculenta. (Testoster-
one, 100 ~g/specimen;
3H-uridine, 10 ~Ci/
c T 5-15 dolys T 20d.ays specimen)
This expression of the anti-oogenetic activity of testosterone does

not necessarily represent a prerequisite for the displayment of the
following medullo-stimulating effect of the androgen hormone. It is
known that a sex reversal by testosterone-induced proliferation of
medullary tissue, may also be obtained in Rana tadpole ovaries com-
pletely deprived of germ cells and consequently also, lacking in oocytes
(Padoa, 1964). It is thought that the sudden diminution of the RNA-
neosynthesis (noted in oocyte nuclei at the twentieth day of testos-
terone treatment) may be but the consequence, and not the cause of the
new proliferation of the medullary tissue. Witschi (1967) has also
stated that the masculinizing medullary tissue of the amphibian gonad
could be provided with anti-oogenetic inducing powers.
D. Suggestions of a Two-step Activity of Testosterone in Inducing

Somatic Cells to Carry Out Sex Reversal
Other experiments, similar to the ones just described, have been carried
out on tadpoles of a sexually undifferentiated race of Rana latastei, a
closely related species to R. dalmatina (Zaccanti, 1974b).
329
1wo identically treated groups of pre-metamorphic specimens were in-

jected with the same testosterone oil-solution, while a control group
had the same oil dose injected without testosterone. After six weeks,
one of the two testosterone-treated groups was additionally injected
with a sub-lethal dose of actinomycin D water-solution. Finally, eight
weeks after the beginning of the experiments and a day after having
been injected with 3H-uridine water-solution, all the control and
treated animals were killed for histological and autoradiographic
studies.
The morphological results were very similar to those previously de-

scribed for the similarly treated Rana dalmatina tadpoles. In control
specimens, the gonads always preserved their typical ovarian structure,
both in the male and female genotypes, while in the specimens treated
with testosterone alone, a sex reversal was constantly induced. In
the specimens treated with testosterone plus actinomycin D, only the
anti-oogenetic effect of testosterone was not hindered, while the
medullo-stimulating effect was suppressed.
As the actinomycin D was administered only during the last two weeks
of testosterone treatment, its effectiveness in interfering with the
masculinizing power of the androgen hormone (in spite of this long
delay in its supply) seems to be significant. It seems to suggest a
certain degree of graduality in the expression of the medullo-stimu-
lating activity of the testosterone. This was confirmed by the quan-
titative autoradiographic study of our material, the results of which
are summarized in a scheme (Fig. 31).
0~~:if~~I
T Fig. 31. Volumetric reconstruction of the
male genotype gonads (hatched areas) and
quantitative representation of 3H-uridine
~~:w ~
incorporation (white areas) in inter-renal
tissue (i); in gonadal somatic tissue (g);
of control (e); testosterone (T) and tes-
tosterone plus actinomycin D (T + A)
treated tadpoles of Rana latastei (testos-
T+A terone, 10 ~g/specimen; actinomycin D,
0,85 ~g/specimen; 3H-uridine, 10 ~Ci/
9
specimen). Further details in the text
In this scheme, the central hatched drawings represent a volumetric

reconstruction of the gonads on both sides of the mesentery, while
the laterally placed white areas respectively indicate, level-by-Ievel
330
of their longitudinal extension, the degrees of RNA-labelling in both

the gonadal somatic cells (g), and the extra-gonadal interrenal terri-
tory (i) from which the new proliferation of medullary tissue is de-
rived, in the cases of sex reversal.
In the control tadpoles, whose gonads are well-preserved ovaries with

a cavity, a very low RNA-labelling (indicative of low RNA-neosynthesis)
is detectable - both in the somatic cells of the gonads and in the
inter-renal territories. In the tadpoles whose compact gonads were
sex-reversed ovaries with a central cavity full of new proliferated
medullary tissue, testosterone treatment resulted in both the gonadal
somatic cells and the inter-renal territories showing a high degree
of RNA-labelling. This increase in labelling (indicative of a strong
RNA-neosynthesis), was greater in the inter-renal territories than in
the gonadal medulla. Finally, in specimens whose gonads were ovaries
with degenerating oocytes (but still well-cavitated owing to the sup-
pression of new medullary tissue proliferation) treatment with testos-
terone plus actinomycin D showed a high degree of RNA-labelling to be
clearly detectable in the inter-renal territories, while only a low
degree of RNA-labelling existed in the gonadal somatic cells.
These facts strongly suggest a two-step activity of testosterone.

Medulloblasts would be activated in the extra-gonadal inter-renal re-
gion, and only later would these activated cells be induced to migrate
towards the gonads. If this is indeed so, our results might be ex-
plained by supposing that the activation of the medulloblasts had al-
ready occurred when actinomycin D was administered to the tadpoles,
after six weeks of treatment with testosterone. As a consequ~nce, this
delayed actinomycin D administration was only able to suppress the,
as yet unrealized, migration of these cells - from the inter-renal
territory toward the gonads, while the activation of the extra-gonadal
medulloblasts was still detectable two weeks after the actinomycin D
injection.
Direct activation by testosterone, of the medullary cells lining the

central cavities in the well-developed ovaries, has never been observed
in our material. If it exists, its importance would seem to be minimal.
E. Conclusion
According to the interpretation of Witschi's studies on amphibian

sexuality (Goldschmidt, 1931), sex differentiation would depend on
the activities of: (1) "primary sex differentiators" localized in the
chromosomes (namely, the female or male sex-genes); (2) "secondary sex
differentia tors " produced by gonado-cortical or medullary somatic tis-
sues (namely, Witschi's inductive substances, "corticin" or "medulla-
rin"); and (3) "tertiary sex differentiators" produced by endocrine
cells and evocating secondary sex characters in adult animals (namely,
estradiol in female and testosterone in male) .
Obviously, the testosterone-induced sex reversal activity analyzed in

Rana tadpoles, is merely to be thought of as a pharmacological action
of this steroid. It does not directly demonstrate a supposed identity
of testosterone, with any natural substance normally operating in spon-
taneous sex differentiation of the gonads (Chieffi, 1965). Therefore,
a genuine participation of testosterone (or some of its metabolites)
in the starting of physiological male-sex differentiation does not
seem probable.
331
Our hypothesis, on the ability of testosterone to de-repress male-genes

justifies, however, the well-known similarity which exists: not between
testosterone and "medullarin" activities, but rather between testoster-
one and male-gene activities. The medullo-stimulating effects which are
attained by administering testosterone are, in fact, similar to those
which would be attained if it were possible to add male-genes to the
genotype.
Our research is continuing, with the aim of testing the relationships

between spontaneous and experimentally-induced sex reversal in tadpole
gonads.
References
Chieffi, G.: Onset of steroidogenesis in the vertebrate embryonic

gonads. In: Organogenesis (eds. R.L. DeHaan and H. Ursprung),
pp. 653-671. New York-Chicago-San Francisco-Toronto-London: Holt,
Rinehart & Winston 1965.
Karlson, P. (ed.): Mechanisms of hormone action. Stuttgart: Thieme.
Padoa, E.: 11 differenziarnento sessuale delle gonadi di Rana esculenta
rese sterili dall'irradiarnento con ultravioletto delle uova indivise.
Boll. Zool. 31, 811-825 (1964).
Smellie, R.M.S.~ed.): The bio~hemistry of steroid hormone action.
Tata, J.R.: Regulation of protein synthesis by growth and developmental
hormones. In: Biochemical Actions of Hormones (ed. G. Litwack),
vol. I, pp. 89-133. New York-London: Academic Press 1970.
Vannini, E., Stagni, A.: Repression by actinomycin D of testosterone
induced sex reversal in Rana dalmatina tadpoles. EXp. Cell Res. 46,
460-463 (1967). -
Vannini, E., Stagni, A.: Repression by puromycin of testosterone-in-
duced sex reversal in Rana dalmatina tadpoles. EXp. Cell Res. 50,
683-687 (1968).
Vannini, E., Stagni, A.: Nuove ricerche sul differenziarnento sessuale
nelle larve degli Anfibi. R.C. Accad. Sc. 1st. Bologna~, 83-94
(1971) .
Vannini, E., Stagni, A.: Inibizione con actinomicina D dei processi
di inversione sessuale provocati dal testosterone sugli bvari dei
girini di Rana dalmatina. Arch. Ital. Anat. Embr. 77, 25-68 (1972).
William-Ashman, H.G., Reddi, A.H.: Adrogenic regulation of tissue
growth and function. In: Biochemical Actions of Hormones (ed. G.
Litwack), vol. II, pp. 257-294. New York-London: Academic Press
1972 .
Witschi, E.: Biochemistry of sex differentiation in Vertebrate embryos.
In: The Biochemistry of Animal Development (ed. R. Weber), vol. II,
Zaccanti, F.: Osservazioni istologiche ed autoradiografiche su ovari
di girini di Rana esculenta in iniziale inversione sessuale indotta
con testosterone. R.C. Accad. Naz. Lincei (in press, 1974a).
Zaccanti, F.: Osservazioni autoradiografiche sopra gonadi larvali di
Rana latastei in inversione sessuale indotta mediante testosterone
e bloccata mediante actinomicina. R.C. Accad. Naz. Lincei (in press,
1974b) .
Temperature and Sex Differentiation in Embryos of Two
Chelonians, Emys orbicularis L. and Testudo graeca L.
C. Pieau
A. Introduction
It is a well-known fact that temperature is a sex-reversal factor in

the larvae of amphibians (Witschi, 1929; Piquet, 1930; Uchida, 1937;
Yoshikura, 1959; 1-1uto, 1961; Hsli et al., 1971). Until recently, this
action of the temperature had not been demonstrated in the other ver-
tebrates, except in fishes. In a reptile, Agama agama (Lacertilia, Agami-
dae), Charnier (1966) observed abnormal sex-ratios in new born individ-
uals arising from eggs which were incubated in different conditions:
when eggs were incubated at 26 - 27°C in the wet sand of a rearing-cage,
most individuals displayed a female phenotype; when eggs were placed on
moist cotton, in an incubator at 29°C, the majority of the individuals
displayed a male phenotype. From this observation, Charnier considered
that the temperature could exert some action on the sex-ratio in embryos
of A. agama and that, under natural conditions, the variations of temper-
ature (mainly the nycthemeral rhythm) could permit the normal differ-
entiation of the sexes (in conformity with the genetical sexes) and the
sex ratios 1~: 1~ • Unfortunately, Charnier did not incubate eggs of
A. agama at different temperatures, while keeping the other conditions
constant.
I have also obtained abnormal sex-ratios in embryos of two chelonians,

Emys orbicularis L. and Testudo graeca L., developed under artificial incu-
bation. Therefore, it seemed that the temperature could have some in-
fluence on the sex differentiation (Pieau, 1971) and the role of this
factor was studied. Some results of this study have been described in
separate publications (Pieau, 1972a,b; 1973 a-c; 1974a,b).
In a first series of experiments, eggs of E. orbicularis L. and T. graeca

L. were incubated at constant temperatures, or presenting only slight
variations. In a second series of experiments, eggs of E. orbicularis L.
were submitted to artificial cyclic variations of temperature, or were
reared under natural conditions in the ground.
C. Results
I. Sex Differentiation in Embryos of Emys orbicularis L. and Testudo graeca

L. Developed at Constant Temperatures
The conditions for obtention and incubation of eggs have described

in earlier publications (Pieau, 1971; 1972b; 1973a). In these experi-
ments the females of the two species laid eggs at the laboratory (the
333
adul ts of Testudo graeca L. were reared in large, well-exposed exterior

terrariums during spring and sununer; females of Einys orbicularis L. were
collected in nature during the egg-laying season and reared at the la-
collected in nature during the egg-laying season and reared at the
laboratory in June and July). The eggs were incubated in glass-dishes
several beds. The lower beds were saturated with distilled water, so
that the relative humidity was a little more elevated at the level of
eggs of E. orbicularis L. (about 9S%), than at the level of eggs of
T. graeca L. (8S - 90%6. Under these conditions, eggs of E. orbicularis L.
were incubated at 2S ± 1°C, 27,So ± O,SoC and 29,So ± O,SoC and eggs
of T. graeca L. at 26, SO ± 0, SoC, 29, SO ± 0, SoC and 31, S ° ± 0, SoC.
The macroscopical and histological study of the genital apparatus gave

the following results:
InoE. orb~cularis L., all the ernJgryos d~splayed a male phenotype at °

2S C ± 1 C (76 cases) and 27,5 ± O,S C (2S cases), whereas at 29,S
± O,SoC, they displayed a female phenotype (117 cases). At 2So ± 1°C
and 27,So ± O,SoC, the gonads acquired testicular structures. In some
cases (probably in the genetical males), this structure was typical,
while in the others (genetical females?) the gonads displayed some
characters indicating a slight degree of intersexuality: persistence
of the germinal. epithelium (Fig. 1) which sometimes formed an anlage
of a cortex; testicular tubules connected with the germinal epithelium
or opening directly in the coelomic cavity (Fig. 2). In all these em-
bryos, the MUllerian ducts degenerated.
At 29,5° ± O,SoC, the gonads acquired an ovarian structure: the cortex

was well-developed, the medullary cords were reduced or formed lacunae
(Fig. 3); in some cases (genetical males?), the cortex was thinner, but
the medullary cords were not differentiated into testicular tubules
and the MUllerian ducts were of a female type.
The ~S-3B-hydroxysteroid dehydrogenasic activity in the gonads dis-

playing a testicular type at 2So ± 1°C and those displaying an ovarian
type at 29,So ± O,SoC, were compared. The procedure was according to
the method of Wattenberg (19S8), modified by Levy et al. (19S9), then
by Baillie et al. (1966). The substrate used was dehydroepiandrosterone
It has been shown that at 2So ± 1°C, activity was evident in the med-
ullary sex-cords of the undifferentiated gonads. This activity increased
during the time of the regression of the MUllerian ducts, in the tes-
ticular tubules derived from these sex-cords and at the end of the
embryonic life, it became very feeble. At 29,So ± o,SoC, a very slight
enzymatic activity was present in the medullary cords of the undiffer-
entiated gonad, which then disappeared once the gonad began to acquire
ovarian characters.
In T. graeca L., all the embryos which developed at 26,So ± o,Soc (19
cases) and 37 out of 38 embryos which developed at 29,So ± O,Soc showed
a male phenotype: the gonads were testes; in many cases a layer of
germinal epithelium persisted at their surface (Fig. 4); in some indi-
viduals, testicular tubules were connected with this germinal epithe-
lium but they did not open in the coelomic cavity, as in E. orbicularis
L.; the MUllerian ducts regressed. In one case (the only exception),
at 29,So ± o,Soc, the genital apparatus displayed a female phenotype:
the gonads showed an ovarian structure (Fig. S) with a well-developed
cortex and a mesenchymal medulla containing thin epithelial cords and
lacunae; the MUllerian ducts were wider than in males, and complete.
The macroscopic examination of all the embryos at 31,So ± o,Soc (16

cases) revealed a female morphology. The histological study of the
gonads showed that the medulla displayed the same structure as that
334
Fig. - 6
335
in the ovarian-type gonads obtained in one individual at 29,5° ± O,5 0 C:

it was formed by a loose mesenchymal tissue in which the medullary
cords were reduced or lacunose (compare Fig. 6 with Fig. 5). If the
germinal epithelium persisted at the surface, it did not form a typical
ovarian cortex (Fig. 6). However, in some cases (genetical females?)
it thickened and began to form a cortex. These gonads cannot be con-
sidered as ovaries, but the interesting fact to note is that at 31,5°
± O,5 0 C, the testicular tubules were not differentiated, whereas below
30°C they differentiated in all the embryos. The b5-3S-hydroxysteroid-
dehydrogenase activity of some gonads have been studied, at 29,5°
± O,5 0 C and 31,5° ± O,5 0 C. As in embryos of E. orbicul=isL. developed
at 25° ± 1°C, this activity was intense in the testicular tubules of
the testes obtained at 29,5° ± O,5 0 C (at the time of the regression
of Mlillerian ducts). Likewise, it was feeble, then decreased and dis-
appeared in the medullary cords of the gonads at 31° ± O,5 0 C, as was
also the case in the ovarian-type gonads of E. orbicul=is L. at 29,5°
± O,5 0 C.
These results indicated that at every temperature used in the two spe-
cies of chelonians which have been studied, some individuals did not
develop a sexual phenotype in conformity with the genetical phenotype.
Therefore, it appears that the temperature was a sex-reversing factor
in the embryos of these chelonians. However, in embryos of T. graeca L.,
at 31° ± O,5 0 C the sexual phenotype was modified but not reversed.
The differences between the temperatures used were not great, so it
was considered (Pieau, 1973c; 1974b) that there was a critical temper-
ature above and below which, one of the sexual phenotypes is modified
.... Fig. 1-6. Transverse sections through the gonads of embryos of chelo-
nians developed at diverse temperatures.
Fig. 1, Testis in an embryo of Emys orbicularis L. developed at 25° ± 1°C
(56 days of incubation; weight: 3,192 g; length of carapace: 21 rom);
wide testicular tubules (t.t.) are differentiated in the medulla; the
germinal epithelium (g.e.) persists at the surface. (x 124)
Fig. 2. Testis in an embryo ofEmysorbicularis L. developed at 25° ± 1°C
a testicular tubule (t.t.) opens in the coelomic cavity. (g.e.: ger-
minal epithelium). (x 160)
Fig. 3. Ovary in an individual of Emys orbicularis L. developed at 29,5°
± O,5 u C, at hatching (55 days of incubation; weight: 4,767 g; length
of carapace: 25 rom); note reduced medullary cords and lacunae in the
medulla (m.) and a well-developed cortex (c.) with oogonia and oocytes
(ov.). (x 124)
Fig. 4. Testis in an individual of Testudo graeca L. developed at 29,5 °
± O,5 u C, at hatching (68 days of incubation; weight: 10,245 g; length
of carapace: 29 rom); note wide testicular tubules (t.t.) in the medulla
and the persistence of a layer of germinal epithelium (g.e.) at the
surface. (x 124)
Fig. 5. Gonad presenting an ovarian structure in one embryo of Testudo
graeca L. developed at 29,5° ± 0,5 0 C (61 days of incubation; weight:
7,910 g; length of carapace: 27 rom); the medulla (m.) is formed by
loose mesenchymal tissue and contains lacunae (l.); a cortex (c.) be-
gins to form at the surface. (x 124)
Fig. 6. Gonad in an embryo of Testudo graeca L. developed at 31,5 ° ± 0,5 °c
the medulla (m.) presents the same structure as that in the gonad of
ovarian type (compare with Fig. 5); at the surface, the germinal epi-
thelium (g.e.) persists, but it does not develop into a cortex. (x 124)
336
or reversed. Probably, this critical temperature is between 28°C and

29°C in E. orbicularis L. and between 30°C and 31°C in T. graeca L. Ev-
idently, such temperatures do not exist constantly in the natural con-
ditions where eggs are submitted to the variations of temperature of
the ground. Therefore, to extend the hypothesis of Charnier (1966),
the effects of cyclic variations of temperature on the sex differen-
tiation in embryos of E. orbicularis L. have been studied.
II. Effects of Cyclic Variations of Temperature on the Sex Differen-

tiation in Embryos of Emys orbicularis L.
The effects of cyclic variations of temperature on the sex differen-

tiation were studied in embryos of E. orbicularis L., which developed
either in artificial incubati.on or in the ground.
In the first experiment (Fig. 7), eggs were submitted to the influence
of variations of temperature between 24°C and 30°C, according to a
daily cycle. This included a time of incubation which was longer below
28°C than above (28°C is near the presumed critical temperature). The
26 sexually differentiated embryos displayed a male phenotype (the
gonads being typical testes, and the MUllerian ducts regressed) or
were intersexes (the gonads were ovotestes with wide testicular tubules
in the medulla and an anlage of a cortex at their surface, while the
MUllerian ducts degenerated). However, female differentiation was not
obtained.
cI' and intersexes
30'
20'
Fig. 7
d" I ~ and intersexes
20'
Fig. 8
Fig. 7. Cyclic variations of temperature between 24°C and 30°C to which

eggs of Emys orbicular,is L. were submitted. The duration of incubation
was longer below 28 C (16 hrs per day) than above
Fig. 8. Cyclic variations of temperature between 26°C and 31· o C to which
eggs of Emys orbicularis L. were submitted. The daily duration of incu-
bation was the same below and above 28°C
337
In the second experiment (Fig. 8), 36 eggs were submitted to variations

of temperature between 26°C and 31°C, according to a daily cycle in-
cluding approximately, the same durations (12 hrs) of incubation above
28°C and below. The two sexes differentiated and some embryos were
intersexes (the gonads were ovotestes, and the Mlillerian ducts re-
gressed) •
These results indicated that the two sexes could differentiate under
the influence of cyclic variations of temperature, but only if certain
conditions were realized. One of these conditions seemed to be, that·
at the time of the sexual differentiation of the gonads, the daily
durations of incubation were almost equally below and above the crit-
ical temperature (Pieau, 1973c). If this condition is not realized
in the ground, it is probable that abnormal sex-ratios can be obtained
under natural conditions, just as in artificial incubation.
Eggs of E. orbiauZaris L. were buried in the ground during summer (Lieau,

1974a). Of 40 embryos which differentiated sexually, there was no fe-
male. One embryo was an intersex (the gonads being an ovotestis), while
all the others displayed a male phenotype. In the latter, the gonads
were typical testes (probably in genetical males), or (in genetical
females?) they displayed some abnormalities with a slight degree of
intersexuality (protuberances, testicular tubules connected with the
germinal epithelium or opening in the coelemic cavity) .
These results were similar to those obtained in artificial incubation

(at 25° ± 1°C and 27,5° ± 0,5 0 C, or with cyclic variations between
24°C and 30°C); they indicated that in some individuals (genetical
females), the gonads had not differentiated in conformity with the
genetical sex. It was demonstrated that at the presumed time of the
action of temperature on the differentiation of the gonads under nat-
ural conditions, the temperature remained longer below the critical
level than above, so that females could not differentiate.
D. Conclusions
It clearly appears that in Emys orbiauZaris L. and Testudo graeaa L.,

(and probably also in other chelonians), the sexual differentiation
of the genital apparatus (gonads and genital tractus) can be modified
or even reversed during the embryonic life, by the action of the tem-
perature. The effects obtained by this action are the contrary to
those observed in amphibians: at low temperatures, all the individuals
displayed a male phenotype, whereas at high temperatures they displayed
a female phenotype. At present, these differences cannot be explained.
A peculiarity shown in this study is that the sexual differentiation

is modified by small differences of temperature. Therefore, it has
been suggested (Pieau, 1973c) that there is a critical temperature,
below and above which one of the sexual phenotypes is modified or re-
versed. Probably, this critical temperature is between 28°C and 29°C
in E. orbiauZaris L., and between 30°C and 31°C in T. graeca L. It must
now be determined if this temperature permits the "normal" sex differ-
entiation (in conformity with the genetical sexes) in the embryos of
these two chelonians species. The results obtained by incubating eggs
of Emys orbicuZaris L. in the ground during summer, show that under natural
conditions the sex-ratio can be also modified by the influence of the
factor "temperature". Likewise it must be determined how the normal
sex differentiation can be realized in these conditions. The artificial
incubation experiments indicate that the two sexes can differentiate
338
under the influence of cyclic variations of temperature, but only if

the daily duration of incubations are almost equally below and above
the critical temperature, at the time of the sexual differentiation
of the gonads.
These conclusions are established only from observations made during

the embryonic life. It is known that in amphibians (Gallien, 1962)
and birds (Wolff, 1936), partially sex-reversed during their develop-
ment by the action of sexual hormones, the gonads tend to revert to
the genetical sex. Is such a reversal possible after exlosion for the
gonads reversed by the action of temperature during the embryonic life
in chelonians? At present, young E. orbiaularis L. hatching from eggs
incubated at 25° ± 1°C (individuals displaying a male phenotype) and
29,5° ± 0,5 0 C (individuals displaying a female phenotype) are being
reared, to try to answer this question.
References
Baillie, A.H., Ferguson, M.M., Hardt, D.M.: Developments in steroid

histochemistry. London-New York: Academic Press 1966.
Charnier, M.: Action de la temperature sur la sex-ratio chez l'embryon
d'Agama agama (Agamidae Lacertilien). C.R. Soc. Biol. 160, 620-622
(1966). -
Gallien, L.: Evolution chez le Triton Pleurodeles waltlii des intersexues
obtenus apres traitement par l'hormone femelle. Bull. Biol. France
et Belgique 96, 249-280 (1962).
Hsli, C.Y., Yli, N.W., Liang, H.M.: Induction of sex-reversal in female
tadpoles of Bana aastesbeiana by temperature treatment. Endocrinol.
Jap. 18, 243-251 (1971).
Levy, H.~Deane, H.W., Rubin, B.L.: Visualization of steroid-3~-01-
dehydrogenase activity in tissue of intact and hypophysectomised
Rats. Endocrinology 65, 932-943 (1959).
Muto, Y.: The gonad of the toad, Bufo vulgaris formosus, cultured at high
temperature. Bull. Aichi Gakugel Univ. 10, 97-110 (1961).
Pieau, c.: Sur la proportion sexuelle chez-res embryons de deux Che-
loniens (Testudo graeaa L. et Emys orbiaularis L.) issus d' oeufs in-
cubes artificiellement. C.R. Acad. Sci. Paris 272, D, 3071-3074
(1971). -
Pieau, c.: Effets de la temperature sur le developpement des glandes
geni tales chez les embryons de deux Cheloniens, Emys orbiaularis L.
et Testudo graeaa L. C.R. Acad. Sci. Paris 274, D, 719-722 (1972a).
Pieau, c.: Incubation artificielle des oeufs de Cheloniens. - Effets
de la temperature sur le sexe gonadique des embryons. Symbioses 4,
253-257 (1972b). -
Pieau, c.: Development of the gonads at diverse temperatures, in em-
bryos of two Chelonians, Emys orbiaularis L. and Testudo graeaa L.
Ontogenez (in Russian) 4,11-16 (1973a).
Pieau, c.: Variation de l'activite enzymatique ~5-3~-hydroxysteroide
deshydrogenasique dans les glandes genitales d'embryons d' Emys
orbiaularis L. (Chelonien), en fonction de la temperature d' incuba-
tion. C.R. Acad. Sci. Paris 276, D, 197-200 (1973b).
Pieau, c.: Nouvelles donnees experimentales concernant les effets de
la temperature sur la differenciation sexuelle chez les embryons
de Cheloniens. C.R. Acad. Sci. Paris. 277, D, 2789-2792 (1973c).
Pieau, c.: Sur la differenciation sexuelle chez des embryons d'Emys
orbiaularis L. (Chelonien) issus d'oeufs incubes dans le sol au cours
de l'ete 1973. Bull. Soc. Zool. France (1974a, in press).
Pieau, c.: Differenciation du sexe en fonction de la temperature chez
les embryons d' Emys orbiaularis L. (Chelonien); effets des hormones
sexuelles. Ann. Embryol. I-1orph. (1974b, in press) •
339
Piquet, J.: Determination du sexe chez les Batraciens en fonction de

la temperature. Rev. Suisse Zool. 12, 173-281 (1930).
Uchida, T.: Studies on the sexuality of Amphibia. II. Sex transforma-
tion in Hynobius l'etal'datus by the function of high temperature.
J. Fac. Sci. Hokkaido Imp. Univ. Ser. VI, Zool. 6, 59-71 (1937).
Wattenberg, L.W.: Microscopic histochemical demonstration of steroid-
3S-ol-dehydrogenase in tissue sections. J. Histochem. Cytochem. 6,
225-232 (1958). -
Witschi, E.: Studies on sex differentiation and sex determination in
Amphibians. II. Sex reversal in female tadpoles of Rana sylvatica
following the application of high temperature. J. expo Zool. ~,
267-292 (1929).
Wolff, Et.: L'evolution, apres l'eclosion, des poulets males transformes
en intersexues par l'hormone femelle injectee aux jeunes embryons.
Arch. Anat. Hist. Embryol. 83, 1-38 (1936).
Yoshikura, M.: The action of the pituitary in sex differentiation and
sex reversal in Amphibians. II. Effects of high temperature on the
gonads of hypophysectomized frog larvae. Kumamoto J. Sci., Ser. B,
Sect. 2, i, 69-101 (1959).
Evolution of Parthenogenetic Species of Reptiles
Ch. J. Cole
The recent reptiles include representatives of four orders; Testudinata

(turtles and tortoises; ca. 335 species survive); Crocodilia (croco-
diles and their allies; ca. two dozen species survive); Rhynchocephalia
(only Sphenodon punctatuB of New Zealand survives); and Squamata (liz-
ards, amphisbaenids and snakes; ca. 5,700 species survive).
A. Reproduction and Sexuality
Most information on reptilian reproduction concerns behavioral, ana-

tomical and ecological aspects in lizards and snakes (Fitch, 1970);
reproductive physiology and environmental stimuli affecting it are
known for only a few species (Callard et al., 1972; Licht, 1972).
Reptiles normally are gonochoristic. Hermaphroditic reproduction and
sex reversal do not occur in reptiles. Nevertheless, intersexuality
does occur normally in adults of some species (see below) .
B. Sex Determination
Although a few experiments indicate that temperatures during embryonic

development may influence sex determination in some reptiles (Charnier,
1966; Pieau, 1974), essentially nothinq is known about mechanisms of
sex determination. Heteromorphic sex chromosomes have evolved inde-
pendently many times; presently they are in various degrees of dif-
ferentiation in the Squamata and a few Testudinata, but many species
lack them (Vorontsov, 1973; Wright, 1973; Moon, 1974).
C. Intersexuality
Intersexuality in reptiles is an anomaly in almost all instances,

other than the varying degrees that occur during normal growth and
development (Stefan, 1963; Forbes, 1964; Vivien, 1964). Intersexuality
in adults usually is the abnormal occurrence in one sex of reproductive
structures that ordinarily are limited to the other sex. Abnormal go-
nads with properties of both sexes (e.g., ovotestes) are very rare.
I know of only two true hermaphrodites, but these are not known to
have reproduced hermaphroditically; one was a snake (Hoge et al.,
1959), the other a lizard (Bons and Bons, 1969).
In two species of snakes, however, adult females normally possess cop-

ulatory organs. These organs (hemipenes; there is one on each side)
occur only in males in other species and serve to transmit spermatozoa
during copulation.
341
In Bothrops insuZaris, the sex of individuals is normally one of the fol-

lowing: True males (with male structures only); true females (with
female structures only); or intersexes (females with a hemipenis on
one or both sides; Hoge et al., 1959, 1961). Intersexes are more abun-
dant than true females, and all pregnant ones have been intersexes.
The karyotype of an intersex had a heteromorphic pair of chromosomes
similar to the ZW pair in females of other species of Bothrops (Becak,
1965). MacLean's (1968) report of a pregnant intersex of Bothrops moojeni
suggests that this may occur with normal frequency in other species
of Bothrops.
Intersexuality is more consistent in Pseudofieimia frontaUs. All adults

are either true males or females having hemipenes with their full suite
of accompanying musculature on each side (Hardy, 1970).
D. Parthenogenesis
Forbes (1964) concluded his review of intersexuality in reptiles by

citing instances of skewed sex ratios. Now it is clear that some spe-
cies of Squamata (about 27) have natural populations consisting en-
tirely, or essentially entirely, of females; their normal reproduction
is strictly parthenogenetic (not gynogenetic), and with few exceptions,
they produce female offspring.
Maslin (1971a) reviewed the facets of parthenogenesis in reptiles and

cited the species for which examples were known. My approach is to
review the information for each species thought to reproduce partheno-
genetically. Thus, the accounts for each species reveal what critical
data are lacking, besides what is known. Although the first substantial
report concerned lizards of the genus Laeerta (Darevsky, 1958), I review
Cnemidophorus first, as it is the group with which I am most familiar.
Family Teiidae: Genus Cnemidophorus
The genus Cnemidophorus is widespread in the western Hemisphere. Thir-

teen of the approximately 40 species reproduce parthenogenetically.
Cnemidophorus neomexieanus: I estimate that more than 600 specimens of

neomexieanus have been examined; all are females (Duellman and Zweifel,
1962; Taylor and Medica, 1966; Wright and Lowe, 1967a; Christiansen,
1971). This includes 34 (100%) that Maslin (1966) hatched in the lab-
oratory. The species survives at some localities where bisexual spe-
cies 1 of the genus are absent (Maslin, 1966).
Tl1e reproductive system and annual reproductive cycle of C. neomexieanu8

are similar to those of females of closely related, bisexual species.
In the mating season of the bisexual species, however, when sperm
apparently from recent matings are found in the females' oviducts,
the oviducts of neomexieanus contain no sperm (Medica, 1967; Cuellar,
1968, 1970; Christiansen, 1971, 1973). Neaves (1971) observed bivalent
lampbrush chromosomes in oocytes that appeared to be at meiotic diplo-
11 refer to unisexual species as those in which all normal individuals

are of one and the same sex (female). I refer to bisexual species as
those which are gonochoristic and for which males and females normally
exist (cf. Tomlinson, 1968).
342
tene. A female that was hatched and raised in the laboratory, without
contact with males of any Cnerrridophorus, produced four eggs. Al though
none hatched, at least two contained well-developed embryos when they
spoiled (Maslin, 1971b).
Cnerrridophorus neomexicanus is a diploid species (Pennock, 1965). Its karyo-

type appears to consist of one haploid complement indistinguishable
from that of the bisexual Cnerrridophorus tigris (including its X chromo-
some) and one complement indistinguishable from those in the sexlineatus
group (Lowe and Wright, 1966a,b; Cole et al., 1969; Lowe et al., 1970b).
This information, along with other characteristics of the species,
indicates that the origin of C. neomexicanus involved hybridization be-
tween C. tigris and C. inornatus (see Lowe and Wright, 1966a,b). Electro-
phoretic mobilities of a variety of proteins are consistent with this
hypothesis, and a combination of biochemical and chromosomal data is
most convincing (Neaves and Gerald, 1968; Neaves, 1969). Cnemidophorus
neomexieanus is highly heterozygous in comparison with bisexual species,
and it has genes most similar to those in C. tigris and C. inornatus. The
fact that neomexicanus and inornatus produce viable hybrids also indicates
genetic similarities; the hybrids have a 1:1 sex ratio (Taylor and
Medica, 1966; Wright and Lowe, 1967a; Christiansen and Ladman, 1968).
Cnemidophorus. tesselatus: About 687 females of tesselatus have been examined

(Tinkle, 1959; Maslin, 1962, 1971a; Zweifel, 1965). This includes 17
(100%) that Maslin (1966) hatched in the laboratory. A grand total of
8 males are known (Taylor et al., 1967; Saxon, 1968). These may be hy-
brids between tesse.latus and one of the bisexual species with which it
coexists at the localities where the males were found (Lowe et al.,
1970a). Interspecific mating involving C. tesselatus occurs in captivity
(Maslin, 1966; Neaves, 1971).
The reproductive system of C. tesselatus is similar to that of females

of bisexual species; in the mating season of the bisexual species,
however, when sperm apparently from recent matings are in the females'
oviducts, oviducts of tesselatus contain no sperm (Cuellar, 1968, 1970).
Neaves (1971) observed bivalent lampbrush chromosomes in oocytes that
appeared to be at meiotic diplotene. A female obtained from an egg
hatching in the field and raised in the laboratory without contact
with males produced two eggs. Although neither hatched, at least one
contained an embryo when preserv~d (Maslin, 1971b).
Zweifel (1965) investigated variation in C. tesselatus and recognized

six distinct color patterns (classes A.through F). Some of these are
diploids (classes C, 0, E and F) and some are triploids (classes A and
B). Karyotypes of the diploids appear to consist of one haploid comple-
ment indistinguishable from that of the bisexual C. tigris (including
its X chromosome) and one complement indistinguishable from those of
the sexlineatus group. Triploids appear to have the same two complements
as the diploids, plus an additional one of the sexlineatus group (Wright
and Lowe, 1967b; Cole et al., 1969; Lowe et al., 1970b). These obser-
vations, along with others, indicate that the origin of the diploids
involved hybridization between the bisexual C. tigris and C. septemvitta-
tus, and the origin of the triploids involved hybridization between
diploid parthenogenetic tesselatus and the bisexual C. sexlineatus (see
\vright and Lowe, 1967b). Electrophoretic mobilities of a variety of
proteins are consistent with this hypothesis, and a combination of
biochemical and karyotypic data is most convincing (Neaves and Gerald,
1968, 1969; Neaves, 1969).
Skin transplants among individuals within a bisexual species (C. sex-

lineatus) are consistently rejected, whereas most transplants among
individuals of tesselatus are accepted as well as control isografts
343
(Maslin, 1967). Grafts among triploids are accepted as are most of

those among diploids. Nearly all grafts rejected among tesseZatus are
transplants from triploids onto diploids (not invariably rejected,
however), whereas skin of diploids is accepted by triploids. These
experiments indicate remarkable similarity in histocompatibility genes
among different populations and different pattern classes of C. tesse-
Zatus.
Cnemidophorus dixoni: Scudday (1973) coined this name for some tesseZatus-
like lizards, including pattern class F of C. tesseZatus. At least 49
specimens, all females, are known (Zweifel, 1965; Scudday, 1973).
Cnemidophorus uniparens: About 750 females have been examined (Duellman and
Zweifel, 1962; Wright and Lowe, 1965; Wright, 1968; Cuellar, 1971).
The only uniparens-like males are three individuals that Wright (1968)
suggested are hybrids between uniparens (~) x inornatus (0"). The males are
from localities where these species coexist. He has found no males
among over 600 uniparens from areas where inornatus does not also occur,
and this interpretation is consistent with other reports of hybrid
males (Lowe et al., 1970a).
The reproductive system of uniparens is similar to that of females of

bisexual species. In the mating season of the bisexual species, how-
ever, when sperm apparently from recent matings are in the females'
oviducts, oviducts of C. uniparens contain no sperm (Cuellar, 1968,
1970, 1971). Seven females that were hatched and raised to maturity
in the laboratory without contact with males produced 23 eggs; some
spoiled, but three hatched and 12 others contained embryos when they
failed (Maslin, 1971b).
Cnemidophorus uniparens is the only parthenogenetic reptile for which oogen-

esis has been investigated in detail (Cuellar, 1971). In this triploid
species, meiosis apparently is preceded by a karyokinesis (division of
the chromosomes) that is not followed by cytokinesis (division of the
cell) so that primary oocytes are hexaploid. An essentially normal
meiosis, perhaps involving orderly synapsis of the newly-formed sister
chromosomes, results in production of triploid ova. Meiosis is soon
followed by cleavage, and the embryo is readily recognized at oviposi-
tion, about a week later (Cuellar, 1971; Maslin, 1971b).
The triploid karyotype of C. uniparens contains three haploid comple-

ments typical of those in the sexZineatus group (Lowe and Wright, 1966a;
Lowe et al., 1970b). Lowe and Wright (1966a,b) concluded that uniparens
evolved through hybridization between the bisexual C. guZaris and C. in-
ornatus, and that uniparens contains one haploid complement from guZaris
and two from inornatus. Electrophoretic mobilities of a variety of pro-
teins indicate that there are indeed two doses of genes from inornatus
and one dose from another species in the sexZineatus group, but the
other species has not yet been so clearly determined (Neaves, 1969).
Some information on uniparens has been reported as applying to C. veZox,

since Maslin (1966) initially disagreed with Wright and Lowe's (1965)
recognition of uniparens. Since then, Maslin (1968: 224) determined that
" .•• reciprocal skin transplants between veZox and uniparens are vigorously
rejected by both forms, thus corroborating Wright and Lowe's treatment".
Cnemidophorus veZox: More than 400 females have been examined (Gehlbach,
1965; Wright, 1966; Taylor et al., 1967); this includes 39 (100%)
hatched in the laboratory (Maslin, 1966). The only male known (Taylor
et al., 1967) is from a locality where veZox coexists with the bisexual
C. tigris, and it may be a hybrid (Lowe et al., 1970a): C. tigris does
mate with veZox in captivity (Maslin, 1966). Cnemidophorus veZox also
344
abounds at some localities where bisexual species of the genus are

absent.
The reproductive system of C. vel-ox is similar to that of females of

bisexual species. In the mating season of the bisexual species, how-
ever, when sperm apparently from recent matings are in the females'
oviducts, oviducts of vel-ox contain no sperm (Cuellar, 1968, 1970).
Cnemidophorus vel-ox is a triploid species; its karyotype contains three

haploid complements similar to those of the sexl-ineatus group (Lowe and
Wright, 1966a,b; Lowe et al., 1970b). A combination of biochemical and
karyotypic data indicates that vel-ox inherited two doses of genes from
C. inoPnatus and one from another species in the sexUneatus group, but
the other species has not yet been so clearly determined (Neaves, 1969).
Cnemidophorus opatae: This rare triploid species is known from only 20

females; its karyotype contains three haploid complements similar to
those of the sexUneatus group· (Wright, 1967; Lowe et al., 1970b).
The Cnemidophorus exsanguis complex: Three sibling species are recognized

(Lowe and Wright, 1964): ~ exsanguis is known from about 400 females,
C. fl-ageHiaaudus from at least 33 females, and C. sonorae from at least
50 females (Duellman and Zweifel, 1962; Lowe and Wright, 1964; Taylor
et al., 1967; Lowe et al., 1970a). These estimates include 21 females
(100%) of exsanguis hatched in the laboratory (Maslin, 1966). All three
species occur at some localities where bisexual species of the genus
are absent. The only known male (Taylor et al., 1967) is an exsanguis
from a locality where it coexists with the bisexual C. inornatus and
C. tigris; it may be a hybrid. Tetraploid hybrids result from syngamy
involving an ovum from a triploid unisexual species and a spermatozoan
from a diploid bisexual species (Lowe et al., 1970a; Neaves, 1971).
The reproductive system of C. exsanguis is similar to that of females

of bisexual species. In the mating season of the bisexual species, how-
ever, when sperm apparently from recent matings are found in the females'
oviducts, oviducts of exsanguis contain no sperm (Cuellar, 1968, 1970).
Neaves (1971) observed bivalent lampbrush chromosomes in oocytes that
appeared to be at meiotic diplotene.
Cnemidophorus e:csanguis. C. fl-ageHiaaudus. and C. sonorae are triploid species

whose karyotypes contain three haploid complements similar to those of
the sexl-ineatus group (Pennock, 1965; Lowe and Wright, 1966a; Lowe et
al., 1970b). Electrophoretic mobilities of a variety of proteins and
karyotypic data indicate that: exsanguis inherited one set of genes from
C. gul-aris, one set from C. inornatus, and one set from possibly a third
species (undetermined) in the sexl-ineatus group (Neaves and Gerald, 1968;
Neaves, 1969).
Cnemidophorus l-aredoensis: This species is known from only 72 females

(McKinney et al., 1973). Eggs, ovaries and oviducts are similar to
those in females of bisexual species. Morphology, geographic distri-
bution, and electrophoretic mobilities of 15 proteins indicate that
this is a diploid species whose origin involved hybridization between
the bisexual C. gul-aris and C. sexUneatus (McKinney et al., 1973).
The Cnemidophorus aozumel-a complex: Two species in Mexico and Guatemala

comprise this complex: C. aozumel-a and C. radeaki. These may represent
clones that have a common ancestor. Fritts (1969: 528) stated that
"of over 533 specimens belonging to the aozumel-a complex examined by
McCoy and Maslin (1962) or myself only two males have been discovered".
He suggested that the males were hybrids between C. aozumel-a un x c.
345
angusticeps (~). Reproductive systems of the females resemble those of

bisexual species (MCCoy and Maslin, 1962).
Cnerrridophorus cozwnela is a diploid species (Fritts, 1969) with "... several

distinct though similar karyotypic forms •.. " (Lowe et al., 1970b: 132).
Although karyotypes of most specimens do not clearly indicate a hybrid
origin, Fritts (1969) found one specimen (among 12 he karyotyped) that
appeared to have one haploid complement from the bisexual C. a:ngusticeps
and one from the bisexual C. deppei. This, together with external mor-
phology, ecology, and geographic distribution, suggests that C. cozwnela
evolved through hybridization between these species (Fritts, 1969).
CnerrridophoT'us lemniscatus: This Central American and South American species

occurs in bisexual populations at many localities and unisexual popu-
lations at others (Vanzolini, 1970; Peccinini, 1971; Hoogmoed, 1973).
A sample of 5 collected at Cbidos in 1965 contained 2 males and 3 fe-
males; a sample of 108 from the same area in 1968 consisted entirely
of females (Vanzolini, 1970). This apparently rapid shift in the local
sex ratio, the absence of sympatric congeneric species and the spotty
distribution of populations in many places suggests that parthenogenesis
evolved several times independently in Z-emniscatus and did not involve
interspecific hybridization (Vanzolini, 1970; Peccinini, 1971; Hoogmoed,
1973). This 'is a diploid species (Lowe et al., 1970b; Peccinini, 1971).
Family Teiidae: Three Additional Genera
Representatives of three other teiid genera appear to be parthenogenet-

ic. Nearly 100 specimens of Gymnophthalmus undeT'IiJoodi from South America
and Caribbean islands are all females (Thomas, 1965; Hoogmoed, 1973).
No males are known for the South American Leposoma peT'caT'inatwn; approx-
imately 40 specimens have proven to be females (Uzzell and Barry, 1971;
Hoogmoed, 1973). No males are known for the South American KentT'Dpyx
bOT'ckianus, of which 27 females were examined by Hoogmoed (1973).
Family Lacertidae: Genus LaceT'ta
Lacerta, with about 40 species, is widespread in the Eastern Hemisphere.

Four sibling species occurring in the Caucasus between the Black Sea
and the Caspian Sea are unisexual: LaceT'ta aromeniaca, L. dahli, L. T'ostombe-
kovi and L. unisexualis. Probably more than 5,000 females of the unisexual
species have been examined (Darevsky, 1966; Uzzell and Darevsky, 1972).
This includes the following lizards (100% in each case) that Darevsky
(1966) hatched from eggs in the laboratory; 276 aromeniaca, 113 dahli,
110 T'ostombekovi and 89 unisexualis.
Not one male of the parthenogenetic species of LaceT'ta has been hatched
in the laboratory; however, some eggs that fail to hatch contain ab-
normal male embryos (Darevsky and Kulikova, 1961; Darevsky, 1966).
A few natural males of aromeniaca and dahli (less than 10) have been
found (Darevsky, 1966, 1972); some actually are intersexes with well-
developed testes and poorly-developed oviducts. Perhaps most, or all,
are hybrids similar to the triploid male hybrid (with well-developed
testes and poorly-developed oviducts) of L. T'ostombekovi (<i?) x L. roaddei
(if) reported by Darevsky et al. (1973). Interspecific mating between
bisexual and unisexual species occurs under natural and artificial
conditions (Darevsky and Kulikova, 1961, 1964; Darevsky, 1966; Darevsky
and Danielyan, 1968; Darevsky et al., 1973). Offspring produced after
mating are either fertile diploid females (from parthenogenetic ova
that were not inseminated) or sterile triploid females (from diploid
ova that received an additional haploid complement from a spermatozoan) •
346
Darevsky (1966) hatched 56 females from a total of 16 armeniaca from

which he obtained eggs after maintaining the adults in artificial
hibernation in isolation from males; this included offspring from six
young adults that were in their first reproductive season. Females of
bisexual species " ... caught in early spring and maintained for a long
time without males laid infertile eggs which died early in the incuba-
tion period" (Darevsky, 1966: 119). Freyse and Muller (1962) raised
six females of armeniaca that were the offspring (100%) of a female given
to them by Darevsky. After reaching maturity in isolation from males,
each of these laid eggs. Although none hatched, some definitely con-
tained partly-developed embryos.
The unisexual species have reproductive systems similar to those of

females of bisexual species. In the mating season of the bisexual spe-
cies, however, when sperm apparently from recent matings are present
in the females' cloacae and oviducts, parthogenetic species contain
no sperm (Darevsky, 1958, 1966; Darevsky and Kulikova, 1961).
The mechanism by which parthenogenetic females produce ova containing

complete chromosome complements has not been fully determined, but
meiosis I appears to be nQrmal (Darevsky and Kulikova, 1961; Darevsky,
1966). Perhaps meiosis is preceded by a karyokinesis without a cyto-
kinesis, as in Cnemidophorus uniparens, and two normal meiotic divisions
restore the maternal condition (Uzzell, 1970; Cuellar, 1971).
Three of the unisexual species of Lacerta (armeniaca, dahZi, unisexualis)

are known to be diploids whose karyotypes are indistinguishable from
one another and from those of several bisexual species, but "some
fluctuation in the chromosome number was discovered in parthenogenetic
species ... " (Kupriyanova, 1969: 814). Although chromosome data are
not suggestive, external morphology, ecology and geographic distribu-
tion do suggest a hybrid origin (Darevsky and Kulikova, 1961; Darevsky,
1962, 1966). Electrophoretic mobilities of a variety of proteins sup-
port this interpretation. Bisexual species differ from one another in
some proteins and are rather homozygous. Unisexual species, however,
are rather heterozygous. The protein mobilities of each partheno-
genetic species suggest which pair of bisexual species constitute its
ancestors (Uzzell, 1972; Uzzell and Darevsky, 1972, 1973a,b). The fact
that various combinations of bisexual and unisexual species produce
viable hybrids is further evidence of their genetic similarities.
Family Xantusiidae: Genus Lepidophyma
Among about 15 species of xantusiids in the Western Hemisphere, Lepido-

phyma flavimaculatv.m is the only one for which all-female populations are
known; not all populations are unisexual, however. Only females (N = 57)
have been found at the southern end of its range in Panama; their re-
productive systems appear to be normal (Telford and Campbell, 1970).
Only one male was found in a series of 10 adults from the northern
edge of the range, but in other samples the sex ratio resembles that
of bisexual species (Bezy, 1972). Seven females from unisexual popula-
tions were diploids with homomorphic pairs of chromosomes; an additional
female was a diploid/triploid mosaic (Bezy, 1972). For L. flavimaculatum,
" ... there is, at present, no morphological, cytogenetic, or biogeo-
graphical evidence that hybridization preceded the evolution of uni-
sexuality" (Bezy, 1972: 23).
Family Gekkonidae: Three Genera

Among more than 650 species of geckoes in the world, three appear to
be parthenogenetic. Hemidactylus garnotii is a triploid species for which
347
no confirmed males are known, although Kluge and Eckardt (1969) examined
218 females; their reproductive systems consist of normal female struc-
tures. Skin grafts among three H. garnotii were all accepted, indicating
their possession of identical or highly similar histocompatibility
genes (Eckardt and Whimster, 1971).
Cuellar and Kluge (1972) examined 674 Lepidodactylus luguhris; 669 were
females, 4 were males. Most of the males and several females had ab-
errant reproductive organs, and one lacked them. Interspecific hybrid-
ization may not account for all the males; one was from Hawaii, where
no other species of the genus is known. No sperm were found in oviducts
of the 19 adult females examined for them. This diploid species with
homomorphic pairs of chromosomes rej ects skin grafts from Hemidactylus,
but different individuals of L. lugvbris are highly histocompatible
(Cuellar and Kluge, 1972).
The high chromosome number of Gehyra variegata ogasawarisimae suggests it

is a triploid species. This suggests it is parthenogenetic, as are all
other triploid species of lizards (Hall, 1970).
Family Agamidae: Genus Leiolepis
Only one species, Leiolepis triploida, is apparently parthenogenetic in

this large family of the Eastern Hemisphere. This triploid species is
known from only 33 females, probably from northern Malaya (Hall, 1970).
Some .are infertile and others are completely fertile (Peters, 1971).
This unisexual species does not seem to have had a hybrid origin (Pe-
ters, 1971).
Family Chamaeleonidae: Genus Brookesia
The only presumably parthenogenetic form in this large family is Broo-

kesia spectrum affinis from Za'ire; all 67 specimens known to Hall (1970)
were females.
Family Typhlopidae: Genus TyphUna
TyphUna (?) bramina is the only presumably parthenogenetic species of

snake; all 114 specimens for which McDowell (1974) determined sex were
females. The genus to which bramina should best be assigned (TyphUna or
Typhlops) is uncertain; the most reliable generic characteristics .for
these typhlopids are on the hemipenes, but no males of bramina are known.
E. Origin of Parthenogenetic Species
Twenty-seven species of reptiles (26 lizards, 1 snake) appear to have

unisexual populations in which females normally reproduce without males.
Parthenogenesis apparently occurs in representatives of six of the ap-
proximately 17 families of lizards and one of the approximately 10
families of snakes. In each of these families, however, most species
reproduce bisexually. Parthenogenesis appears to be relatively recently
derived, and probably evolved many times independently.
Evidence from external morphology, ecology, geographic distribution,

chromosome cytology and/or biochemistry indicates that the origin of
10 parthenogenetic species involved interspecific hybridization:
348
Cnemidophorus neomexicanus, C. tesselatus, C. uniparens, C. velox, C. exsanguis,

C. laredoensis, Lacerta armeniaca, L. dahli, L. rostombekovi and L. unisexualis.
Populations of these species are maintained parthenogenetically, al-
though individuals occasionally may hybridize with sympatric bisexual
species. There is no evidence suggesting a hybrid origin for three
parthenogenetic species: Cnemidophorus lemniscatus, Lepidophyma flavimaeulatum
and Leiolepis triploida. Two of these (C. lemniscatus, L. flavimaculatum) exist
as bisexual populations at some localities and unisexual populations
at others. Additional evidence is needed regarding the origin of 14
species, although it has been suggested that several evolved through
hybridization: Cnemidophorus dixoni, C. opatae, C. sonorae, C. flageUicaudus,
C. cozumela, C. rodecki, Gymnophthalmus underwoodi, Kentropyx borckianus, Leposoma
percarinatum, Hemidactylus garnotii, Lepidodactylus lugubris, Gehyra variegata
ogasa:warisimae, Brookesia spectrum affinis and Typhlina (?) bramina. These species
appear to be normally parthenogenetic, but individuals of some of them
occasionally may hybridize with sympatric bisexual species.
I. Evolution through Hybridization
A theory on the origin of parthenogenetic species through hybridization

involves several stages.
1. Two distinct diploid bisexual species, each adapted to a particular

set of environmental conditions, are distributed geographically such
that they do not coexist at any locality. Geographic separation, al-
though possibly slight, results from historical factors (evolutionary
history) and environmental factors (habitat distribution, physical and
ecological barriers) •
2. Changes in environmental factors (e.g., humidity, precipitation,

temperature) result in shifts in habitat distributions (Darevsky and
Kulikova, 1961; Wright, 1968; Wright and Lowe, 1968; Lowe et al.,
1970a). Distributions of the lizards shift with their habitats, which
brings them into contact in ecotones where their habitats merge. Each
species is well-adapted to its own habitat, so neither displaces the
other.
3. Interspecific hybridization occurs in the ecotone (Darevsky and

Kulikova, 1961; Lowe and Wright, 1966a,b; Lowe et al., 1970a). Male
and female hybrids are produced in equal numbers, since sex determining
mechanisms are similar in these species. Hybrids are sterile or have
low fertility; otherwise there would be no selection against bisexual-
ity among the hybrids and one or both of the original species may be-
come genetically swamped by extensive crossing and back-crossing. Thus,
hybridization is unproductive for both populations, and selection fa-
vors the evolution of premating isolating mechanisms (reinforcing mech-
anisms) that reduce or essentially eliminate continued hybridization
(Uzzell and Darevsky, 1972). Parents with a tendency to breed intra-
specifically produce successful offspring most similar to themselves
(including the tendency for intraspecific breeding), whereas those
that breed interspecifically produce relatively unproductive offspring.
4. Although they are reproductively unfit, the hybrids exhibit heter-

osis in other respects. They are healthy, robust, and able to outcom-
pete both parental species in the ecotone, even though they may be
outcompeted by each parental species in its respective habitat (Wright
and Lowe, 1968; Uzzell and Darevsky, 1972).
5. Evolution of premating isolating mechanisms (stage 3) requires gen-

erations, and production of reproductively unfit hybrids occurs for
some time (Neaves, 1971). These are diploids having a complete haploid
349
complement of chromosomes from each parental species (Lowe and Wright,

1966a,b). Their reduced fertility may result from faulty meiosis, per-
haps involving synapsis and balanced segregation. If any hybrid or hy-
brids happen to possess gene combinations that correct this, they may
be productive (Neaves, 1971). If their offspring also retained the het-
erotic traits of the FI hybrids, they would be strong ecological compet-
itors in the ecotone and reproductively successful. Events that would
permit this would be a premeiotic karyokinesis that is not followed by
cytokinesis in ovarian cells, so that primary oocytes are tetraploid.
This followed by meiosis involving synapsis or sister-chromosomes (pro-
duced by the endoduplication, which would avoi~ formation of complicated
multivalents), would result in production of diploid ova having all the
successful heterozygosity of the hybrid female. This, in fact, may be
the ~echanism of oogenesis in most unisexual vertebrates (Uzzell, 1970)
and apparently is in CnemidophoY'Us uniparens, the only parthenogenetic
reptile in which meiosis has been studied thoroughly (Cuellar, 1971).
Thus, it is hypothesized that an unusual hybrid lizard could happen
to have the combination of genes that results in production of diploid
ova, which then undergo parthenogenetic development (Neaves, 1969,
1971). The resulting offspring have essentially the same adaptive
traits as their mother, excepting random mutations. Among the hybrids,
only the parthenogenetic females are successful reproductively. The
males become. extinct as selection continues against additional inter-
specific hybridization.
6. Once the hybrids occupy a habitat and perpetuate themselves, they

constitute a new species (Darevsky, 1966, 1967; Lowe and Wright, 1966a,b;
Wright and Lowe, 1968). This species, as all others, is subject to nat-
ural selection.
7. The new parthenogenetic species has some advantages over its bisexual
ancestors. Each normal individual can produce offspring, and any indi-
vidual dispersing into a satisfactory and yet unoccupied area may es-
tablish a colony (Darevsky, 1966; Wright and Lowe, 1968). The mode of
reproduction, however, while preserving heterozygosity, sacrifices the
rapid genetic adaptability allowed by bisexual reproduction. It would
appear that under conditions of severe selection, unisexual species
may become extinct more readily than bisexual species (Uzzell, 1970).
However, under conditions of catastrophic selection in which the uni-
sexual species is not quite exterminated, several well-adapted partheno-
genetic clones could emerge; these would efficiently colonize or recol-
onize suitable areas when the stress is relieved (Zweifel, 1965). Adapt-
ability of parthenogenetic species may be aided also by tolerance of
higher mutation rates than are withstood by bisexual congeners; this
may be reflected in chromosome polyrnorphisms (Zweifel, 1965; Kupriya-
nova, 1969; Lowe et al., 1970a,b).
8. Hybridization between the diploid parthenogenetic species and a

bisexual species (one of the parental species or a third species) may
occur at localities where they coexist. The new hybrids would be tri-
ploids, containing both sets of chromosomes from the parthenogenetic
species and a third set from the male (Darevsky, 1966; Lowe and \vright,
1966a,b). Since the male delivers two kinds of sperm (with an X- or
Y-chromosome), triploids occur in both sexes in equal frequency (wright
and Lowe, 1967a; Cole et al., 1969; Darevsky et al., 1973). Depending
on the circumstances of competition, mutation, and selection, the tri-
ploid hybrids could either evolve into a parthenogenetic species as
did the diploids earlier, or they could become extinct. Once they occupy
a habitat and perpetuate themselves, they constitute a new species
(Lowe and Wright, 1966a,b).
350
9. No tetraploid species of reptiles are known, but tetraploid hybrids

resulting from interbreeding between triploid parthenogenetic species
and diploid bisexual species are known (Lowe et al., 1970a; Neaves,
1971) .
II. Evolution without Hybridization
Bezy (1972) discussed the factors that may have resulted in evolution
of parthenogenesis in Lepidophyma. They may also apply to other species.
1. Bisexual populations have highly restricted habitat requirements

that are best fulfilled in relatively limited areas. Such areas are
ecological edge situations (including temporary clearings) within con-
siderably larger areas of habitat (e.g., forest) that is relatively
unsuitable for these lizards. Thus, bisexual populations are locally
dense and widely separated; normally they have little or no contact
with one another.
2. Natural selection occurs essentially independently on the local pop-

ulations. They are under strong selective pressure as ecological suc-
cession proceeds and the locality becomes increasingly inhospitable
for them. Many populations become exterminated as succession occurs
and their genes survive only through those few individuals that dis-
perse successfully and find suitable habitat to populate elsewhere.
3. If any fit individual happens to evolve a gene combination that al-

lows parthenogenetic reproduction, its offspring would be at a selec-
tive advantage in dispersing and establishing populations.
In both theories, parthenogenesis is visualized as emanating from the

fortuitous occurrence of a gene combination that permits it in one
female (or a few). In species that evolved through hybridization, the
importance of hybridization was in producing heterotic gene combinations
that were successful in the hybrid habitat. The primary selective ad-
vantage of parthenogenesis was in overcoming sterility, or low fertil-
ity in the hybrids, and maintaining their heterozygosity. In the absence
of suitable habitat in which they could outcompete their parents, the
hybrids would probably become extinct. In Lepidophyma, however, the
primary selective advantage of parthenogenesis was in facilitating
colonization of new localities but not new habitats, since most popu-
lations experience strong selection for dispersal.
For both theories, the same critical information is lacking to fully

explain the origin of parthenogenesis in individual females. What
genetic changes are involved? Apparently they can occur in either hy-
brid or non-hybrid animals, and this seems to suggest that partheno-
genetic species could evolve spontaneously within any population of
lizards. However, in the absence of ecological conditions such as those
discussed above, unusual females that may fortuitously develop partheno-
genetic reproduction probably would be inseminated by conspecific males
and the polyploid offspring; which would be sterile or have a low fer-
tility, would be at a selective disadvantage.
F. Problems for Study
1. Egg Activation: ,ihat activates the diploid or triploid ova to de-

velop in parthenogenetic species? Perhaps it involves the presence of
two or more complete chromosome complements and their accompanying
micro-tubule systems in the environment of the ovum? The answer also
351
may lie among the mechanism that causes haploid ootids and spermatids
to cease division activity.
2. Sex Determination: Is the basis of all-femaleness in triploid spe-

cies due to possession of only X-chromosomes, the absence of Y-chromo-
somes, possession of considerably more heterochromatin than occurs in
diploid cells, or a combination of factors?
3. Dosage Compensation: What problems of gene dosage are encountered

in polyploid species? Electrophoretic mobilities of proteins indicate
that more genes are active in triploids than in diploids (Neaves and
Gerald, 1969; Neaves, 1969). In cells of triploids, however, secondary
constrictions normally occur on only two homologous chromosomes instead
of all tnree (Lowe et al., 1970a).
4. Mutation Rate: How does mutation rate of parthenogenetic species

compare with that of bisexual species, and how does this affect their
relative adaptability (Zweifel, 1965)?
5. Ecology: What are the important ecological interactions between

parthe'nogenetic and bisexual species that coexist at particular local-
ities (Medica, 1967; Christiansen, 1971)?
References
Becak, W.: Constituicao cromossomica e mecanismo de determinacao do

sexo em ofidios Sul-Americanos. I. Aspectos Cariotipicos. Mem.
Inst. Butantan 32, 37-78 (1965).
Bezy, R.L.: Karyotypic variation and evolution of the lizards in the
family Xantusiidae. Contr. Sci., Nat. Hist. Mus., Los Angeles County,
no. 227, 1 - 29 ( 1 972) .
Bons, J~Bons, N.: Un cas d'intersexualite complete chez un lezard
anguide du 14aroc Ophisaurus koeUikeri (Gunther 1873). C.R. Acad. Sci.
268, Ser. D, 695-696 (1969).
Callard, I.P., Chan, S.W.C., Potts, M.A.: The control of the reptilian
gonad. Amer. Zool. 12, 273-287 (1972).
Charnier, M.: Action de-la temperature sur la sex-ratio chez l'embryon
d'Agamaagama (Agamidae, Lacertilien). C.R. Soc. Biol. 160, 620-622
(1966).
Christiansen, J .L.: Reproduction of Cnemidophorus inornatus and Cnemido-
phorus neomexicanus (Sauria, Teiidae) in northern New Mexico. Amer.
Mus. Novitates 2442,1-48 (1971).
Christiansen, J.L.: Natural and artificially induced oviducal and
ovarian growth in two species of Cnemidophorus (Sauria, Teiidae).
Herpetologica 29, 195-204 (1973).
Christiansen, J.L~ Ladman, A.J.: The reproductive morphology of Cnemi-
dophorusneomexicanus xC. inornatus hybrid males. J. Morph. 125,367-
377 (1968). -
Christiansen, J.L., Degenhardt, W.G., White, J.E.: Habitat preferences
of Cnemidophorus inornatus and C. neomexicanus with reference to conditions
contributing to their hybridization. Copeia 1971, 357-359.
Cole, C.J., Lowe, C.H., Wright, J.W.: Sex chromosomes in teiid whiptail
lizards (genus Cnemidophorus). Amer. Mus. Novitates 2395, 1-14 (1969)
Cuellar, 0.: Additional evidence for true parthenogenesis in lizards
of the genus Cnemidophorus. Herpetologica 24, 146-150 (1968).
Cuellar, 0.: Egg transport in lizards. J. Morph. 130, 129-135 (1970).
Cuellar, 0.: Reproduction and the mechanism of meiotic restitution in
the parthenogenetic lizard Cnemidophorus uniparens. J. Morph. 133, 139-
165 (1971).
352
Cuellar, 0., Kluge, A.G.: Natural parthenogenesis in the gekkonid liz-

ard LepidodactyZus lugubris. J. Genet. .§.1, 14-26 (1972).
Danielyan, F.D.: A comparative study of the population density and
migrations of parthenogenetic and bisexual rock lizards in Armenia.
Zool. Zhur. 50, 145-147 (1971). (Russian, English summ.)
Darevsky, I.S.:~atural parthenogenesis in certain subspecies of rock
lizards, Lacerta s=icoZa Eversmann. Doklady Akad. Nauk SSSR, Biol.
Scis. (English Transl.) 122, 877-879 (1958).
Darevsky, 1.5.: On the origrn-and biological role of natural partheno-
genesis in the polymorphous group of Caucasian rock lizards Lacerta
saxicola Eversmann. Zool. Zhur. 41, 397-408 (1962). (Russian, English
summ.) -
Darevsky, 1.5.: Natural parthenogenesis in a polymorphic group of Cau-
casian rock lizards related to Lacerta s=icola Eversmann. J. Ohio
Herpetol. Soc. 5, 115-152 (1966).
Darevsky, 1.5.: On-the taxonomic status of parthenogenetic forms of
rock lizards (Lacerta saxicola Eversmann) as related to the applica-
tion of species criteria to agamic species. Zool. Zhur. 46, 413-419
(1967). (Russian, English summ.) -
Darevsky, 1.5.: Zur Verbreitung einiger Felseidechsen des Subgenus
Archaeolacerta in der TurkeL Bonner Zool. Beitr. 23, 347-351 (1972)
Darevsky, 1.5., Danielyan, F.D.: Diploid and triploid progeny ariSing
from natural mating of parthenogenetic Lacerta armeniaca and L. v.ni-
sexualis with bisexual L. saxicola valentini. J. Herpetol. 2, 65-69
(1968). -
Darevsky, 1.5., Kulikova, V.N.: Naturliche Parthenogenese in der poly-
morphen Gruppe der Kaukasischen Felseidechse (Lacerta saxicola Evers-
mann). Zool. Jb. Abt. Syst. Okol. Geogr. Tiere 89, 119-176 (1961).
Darevsky, 1.5., Julikova, V.N.: Natural triploidy within a polymorphous
group of Lacel'ta saxicola Eversmann, resulting from hybridization
between bisexual and parthenogenetic varieties of this species.
Doklady Akad. Nauk SSSR, Bioi. Scis. 158, 202-205 (1964). (Transl.,
Janis A. Roze). --
Darevsky, 1.5., Uzzell, T., Kupriyanova, L.A., Danielyan, F.D.: TriplOid
hybrid males in sympatric populations of some parthenogenetic and
bisexual species of rock lizards of the genus Lacerta L. Bull. Moscow
Soc. Nat., Biol. Ser. 78, 48-58 (1973). (Russian, English summ.)
Duellman, W.E., Zweifel, ~G.: A synopsis of the lizards of the sexli-
neatus group (genus Cnemidophorus). Bull. Amer. Mus. Nat. Hist. 123,
155-210 (1962).
Eckardt, M.J., Whimster, I.'L: Skin homografts in the all-female gek-
konid lizard Hemidactylus garnotii. Copeia 1971, 152-154.
Fitch, H.S.: Reproductive cycles in lizards and snakes. Musc. Publ.
no. 52, Univ. Kansas Mus. Nat. Hist., 1-247 (1970).
Forbes, T.R.: Intersexuality in reptiles. In: Intersexuality in Verte-
brates including Man (eds. C.N. Armstrong and A.J. Marshall), pp.
273-283. London-New York: Academic Press 1964.
Freyse, K., Muller, G.: Die Parthenogenese bei Lacerta so.xicola armeniaca
Mahely, 1909. Zool. Garten 26 (new ser.), 348 (1962).
Fritts, T.H.: The systematics of the parthenogenetic lizards of the
CnemidophoY'Us cozwnela complex. Copeia 1969, 519-535.
Gehlbach, F.R.: Herpetology of the Zuni Mountains region, northwestern
New Mexico. Proc. U.S. Natl. Mus. 116, 243-332 (1965).
Hall, W.P.: Three probable cases of parthenogenesis in lizards (Agami-
dae, Chamaeleontidae, Gekkonidae). Experientia 26, 1271-1273 (1970).
Hardy, L .M.: Intersexuality in a Mexican colubrid snake (Pseudoficimia).
Herpetologica 26, 336-343 (1970).
Hoge, A.R., Belluomini, H.E., Schreiber, G., Penha, A.M.: Sexual ab-
normali ties in Bothrops insularis (Amaral) 1921. Mem. Inst. Butantan
29, 17-88 (1959).
353
Hoge, A.R., Belluomini, H.E., Schreiber, G., Penha, A.M.: Anomalias

sexuais em Bothrops insula:ris (Amaral) 1921. (Serp. Crot.). Anais da
Acad. Brasileira de Cienc. 33, 259-264 (1961).
Hoogmoed, M.S.: Notes on the herpetofauna of Surinam. IV. The lizards
and amphisbaenians of Surinam. The Hague: Junk 1973.
Kluge, A.G., Eckardt, M.J.: Hemidactylus garnotii Dumeril and Bibron,
a triploid all-female species of gekkonid lizard. Copeia 1969, 651-
664. --
Kupriyanova, L.A.: Karyological analysis of lizards of subgenus Archaeo-
lacerta. Tsitologiya 11, 803-814 (1969). (Russian, English summ.)
Licht, P.: Environmentar-physiology of reptilian breeding cycles: Role
of temperature. Gen. Compo Endocrinol. Suppl. 3, 477-488 (1972).
Lowe, C.H., Wright, J. W.: Species of the Cnemidophorus ex sanguis subgroup
of whiptail lizards. J. Arizona Acad. Sci. 3, 78-80 (1964).
Lowe, C.H., Wright, J.W.: Chromosomes and karyotypes of cnemidophorine
teiid lizards. Mammal. Chromosomes Newslett., no. 22, 199-200 (1966a).
Lowe, C.H., Wright, J.W.: Evolution of parthenogenetiC-species of
Cnemidophorus (whiptail lizards) in western North America. J.
Arizona Acad. Sci. 4, 81-87 (1966b).
Lowe, C.H., Wright, J.W., Cole, C.J., Bezy, R.L.: Natural hybridization
between the teiid lizards CnemidophoY'Us sonorae (parthenogenetic) and
Cnemidophorus tigris (bisexual). Syst. Zool. 19, 114-127 (1970a).
Lowe, C.H., Wright, J.W., Cole, C.J., Bezy, R~.: Chromosomes and evo-
lution of the species groups of CnemidophoY'Us (Reptilia: Teiidae).
Syst. Zool. li, 128-141 (1970b).
MacLean, W.P., III: A case of intersexuality in Bothrops moojeni Hoge.
Copeia 1968, 170.
Maslin, T.:P":"":All-female species of the lizard genus CnemidophoY'Us,
Teiidae. Science 135, 212-213 (1962).
Maslin, T.P.: The sex of hatchlings of five apparently unisexual spe-
cies of whiptail lizards (Cnemidophorus, Teiidae). Amer. Midl. Nat.
76, 369-378 (1966).
Maslin, T.P.: Skin grafting in the bisexual teiid lizard Cnemidophorus
sexlineatus and in the unisexual C. tesselatus. J. EXp. Zool. 166, 137
-149 (1967).
Maslin, T.P.: Taxonomic problems in parthenogenetic vertebrates. Syst.
Zool. 17, 21 9 - 2 3 1 ( 1 96 8) .
Maslin, T~.: Parthenogenesis in reptiles. Amer. Zool. 11, 361-380
(1971a) .
Maslin, T.P.: Conclusive evidence of parthenogenesis in three species
of CnemidophoY'Us (Teiidae). Copeia 1971,156-158.
McCoy, C.J., Maslin, T.P.: A review of the teiid lizardCnemidophoY'Us
cozume lus and the recognition of a new race, Cnemidophorus cozume lus
rodecki. Copeia 1962, 620-627.
MCDowell, S.B.: A catalogue of the snakes of New Guinea and the Solo-
mons, with special reference to those in the Bernice P. Bishop
Museum, Part I. Scolecophidia. J. Herpetol. 8, 1-57 (1974).
McKinney, C.O., Kay, F.R., Anderson, R.A.: A newall-female species
of the genus Cnemidophorus. Herpetologica 29, 361-366 (1973).
Medica, P.A.: Food habits, habitat preference, reproduction, and diurnal
activi ty in four sympatric species of whiptail lizards (Cnemidophorus)
in south central New Mexico. Bull. Southern California Acad. Sci.
66, 251-276 (1967).
Moon, R.G.: Heteromorphism in a kinosternid turtle. Mammal. Chromosomes
Newslett. 15, 10-11 (1974).
Neaves, W.B.:~denosine deaminase phenotypes among sexual and partheno-
genetic lizards in the genus CnemidophoY'Us (Teiidae). J. Exp. Zool.
171,175-183 (1969).
Neaves, W.B.: Tetraploidy in a hybrid lizard of the genus Cnemidophorus
(Teiidae). Breviora 381, 1-25 (1971).
Neaves, W.B., Gerald, P:5:: Lactate dehydrogenase isozymes in partheno-
genetic teiid lizards (CnemidophoY'Us). Science 160, 1004-1005 (1968).
354
Neaves, W.B., Gerald, P.S.: Gene dosage at the lactate dehydrogenase

b locus in triploid and diploid teiid lizards. Science 164, 557-
559 (1969). -
Peccinini, D.: Chromosome variation in populations of Cnerrridophor'Us
~emniseatus in the Amazon Valley (Sauria, Teiidae). Ciencia e Cultura
23,133-136 (1971).
Pennock, L.A.: Triploidy in parthenogenetic species of the teiid lizard,
genus Cnemidophor'Us. Science ill, 539-540 (1965).
Peters, G.: Die intragenerischen Gruppen und die Phylogenese der
Schmetterlingsagamen (Agamidae: Leio~epis). Zool. Jb. 98, 11-130
(1971) •
Pieau, C.: Gonadogenesis in embryos of two tortoise species, &ws orbi-
eu~aris L. and Testudo graeea L., at different temperatures. Soviet.
J. Develop. Biol. 4, 8-14 (1974). (Transl. from Ontogenez 4, 11-16
(1973». - -
Saxon, J.G.: Sexual behavior of a male checkered whiptail lizard,
CnerrridophoT'Us tesse~atU8 (Say). Southwestern Nat. 13, 454-455 (1968).
Scudday, J.F.: A new species of lizard of the Cnerrridophorus tesse~atus
group from Texas. J. Herpetol. 7, 363-371 (1973).
Stefan, Y.: Contribution a l'etude-experimentale de l'intersexualite
chez un Chelonien: &ws ~eprosaS. Bull. Biol. France Belgique 97,
363-467 (1963).
Taylor, H.L.: Morphological variation in selected populations of the
teiid lizards Cnemidophor'Us ve~ox and Cnerrridophor'Us inornatus. Univ.
Colorado Studies, Ser. Biol. 21, 1-27 (1965).
Taylor, H.L., Medica, P.A.: Natural hybridization of the bisexual teiid
lizard Cnerrridophorus inornatus and the unisexual Cnerrridophorus perp~exus
in southern New Mexico. Univ. Colorado Studies, Ser. Biol. 22, 1-9
(1966) .
Taylor, H.L., Walker, J.M., Medica, P.A.: Males of three normally
parthenogenetic species of teiid lizards (genus Cnemidophorus). Copeia
1967, 737-743.
Telford, S.R., Campbell, H.W.: Ecological observations on an all female
population of the lizard Lepidophyma fZavimaeu~atum (Xantusiidae) in
Panama. Copeia 1970, 379-381.
Thomas, R.: The smaller teiid lizards (Gymnophtha~mus and Baehia) of the
southeastern Caribbean. Proc. Biol. Soc. Wash. 78, 141-154 (1965).
Tinkle, D.W.: Observations on the lizards Cnemidophor'Us tigris, Cnemido-
phorus tesselZatus and Crotaphytus wislizeni. Southwestern Nat. 4, 195-
200 (1959). -
Tomlinson, J.T.: Improper use of the word bisexual. Syst. Zool. 11,
212 (1968).
Uzzell, T.: Meiotic mechanisms of naturally occurring unisexual verte-
brates. Amer. Nat. 104, 433-445 (1970).
Uzzell, 'r.: Hybrid origin of unisexual lizards, genus Laeerta. Amer.
Philos. Soc. Yr. Book 1972, 401-402 (1972).
Uzzell, T., Barry, J .C.: Leposoma perearinatum, a unisexual species related
to L. guianense; and Leposoma ioanna, a new species from pacific coastal
Colombia (Sauria, Teiidae). Postilla 154, 1-39 (1971).
Uzzell, T., Darevsky, I.S.: Biochemical and evolutionary studies of
unisexual and bisexual lizards of the Laeerta s=ieo~a complex. In:
Theme No.5: ttudes moleculaires des differences entre les especes.
Proc. 17th Internatl. Congr. Zool" Monte-Carlo, 25-30 Sept. 1972.
Uzzell, T., Darevsky, 1. S.: The relationships of Laeerta portsehinskii
and Laeertaraddei (Sauria, Lacertidae). Herpetologica 29,1-6 (1973a).
Uzzell, T., Darevsky, I.S.: Electrophoretic examination of Laeerta mixta,
a possible hybrid species (Sauria, Lacertidae). J. Herpetol. 7,
11-15 (1973b). -
Vanzolini, P. E.: Unisexual Cnerrridophor'Us ~emniseatus in the Amazonas Valley:
A preliminary note (Sauria, Teiidae). Papeis, Avulsos Zool., Sao
Paulo 23, 63-68 (1970).
355
Vivien, J.H.: Intersexualite spontanee chez Pseudemys scripta eZegw!S S.,

un Chelonien aquatique. C.R. Soc. Biol. 158, 185-187 (1964).
Vorontsov, N.N.: The evolution of the sex chromosomes. In: Cytotaxo-
nomy and Vertebrate Evolution (eds. A.B. Chiarelli and E. Capanna),
pp. 619-657. London-New York: Academic Press 1973.
Wright, J .W.: Variation in two sympatric whiptail lizards (Cnemidophorus
inornatus and C. veZox) in New Mexico. Southwestern Nat. 11, 54-71
(1966). -
Wright, J .W.: A new uniparental whiptail lizard (genus Cnemidophorus)
from Sonora, Mexico. J. Arizona Acad. Sci. 4, 185-193 (1967).
Wright, J.W.: Variation in three sympatric sibling species of whiptail
lizards, genus Cnemidophorus. J. Herpetol. 1, 1-20 (1968).
Wright, J.W.: Evolution of the X1X2Y sex chromosome mechanism in the
scincid lizard ScinceZZa ZateraZe (Say). Chromo soma 43, 101-108 (1973).
Wright, J.W., Lowe, C.H.: The rediscovery of Cnemidophorus arizonae Van
Denburgh. J. Arizona Acad. Sci. 3, 164-168 (1965).
Wright, J.W., Lowe, C.H.: Hybridization in nature between parthenoge-
netic and bisexual species of whiptail lizards (genus Cnemidophorus ) .
Amer. Mus. Novitates 2286, 1-36 (1967a).
Wright, J.W., Lowe, C.H.: Evolution of the alloploid parthenospecies
Cnemidophorus tesseZatus (Say)., Mammal. Chromosomes Newslett. 8, 95-96,
(1967b). -
Wright, J .W.', Lowe, C.H.: Weeds, polyploids, parthenogenesis, and the
geographical and ecological distribution of all-female species of
Cnemidophorus. Copeia 1968, 128-138.
Zweifel, R.G.: Variation in and distribution of the unisexual lizard,
Cnemidophorus tesseZatus. Amer. Mus. Novitates 2235, 1-49 (1965).
Intersexuality in Birds. Study of the Effects of Hybridization
and Post-Embryonic Ovariectomy
L. Gamat
Many examples of spontaneous and experimentally induced intersexuality

have been reported in birds. Several reports have been devoted to the
presentation of known cases and interpretations have been proposed
taking into consideration the process of sexual differentiation (Domm,
1939; Benoit, 1950a,b; Taber, 1964).
Spontaneous sexual-inversion of the gonads occurs mostly in the female.
Riddle et al. (1942) however, have successfully bred a line of pigeons
in which the males frequently had a hermaphroditic left gonad. In ex-
periments with birds, a kind of paradoxical situation is encountered
concerning .the factors which control the differentiation of the gonads.
Indeed, several researchers have succeeded in feminizing the left gonad
of males by the administration of estrogens to the fowl embryo (Wolff,
1936; Van Tienhoven, 1957) and the quail (Haffen, 1965-1969). Such an
artificial feminization, which is usually transitory despite repeated
hormonal injections after hatching, sometimes persists for several
years in the hen (Domm, 1939), the herring gull (Boss and Witschi,
1947) and the turkey (Japp et al., 1951). Such a feminization as a
spontaneous event is unknown in these species. On the other hand, the
treatment ·of embryos with androgenic steroids (Wolff et al., 1948;
Van Tienhoven, 1957) does not affect the differentiation of the gonads
of genetic females although spontaneous modifications do occur in them.
In vitro cultures of embryonic gonads carried out by Wolff and Haffen

(1952 a,b) as well as results of ovariectomies (Benoit, 1923-1950;
Domm, 1939) show that the development of the female gonads is strongly
influenced by the asymmetry of their development and by the fact that
the ovary usually retains a part of the medullary component of the
early gonad (anlagen).
The frequent occurrence of anomalies in the sexual differentiation and

gametogenesis in hybrids of birds suggests a genetic control which is
attributed to an aberrant chromosomal constitution of the cells (Riddle
and Johnson, 1939) or to a decrease in the activity of female deter-
mining genes (Bissonnette, 1942) or which, more generally may arise
from a genetic disturbance of the control over the functions of the
gonadal constituents.
Although most of the observations and experiments up to this time have

been carried out on chickens, pheasants and pigeons, we became inter-
ested in intersexuality ensuing from the hybridization of two genera
of ducks (Anas platyrhynchos and Cairina moschata). The analysis of the phe-
nomenon was accomplished histologically and by the use of a histochem-
ical method which localizes emzymes (hydroxysteroido-dehydrogenases:
~5-3S-HSDH and ~5-17S-HSDH) occurring in steroidogenesis. It is there-
fore possible to follow the development of gonadal constituents during
normal ontogenesis of the hybrids and the parents, as well as after
experimentation (action of sexual hormones and ovariectomy) .
357
A. Development of the Gonads and of Gametogenesis in Hybrid Ducks
I. Investigation of the Causes for the Changes in Males Gametogenesis

(in Collaboration with C.R. Marchand)
The male hybrids of the two reciprocal cross-breedings (A. platyrhynchos

if x C. moschata 'i' and C. moschata (j' x A. platyrhynchos 'i' are sterile because
spermatogenesis fails at the meiotic prophase. The cytological studies
of Crew and Koller (1935-1936) indicate that the sterility is not nec~
essarily due to a failure in the pairing of chromosomes but that it
may arise from an alteration in the spindle of the dividing cell. An
analysis of the chromosomal constitution has been carried out by Yama-
shina (1941-1942), then by Mott et al. (1968) who noted that differ-
ences between the first and fourth pairs of chromosomes can induce
the sterility of the hybrids.
From endocrinological investigations of these animals it is evident

that the failure of spermatogenesis cannot be attributed to a lack of
hormones since injections of gonadotropic hormones or steroids do not
modify the development of the germ cells of .the testes (Lavedan and
Demay, 1951; Popescu, 1960; Pavlov et al., 1966; Rigdon, 1968). From
the present cytophysiological studies on the pituitary and the gonads
(Faber, 1957; Marchand et al., 1969-1972; Marchand, 1972; Marchand and
Gomot, 1972-1973; Marchand and Bugnon, 1973), it seems that the hor-
monal factors involved in gametogenesis are comparable to those of the
parent genera of the ducks. One proof of the hypophysial activity of
the hybrid male was provided by the normal development of spermato-
genesis in the testis of a Peking duck which was transplanted into a
male hybrid that had been castrated beforehand (Gomot, 1970). With
regard to the testis, the cytological studies do not show indications
of intersexual characters but experiments with embryonic testicular
grafts (Lutz-Ostertag and David, 1961; Lutz-Ostertag, 1965) demonstrate
that these gonads do not have a normal male constitution since they
have a feminizing effect on the embryos when the chicken host is a
male. It is therefore probable that the hybridization impairs the ge-
netic control of the production of estrogens in the male, which accord-
ing to Narbaitz and Teitelman (1965) is involved in the mechanism of
the differentiation of the testes in the embryos of the chicken. The
feminizing properties of the testis do not persist because the grafting
of fragments from the testes of an adult hybrid has no effect on the
embryos of the chicken.
II. Development of the Female Gonads
In female hybrids intersexual characters are much more distinct than

in males and appear different according to the type of cross-breeding.
1. Hybridization Cairina moschata if x Anas platyrhynchos'i'
During the embryonic life there is a very distinct predominance of the

medulla. In the weeks after hatching a regression of the cortex and
of the germinal cells of the ovary can be seen (Lutz-Ostertag and Gomot,
1960). The right gonad, although reduced in size in the embryo (Fig. 1a),
sometimes persists in the adult (Fig. 1c), but it is smaller than the
left one. The left gonad of the adult (1 year) has the form of a red
colored band (Fig. 1b), consisting of a network of connective tissue
with rich vascularization which surrounds islets and cords of indif-
ferentiated cells and large cells with 3B-HSDH-activity. We have never
seen follicles, although Rigdon (1967-1968) reports observing some
358
degenerating follicles among the structures which resemble the sterile

seminiferous tubules of the left ovotestis. He also observed that their
number and their size increase after administration of estradiolbenzoate
(15 mg in 9 intramuscular injections apportioned during 35 days before
death). This author believes that a deficiency in the secretion of
estrogens could be the cause of the sterility of the gonads.
Fig. 1 a-c. Genital apparatus of the female hybrid (d'C. moschata x 'i' A.
platyrhynchos) . (a) Gonads of an embryo of 20 days. The cortex of the
right gonad (rg) is regressed nearly completely. 19 left gonad; ov
oviduct (x 10). (b) Aspect of the left gonad of an adult female hybrid
(1 year) (x 1). (c) Hypertrophied left gonad (lg) and right gonad (rg)
of a 2 year old female (x 0,8)
The size of the right gonad shows a high degree of individual variation.
In most cases it is very rudimentary and reduced to a few strings of
cells in connection with the mesonephric tubules which persist along-
side the vena cava. Sometimes it is as well developed as the left gonad
and has the same structure.
These gonads can be compared to the proliferations of cells originating
from the medulla in close association with the mesonephric rudiment.
They have no stimulatory action on the oviduct which, however, remains
very sensi ti ve to es trogens.. Its weight increases from 2 g to 57 g
after 9 days of treatment with estradiolbenzoate at the rate of 5 mg/day.
The uterus secretes considerable amounts of viscous material.
2. Hybridization Anas platyrhynchos d' x Cairina moschata 'i'

(in Collaboration with A. Deray)
In this cross breeding the gonads of the females have a normal morpho-
genesis with predominance of the medullary part which is comparable to
the situation which has been described by Lutz-Ostertag (1965) in the
embryos of the crossing between Khaki-Campbell d' x C. moschata 'i'. After
hatching the ovary develops normally and the hybrids start egg-laying
at the age of 7 1/2 months. Their eggs are small (26 - 39 g) whereas
those of C. moschata weigh between 65 and 89 g and the eggs of Peking
359
ducks (A. platyrhynehos) 60 - 75 g. The number of eggs laid, which is

around 30 to 40 during the first year, decreases to a dozen the second
year after which egg-laying ceases. The histological study of the go-
nads (Gomot and Ardiet, 1966; Gomot et al., 1966; Deray and Gomot,
1969; Deray, 1970) shows that the cortex of the left ovary atrophies
progressively, whereas the differentiation of testicular tubules occurs
with or without gonocytes. During this transformation of the left go-
nad the development of gonadic nodules corresponding to the right ovary
are often observed (Table 1). The structure of these nodules is vari-
able. Sometimes they consist of small ovotestis (Fig. 2 a-c) and some-
times they are devoid of germinal cells. The interpretation of the
histological observations which can explain the formation of the right
gonad is summarized in Fig. 3.
Fig. 2 a-c. Structure of the hypertrophied right gonad of a female

hybrid (d' A. platyrhynehos x <jl C. mosehata) 18 months old. (a) General
aspect of a transversal section with testicular tubules (male compo-
nent) and some parietal oocytes (oe) (x 60). (b) Aspect of testicular
tubules in which spermatogenesis becomes abnormal at the onset of
meiosis (x 250). (c) Section of one of the persisting oocytes (x 250)
Table 1. Observations on the development of a right gonad under natural and experimental conditions in '"mo
females of A. platyrhynchos, C. moschata and the hybrid d' A. pZatyrhynchos x 'i' C. moschata
Female ducks Conditions observed in the rights gonads

Germinal component present
Conditions lotal number Number of Germinal
Type of observa- of observa-
? component only component
right gonads
tlon tions present well de- atresia d' component Ovo- absent
veloped only only testis
ovary
natural 24 4 4 0 0 0 0 Ii
e exogenous i
Coirino x 7 3 3 0 0 0 0
moschofa p hormones
e
r ovari -
i
m ectomy 21 21 7 9 1 3 1
. I
natural 22 0 0 0 0 0 0
I
Anos e exogenous !
x 10 0 0, t 0 0 0 0
plofyrhynchos p hormones
I
e
r ovari- I
i ectomy 6 0 0 I 0 0 0 0
!
m. I
natural 16 7 4 I 0 0 1 2
Hybrid I
(Anos d' x e exogenous I I
x 16 2 2 0 0 0 0
Co/rino ~ ) p I
e hormones I
r. I
I ovari-
m . ectomy I
14 14 0 I 2 1 0 11
._- . ___ . .1 t.....- .. ---~--
361
The study of the development of germinal cells and steroidogenic cells

of the embryos, juvenile and adult females, provides indications for
the understanding of sex-inversion (Deray and Gomot, 1974).
On the fifth day of incubation, the gonocytes start to invade the geni-
tal ridges, this being earlier than in the females of the parental
species. But after the seventh day the number of germinal cells is
smaller than in the parents. However the stage of gonial multiplica-
tion is as active as in the females of A. platyrhynchos and C. moschata
and consequently the number of germ cells is nearly as large as in the
parental females when the embryos have reached the age of 18 - 20 days.
But between the 20th and 22nd days the hybrid ovary undergoes a con-
siderable loss of germ cells as it starts to enter meiosis. Beginning
at this stage the number of germ cells is reduced but those which re-
main enter vitellogenesis after hatching. Vitellogenesis progresses
more rapidly in the hybrid than in the females of the parent species.
The first egg-laying therefore is precocious, and the eggs are small.
The localization of steroidogenic cells with 3S-HSDH and 17S-HSDH ac-

tivity has been studied throughout the development of the gonads of the
hybrids and compared with that of females from A. platyrhynchos and C.
moschata (Der:ay, 1.970; Deray, 1973a,b; Deray and Gomot, 1973-1974).
The appearance of 3S-HSDH activity is precocious in the three catego-
ries of females (8th, 9th day of incubation). In the embryo the abun-
dance of "clear cells" or "interstitial medullary cells" which are
responsible for this enzymatic activity, is rapidly visible and their
migration towards the cortex where they will form the theca interna
is more rapid than in the females of the parents.
The subepithelial cells with 11B-HSDH activity appear later on day 19

and form a nearly· continuous border below the germinal epithelium of
the ovary of the hybrids, whereas in the parental species they show
a tendency to thin out in certain regions.
After hatching, the ovaries of the juvenile ducks show a particular

development of the steroidogenic cells. During the first months there
is a very strong reactivity of the cells with 3S-HSDH activity which
form a thick border line (Fig. 4e). In the parental females, however,
(Fig. 4a and c) they are arranged as islets localized at the base of
the follicles situated in an ovarian cortex which is thicker than that
of the hybrid.
The subepithelial cells with 17S-HSDH activity are more abundant in

the female of C. moschata than in the Peking female (Fig. 4b and d)
In the hybrid they are arranged as a fragmented band (Fig. 4f).
In the adult ducks this distribution of the two steroidogenic cell

types is further .developed. During the egg-laying period the ovary of
the hybrids is smaller than that of the parents, the medullary zone is
strongly developed and the follicles are less numerous. As in the ova-
ries of the females of both parental species there are "clear cells"
with 3B-HSDH activity in the theca where they form a fairly thick en-
velope. These cells also occur in the cortical zones which are without
follicles. The histochemical 17S-HSDH reaction is also very strong in
the female hybrid in the regions of contact between the germinal epi-
thelium and the follicles and is even much more evident than in the
female of C. moschata whose reaction is more distinct than in the female
Peking (Deray, 1973b). But after the 15th month, when egg-laying has
definitely come to an end, the follicles can no longer be seen but the
cells with 3B-HSDH activity persist and form increasingly important
nodules.
362
large tubules (cell. dual.)
dorsal muscle _*~~
atresia
nodule of
.'"'t-~t--- clear cells
Islets of
clear cells
vasc o
Fig. 3. Schematic drawing of the progressive hypertrophy of the right

gonad of the female hybrid (if A. pZatyrhynchos x <jl C. moschata) under nat-
ural conditions. Clear cells which seem to derive from mesonephric
tubules are forming "caps" which associate with the theca of the fol-
licles (1), are forming nodules (2) or islets (3) of clear cells. In
certain cases testicular tubules differentiate which are sterile or
contain germinal cells which develop until the onset of meiosis. After-
wards they become necrotic as in male hybrids
The study of the ovary of the hybrid during ontogenesis thus enables
us to distinguish several stages in its development towards sexual
inversion. The first event is observed on the 21st day when a large
number of germinal cells die at the moment they enter precocious mei-
osis. It might be suggested that this necrosis is the consequence of
a failure in the reduction process comparable to the situation which
is encountered in male cells. But it is probable that the endocrine
factors which are due to the particular ratio between 3B-HSDH and 17B-
363
o.
b c
-------*"
c
.. .....
1
...... '.
M
~~--~~------!--------~
Fig. 4. Distribution of the enzymes ~5-3 S -HSDH and ~ 5-17S-HSDH in the
ovary of female A. platyrhynchos , C. moschata and hybrids, 2 months olCl
(C corte x, M medulla, 0 oocytes). (a and b) Ovary of A. pla tyr hyncho s;
3S-HSDH activity (a) is localized at the base of th e follicles whereas
17S-HSDH activity (b) is seen i n the small subepithelial cells (x 80) .
(c and d) Ovary of C. moschata ; the intensity of the reactions for 3S-
HSDH (c) and 17S-HSDH (d) is reciprocal to the situation in the ovary
of the female Peking duck (x 80). (e and f) Localization of the same
enzymes in the female hybrid ( a" A . platyrhynchos x 'i' C. moschata) : 3S -HSDH
(e) and 17S-HSDH (f). The cortex is less thick than in the females of
the parental species (x 80)
364
HSDH cells have some influence on this phenomenon. We assume that the
hormonal balance in the o v ary of the juvenile and adult hybrid duck,
which is different from that of the parental species, is responsible
for its subsequent development and determines the differentiation of
the germinal cells in oocytes and later in spermatocytes which, how-
ever, do not exceed the stage which has been observed in the testes
of hybrid males (Fib. 2b). The progressive hormonal imbalance which
is the result of the activity of a hybrid genome finally leads to sex-
ual inversion which ends by a total des integration of the gonad.
B. Experimental Interventions
I. Effects of Estrogenic Hormone

(in Collaboration with A. Deray and Y. Lutz-Ostertag)
Ridgon (1968) has suggested that one of the causes for the sterility
of female hybrids (if C. moschata x 'i' A. p l atyrhynchos) might be the de-
ficiency or absence of secretion of estrogenic hormone. We carried
out experiments by moistening incubated eggs following the technique
of Seltzer and Wineland (1956), in order to study the influence of
diethylstilbestrol on the differentiation of female gonads of hybrids
(Deray et al., 1970; Deray, 1970). For comparative purposes female
embryos of Anas and Cairina were also treated. In these genera diethyl-
stilbestrol does not change the differentiation of the left ovary, but
it has a different effect on the right gonad and the right oviduct de-
pending on the genus. In the female of A. platyrhynchos (P e king race )
the hormonal treatment does not affect the regression of the right
gonad but the posterior half of the oviduct persists (Fig. Sa) whereas
in the female of C. m oschata the right oviduct regresses nearly completely
while the right gonad persists as a small ovary (Fig. 5b).
In the f e male hybrids the whole genital apparatus reacts to the hor-
monal treatment. In both cross-bre edings the right oviduct is mai nt ained
and its anterior end is closed. The right gonad also persists more fre-
quently in the treated hybrids. In the hybrid Anas if x Cairina 'i' in
which the right gonad appears spontaneously during the third year, the
female hormone in 3 out of 6 cases has induced the differentiation of
( a) (b) ( c) (d)
Fig. 5 a-d. Influenc e of diethylstilbestrol on the differentia t ion of

the g e nital apparatus of A. platyrhynchos ( a), C. moschata (b), hybrid
A. p l a tyrhynchos if x C. moschata 'i' (c) and hybrid C. moschata if x A. platy-
rhynch os 'i' (d) (x 1/3)
365
an ovary in which the oocytes enter vitellogenesis but remain smaller

than those of the left ovary (Fig. 5c). In the reciprocal crossing
( Cairina d' x Anas ~) the right gonad is also maintained by the treatment
(Fig. 5d) but both gonads consist of a medullary stroma without gono-
cytes.
The treatment of duck embryos with diethylstilbestrol demonstrates

that the regression of the right gonad is genetically fixed in a strict
way in the female Peking duck whereas it can be influenced by the hor-
mone in the female of C. mosehata in which a right ovary sometimes per-
sists spontaneously (4 out of 24 cases, Table 1) (Gomot et al., 1973).
In the female hybrids the hormone causes a dominance of the ovarian
character of the female whose gonads are affected to a lesser degree
by the hybridization. But in the sterile hybrid (d' Cairina x ~ Anas)
it favors a pathological cellular proliferation of the medulla instead
of a promotion of ovarian differentiation.
II. Influence of Post-embryonic Removal of the Left Ovary

(in Collaboration with A. Ueray)
Removal of tpe left ovary in young hens elicits the development and
hypertrophy of the rudimentary right gonad (Goodale, 1916; Benoit,
1923, 1932; Zawadowsky, 1926; Domm, 1927, 1939). Generally this right
gonad consists of medullary cords and sterile testicular tubules, but
when an ovariectomy is carried out shortly after hatching, the right
gonad may show complete spermatogenesis (Benoit, 1923; Miller, 1938;
Domm, 1939). In certain cases Benoit (1950) observed the formation of
ovotestes. In hens the ovariectomy causes the development of the male
(or neuter) plumage but later on the female plumage reappears. Ablation
of the left ovary in the turkey and the pheasant does not provoke a
development of the right gonad. This gonad rarely hypertrophies in the
duck (Domm, 1939). In the quail, Kannankeril and Domm (1968) have ob-
served the hypertrophy of the rudimentary right gonad in 80% of the
birds which have been operated on between the 2nd and 68th days. The
sterile gonad consists of epithelial cells, of "fat laden cells" and
of connective tissue.
We have carried out corresponding experiments. in female hybrids in

order to observe the behavior of the right rudiment which normally
starts to develop during aging. The same operations have been done in
the females of the parental species. The results of the ablation of
the left ovary in ducks between 2 days and 6 months old are presented
in Table 1.
There is a fundamental difference between the results in both parental

females. Ablation of the left ovary in the Peking female ducklings
between 2 and 15 days old is never followed by hypertrophy of the
right rudiment (Fig. 6b). In the female of C. mosehata removal of the
left ovary always causes an onset of the development of the right go-
nad (Fig. 6a). The competence of the right gonad to form a fertile
structure decreases with age, but the influence of this factor seems
to be much less important than in hens. In 13 ovariectomized animals
between days 2 - 20, the right gonads had the following structure:
11 had normal or atretic follicles (Fig. 7a and b), 1 presented cells
in meiotic prophase of spermatogenesis (Fig. 7c) and 1 consisted only
of "clear cells".
In the C. mosehata females that have been castrated between the 40th
day and the 6th month, the right gonads in most cases still have ger-
minal cells (oocytes 1/5, spermatocytes 3/5 (Fig. 7d), "clear cells"
366
1/5). Finally, in adult ducks which have been castrated after 20 months,
the right rudiment does not develop and maintains the appearance of a
small white filament.
Whereas the kind of gamete differentiation is easily recognizable histo-

logically (male meiosis and ovarian follicles) in the righthand gonads,
this is not true with regard to the endocrine function which neverthe-
less is evident in certain cases by a remarkable stimulation of the
left oviduct. The histochemical study shows that 3S-HSDH and 17S-HSDH
are active in the "clear cells" and the subepithelial cells of the
right gonads which have the appearance of an ovary (Fig. 8 a-c). Gener-
ally the 3S-HSDH cells are more numerous and it may be that 17S-HSDH
activity cannot be detected. The interpretation of the role of cells
with 3S-HSDH activity is difficult since we know that it is present in
the "clear cells" of the theca of the ovary as well as in the inter-
stitial cells of the testis. The histochemical reaction does not indi-
cate the kind of hormone which is secreted.
The comparative study of the regression of the right gonad during the
embryonic development and its reappearance after ovariectomy provides
some hints. It must be emphas~zed in particular that contrary to the
conditions in the hen, it is the female component which appears most
frequently. ,The differentiation of male cells is never much advanced
and remains restricted to the posterior part. Furthermore, when the
regression of the right gonad in the embryo is observed, it becomes
apparent that the cortex persists frequently in the anterior region
and regresses completely posteriorly (Fig. 9). At the moment of the
awakening, ovarian development takes place in the anterior part depend-
ing on the amount of cortical "remnants". When these are either much
reduced or nonexistent, the newly developing organ is very small and
consists exclusively of mesonephric tubules and a few "clear cells"
("interstitial medullary cells" or "fat laden cells" or "luteal cells")
When the hypertrophy of the right rudiment is more pronounced, the
"clear cells" are more numerous and their close association with the
mesonephric tubules suggests a progressive transformation of the tu-
bular cells to clear cells. The drawing (Fig. 9) shows the probable
evolution of the rudiment of the right gonad as function of an antero-
posterior gradient with cortical dominance.
Considering the female hybrid (if A. platyrhyncho$ x ~ C. mO$chata) one

observes that removal of the left ovary always elicits a development
of the rudiment of the right gonad but often this is of only a small
Fig. 6 a-d. Effects of ova~iectomy after hatching on the genital ap- ~

paratus. (a) C. mO$chata ovariectomized at 4 days and sacrificed when
16 months old. The well developed right ovary (oc) has stimulated the
oviduct (ovd) (x 1). (b) A. platyrhyncho$ ovariectomized at day 2 and
sacrificed when 9 months old. No hypertrophy of the right rudiment (rr).
On the vena cava posterior only a small thickening can be observed
which corresponds to a remain of the mesonephros (x 1). (c) Female
hybrid (A. platyrhyncho$ if x C. mO$chata ~ ) ovariectomized at 20 days,
sacrificed 13 months old. The right gonad is considerably hypertrophied.
It consists of large clear cells without germ cells and has a stimulat-
ing action on the left oviduct (x 1). (d) Female hybrid ( if A.platy-
rhyncho$ x ~ C. mO$chata) ovariectomized at day 2 and sacrified when 12
months old. The right gonad (rg) is slightly developed and is formed
by a nodule of clear cells with 3S-HSDH activity and by some atretic
follicles (x 1).
367
Fig. 6 a - d
368
Fig. 7 a-d. Histological aspect of the right gonads of C. mosehata after

ablation of the left ovary. (a) Presence of numerous oocytes (oe) in
the hypertrophied right ovary of a female which was operated on day 4
and sacrificed after 7 months (x 10). (b) Oocyte of the same ovary
(Fig. a). The granulosa (g) and the theca which consists of islets of
clear cells (eel) are visible (x 100). (c) Figures of male meiosis
(se) amidst clear cells in the right gonad of a female which was ovari-
ectomized at day 2 and sacrified after 13 months (x 400). (d) Testic-
ular tubules (tt) at the base of the globular nodule which differen-
tiated instead of the right gonad in a female which was ovariectomized
at the age of 40 days and sacrificed when 27 months old (x 400)
size (Fig. 6c and d). Germ cells have been found there in only 3 out
of 14 cases (Table 1). This situation is certainly the consequence of
the marked necrosis which affects the germinal cells in the 2 embryonic
gonads on the 19th day of incubation. In the right gonads which have
been obtained in this way, 2 cases of female atretic development and
1 case with male development with stoppage of the meiosis as in the
testes of the male have been observed. The rest of the gonad consists
of connective tissue and mainly of very abundant clear cells. Some of
these cells become necrotic but the majority shows 3B-HSDH activity
which probably is responsible for the secretion of estrogens which
stimulate the oviduct. The origin of these cells seems to be comparable
to those of the right gonads of females of C. mosehata (Fig. 9) and to
those which appear under normal conditions (Fig. 3). After removal of
the left gonad of the female hybrid (d' C. mosehata x 'i' A. platyrhynchos )
one observes on the right side, the presence of a nodule whose struc-
ture is analogous to the gonad which persists sometimes under natural
conditions (Fig. 1).
369
1713+
Fig. 8 a-c. Histochemical localization of 3S-HSDH and 17S-HSDH activity

in the right gonad of a female of C. moschata which was ovariectomized
at day 3 and sacrificed after 13 months. The clear cells of the theca
(b) are numerous and show a positive reaction for 3S-HSDH (a) whereas
17S-HSDH activity is localized in the subepithelial cells (c). (x 250)
C. Conclusion
The hybridization between two species of ducks of different genera

( Anas platyrhynchos and Cairina moschata) induces in females an intersex-
uality which expresses itself either by the rapid disappearance of
germ cells and of the cortex in the females (C. moschata if x A. platy-
rhynchos ~ ) or by becoming progressively manifest during the aging
(A. platyrhynchos if x C. moschata ~ ). Effects which are comparable to the
second form have been described in hybrid doves (Riddle and Johnson,
1939) or pheasants (Bissonnette, 1942). Until now no explanation other
than that of an aberrant chromosomic constitution or a diminution of
the activity of female determining genes as proposed by the fore-
mentioned authors, seemed to be reasonable. The differences which have
been described between the chromosomes of the hybrids and the parental
species (Yamashina, 1941; Mott et al., 1968; Stasko and Chalupa, 1972)
or between the contents of nuclear DNA of erythrocytes of these animals
(Gerzeli, 1956; Chilingarjan and Pavlov, 1961) provide some indications
for the causes of the failure of meiosis in males but they do not ex-
plain the endocrinological changes.
When we consider the results of histological and histochemical obser-
vations we are struck by the strong development of the medullary com-
ponent of the ovary of the hybrids and by the distribution of 3S-HSDH
and 17S-HSDH activity. Unfortunately the presence of these enzymes
does not provide information about the nature of the secreted hormones ..
Some pilot studies on the uptake of labelled progesterone and pregne-
nolone by ovarian fragments of adults show us, however, that the per-
centage of uptake corresponds remarkably well to the extent of 3S-HSDH
reaction to the histochemical preparations. Therefore we may be able
370
in the embryo: cortex present cortex absent
caudal -"
part U
Fig. 9. Schematic drawing.of the hypertrophy of the right gonad of the

female of C. moschata in relation to its embryonic development. In the
cranial part of the gonad normal (1) and atretic follicles (2) can be
seen. In this region clear cells (3) remain at the base of the folli-
cles. In the middle region the cortex contains subepithelial cells
with 17S-HSDH activity (4), thecal cells (5) in a rosette-like grouping
and spermatogenetic figures (6). In the posterior region there are
clear cells with 3S-HSDH activity (7) and degenerating clear cells.
In the zone which is in contact with the mesonephros (8) one observes
structures which suggest that the clear cells derive from tubules of
the mesonephric region. In the parts in which the new development is
less marked there are small islets of clear cells (9) between the meso-
nephric tubules whereas in the cases of strong hypertrophy large nod-
ules (10) of active clear cells can be seen
to say that the secretion of sex hormones which arise from these pre-
cursors is different in the parental species - higher uptake in the
ovary of the female Peking duck - and that it is distinctly lower in
the female hybrid. Since hybridization diminishes the amount of hor-
monal secretion by the ovary we have subjected embryos to a treatment
with estrogenic hormone and the results show that this procedure fa-
vors the differentiation of the ovarian cortex of the right gonad of
C. moschata and the hybrid, whereas it remains wi thou t effect on the
regression of the right gonad of Peking duck. The regression of the
right gonad of the female Anas seems to be under strong genetical con-
trol from the beginning of embryonic development. For Wolff and Haffen
(1952a,b) have shown that the right gonadal (Anlage) of Anas continues
to regress during culture in vitro whereas the corresponding gonad of
embryos of fowl develops to an ovotestis. These results also explain
the fate of the right rudiment of the Peking duck following ovariectomy
after hatching. The ovariectomy in Peking ducks elicits the differen-
tiation of tail-feathers which are characteristic for the male (neutral)
sex. This phenomenon has been observed earlier by Goodale (1910 - 1911)
in the "Rouen"-duck simultaneously with the changement of coloration.
371
The observation on a rebeginning of ovarian development in the female

duck of C. moschata indicates that the right gonad of birds has not
necessarily a testicular potency as experiments by ablation of the
left ovary in the hen might suggest (Benoit, 1932 - 1950). In the fe-
male of C. moschata the inverse situation develops because a part of
the cortex of the embryonic ovary does not regress totally. Comparable
resul ts have been obtained in Streptope lia risoria by Cheng (1 973). I'iore-
over, we observed that in the female of C. moschata, a longitudinal gra-
dient with cortical dominance exists; the cortex regresses mainly in
the posterior region and cortical islets of the anterior zone become
an ovary either spontaneously or more frequently after ovariectomy on
the left side. In these cases vitellogenesis never proceeds as far as
in the left ovary (the follicles never exceed a diameter of 1 cm).
We assume that this must be attributed to the persistence of a medul-
lary "nucleus", but we do not yet have the proof for this.
Left ovariectomy in C. moschata causes a retardation in the differentia-

tion of the red tubercles which differentiate on the" bill and which
surround the eye.
The right gonad of the female hybrid reacts to ovariectomy but its
hypertrophy is weak. It seems that its development is genetically pro-
grammed at the same time as the transformation of the left gonad. It
is therefore concluded that the phenotypic expression of the hybrid
genome of the female changes gradually in time and that after a domi-
nance of female genes of the Peking type during the embryonic life,
one observes a progressive expression of regulatory genes of the ma-
ternal type which permit the differentiation of the right rudiment in
the hybrid (if A. platyrhynchos x ~ C. moschata). As far as the inverse
cross breeding is concerned (if C. moschata x ~ A. platyrhynchos) it is
probable that structural genes correlated with cortex development re-
main inactive or are strongly repressed after the end of the embryonic
life. Ovariectomy in female hybrids has no effect on the external fea-
tures of the animals.
References
Benoit, J.: A propos du changement experimental de sexe par ovariec-

tomie chez la poule. C.R. Soc. Biol. 89, 1326-1328 (1923).
Benoit, J.: Etude physiologique, histologique et histophysiologique
de l'inversion sexuelle de la poule determinee par l'ablation de
l'ovaire gauche. Arch. Zool. EXp. Gen. 11, 1-112 (1932).
Benoit, J.: Reproduction, caracteres sexuels et hormones. Determinisme
du cycle sexuel saisonnier. In: Traite de Zoologie (ed. P.P. Grasse),
T. XV, Oiseaux, pp. 384-478. Paris: Masson 1950a.
Benoit, J.: Differenciation sexuelle chez les Oiseaux au cours du de-
veloppement normal et de l'intersexualite experimentale par ovari-
ectomie. Arch. Anat. micro morpho expo 12, 395-414 (1950b).
Bissonnette, T.R.: The "mule" pheasant. Proc. Amer. Phil. Soc. 85,
49-70 (1942).
Boss, W.R., Witschi, E.: The permanent effects of early stilbestrol
injections on the sex organs of the Herring gull Larus argentatus.
J. Exp. Zool. 105, 61-77 (1947).
Cheng, M.F.: Effect of ovariectomy on the reproductive behavior of
female ring doves (Streptopelia risoria). J. Comp. Physiol. Psychol.
U.S.A. 83, 221-233 (1973).
Chilingarjan, A.A., Pavlov, E.F.: Dosage du taux moyen d'acide desoxy-
ribonucleique dans les noyaux des erythrocytes de Canards musques
et Pekin, ainsi que de leurs hybrides, et de 2 sous-especes de
Lacerta saxicola, ainsi que de leurs hybrides; resultats numeriques
obtenus. Akad. Nauk. arm. SSR, Dokl. 11, 55-60 (1961) (in Russian).
372
Crew, F.A.E., Koller, P.C.: Genetical and cytological studies of the

intergeneric hybrid of Cairina moschata and Anas platyrhyncha. Proc.
Roy. Soc. Edinb. 56, 210-241 (1935/1936).
Deray, A.: Etude deslmodifications progressives affectant les gonades
des feme lies hybrides du croisement: Cairina moschata 'i' x Anas platyrhyn-
chos a". These 3e cycle, Fac. Sc. Besan<;:on, no d' ordre: 11-3 (1970).
Deray, A.: Localisation et appreciation quantitative des cellules a
reaction positive pour l'enzyme A5-3B-HSDH au niveau de l'ovaire
gauche de Cane de Barbarie adul te (Cairina moschata L.). Etude com-
paree avec l'ovaire de Cane Pekin (Anas platyrhynchos). C.R. Acad.
Sci. Par i s 276, 1 863 -1 866 ( 1 973 a) .
Deray, A.: Localisation et extension prise par les sites d'activite
des enzymes A5-3S-HSDH et A5-17S-HSDH dans l'ovaire de femelle
hybride adulte (Pekin d' x Barbarie 'i') avant sa transformation.
C.R. Acad. Sci. Paris 277, 1161-1164 (1973b).
Deray, A., Gomot, L.: Sur un stade ovotestis des gonades droite et
gauche de Cane hybride (Cairina moschata 'i' x Anas platyrhynchos a" )
en cours d'inversion sexuelle. C.R. Acad. Sci. Paris 268, 2485-2488
(1969). -
Deray, A., Gomot, L.: Etude comparee des rapports entre les quantites
de cellules a activite 3S-HSDH et 17S-HSDH dans les ovaires de Cane
Pekin (Anas platyrhynchos) et Barbarie (Cairina moschata). Journees de
recherches avicoles et cunicoles, 12, 13, 14 decembre 1973. Paris.
Deray, A., Gomot, L.: Etude comparee de l'appareil genital des Canards
Pekin, Barbarie et hybrides. Activites et travaux du Laboratoire
de Zoologie et Embryologie de la Faculte des Sciences de Besan<;:on,
1974.
Deray, A., Gomot, L., Lutz-Ostertag, Y.: Modifications de la diffe-
renciation de l'appareil genital de canards femelles des genres
Cairina, Anas et de leurs hybrides sous l' effet d' une hormone femelle.
C.R. Soc. Biol. 164, 303-307 (1970).
Domm, L.V.: New experiments on ovariotomy and the problem of sex in-
version in the fowl. J. Exp. Zool. ~, 31-173 (1927).
Oomm, L.V.: Modifications of sex and secondary sexual characters in
birds. In: Sex and Internal Secretions (ed. E. Allen), pp. 227-327.
Baltimore, Md.: Williams and Wilkins 1939. '
Faber, H.V.: tiber den extremen Geschlechtsdimorphismus im Wachstum der
Moschusente, Cairina moschata Flemm. III. Die Bastarde von Moschus-
erpel x Hausente. Acta endocrinol. l&., 135-141 (1957)·.
Gerzeli, G.: 11 tenore di acido desossiribonucleico negli eritrociti
di Anas plaryrhynchos, Cairina moschata e dei lora ibridi diretti e '
reciproci. Experientia 12 (7), 265-267 (1956).
Gomot, L.: La transplantation des gonades dans la recherche du deter-
minisme de la sterilite des hybrides intergenetiques de Canards.
Ann. Bio. 9, 7-8, 9-10, 269-277 (1970).
Gomot, L., Ardiet, P.: L'hybridation Canard Pekin x Cane de Barbarie.
Etude comparee des canards femelles Pekin, Barbarie et hybride.
An. Sci. Univ. Besancon 2, 3-25 (1966).
Gomot, L., Ardiet, P., Lutz=Ostertag, Y.: Intersexualite gonadique des
hybrides femelles provenant du croisement Canard Pekin x Cane de
Barbarie. C.R. Soc. Biol. 160, 1572-1573 (1966).
Gomot, L., Oeray, A., Marchan~C.R.: Nouveaux cas d'exceptions a la
regle generale de la subsistance du seul ovaire gauche chez les
femelles d'oiseaux adultes. (Observations chez les canes de Barbarie
et les hybrides du croisement Pekin male x Barbarie femelle). Bull.
Ass. Anat. 22, 333-340 (1973).
Goodale, H.O.: Some results of castration in ducks. Bioi. Bull. 20,
35- 66 ( 1 91 0-1 911) . -
Goodale, H.O.: Further developments in ovariectomized fowl. Bioi. Bull.
30, 286-293 (1916).
373
Haffen, K.: Intersexuali te chez la Caille (CotuPnix coturnix). Obtention

d'un cas de ponte ovulaire par un male genetique. C.R. Acad. Sci.
ill, 3876-3879 (1965).
Haffen, K.: Quelques aspects de l'intersexualite experimentale chez
le Poulet (Gallus gallus) et la Caille (CotuPnix cotuPnix japonica). Bull.
Biol. 103, 401-417 (1969).
Jaap, R.G-.-,-Ingram, R.S., Godfrey, E.M.: Ninth World's Poultry Congress
2, 30- 35 (1 951) .
Kannankeril, J.V., Dornrn, L.V.: The influence of gonadectomy on sexual
characters in the Japanese Q1!ail. J. Morph. 126, 395-412 (1968).
Lavedan, J., Demay, M.: Etude sur la sterilite des hybrides. Arch.
Zool. EXp. Gen. 88, 85-95 (1 Q 51).
Lutz-Ostertag, Y.: Le croisement Canard Khaki-Campbell (An as platyrhynchos
L.) x Cane Barbarie CCairina moschata L.). Arch. Anat. Histol. Embryol.
48, 300-327 (1965).
Lutz-Ostertag, Y., David, D.: Sur l'activite des testicules de l'em-
bryon de Canard hybride (Cairina moschata male x Anas platyrhynchos
femelle). C.R. Soc. Biol. 155, 2330-2331 (1961).
Lutz-Ostertag, Y., Gomot, L.: Sur l'hybride femelle issue du croise-
ment Cairina moschata L. (male) et Anas platyrhynchos L. (femelle).
C.R. Soc. Biol. 154, 2282-2283 (1960).
Marchand, C.R.: Recherches histophysiologiques sur l'adenohypophyse
et les testicules des Canards hybrides steriles issus du croisement
male Pekin x femelle Barbarie: 1. Comparaison du cycle testiculaire
des hybrides avec celui des genres parentaux. 2. Caracterisation
et variations des cellules gonadotropes et thyreotropes chez les
males des trois types. These 3e cycle, Fac. Sc. Besan90n 147 (1972).
Marchand, C.R., Bugnon, C.: La reponse "thyreoprive" de l'adenohypophyse
des Canards males et femelles Barbarie, Pekin et hybrides (du croise-
ment Canard Pekin x Cane de Barbarie) comparee aux effets de la
castration. Bull. Ass. Anat. 57, 157-164 (1973).
Marchand, C.R., Bugnon, C., Gomo~ L.: Reponses cytophysiologiques de
l'adenohypophyse a la castration des Canards males Pekin ou Barbarie
et hybrides des deux genres. Ann. Sci. Univ. Besan90n, med. 6, 1-6
(1969). -
Marchand, C_R., Bugnon, C., Gomot, L.: Les effets de la thyroxine sur
l'adenohypophyse et la thyroide des canards males Pekin, Barbarie
et hybrides (du croisement male Pekin x femelle Barbarie), castres
ou non. C.R. Acad. Sci. Paris 274, 2204-2207 (1972).
Marchand, C.R., Gomot, L.: Variations annuelles de l'activite testi-
culaire du Canard hybride du croisernent Pekin male (Anas platyrhynchos
L.) x femelle Barbarie (Cairina moschata L.). C.R. Soc. Biol. 166,
556-563 (1972). -
Marchand, C.R., Gomot, L.: Ultrastructure des cellules de Leydig des
testicules de Canard doe Barbarie (Cairina moschata L.) et du Canard
hybride (Anas platyrhynchos L. male x Cairina moschata L. femelle) pen-
dant la phase d'activite sexuelle. Gen. Compo Endocrinol. 12, 395
(1973) .
Miller, R.A.: Spermatogenesis in a sex-reversed female and in normal
males of the domestic fowl, GaUusdomesticus. Anat. Rec. 70, 155-189
(1938) .
Mott, C.L., Lockhart, L.H., Rigdon, R.H.: Chromosomes of the sterile
hybrid duck. Cytogenetics 7, 403-412 (1968).
Narbaitz, R., Teitelman, G.: A histochemical study of sex inversion
produced by estradiol in chick embryos. J.Embryol. Exp. Morphol.
G.B. 13, 46-50 (1965).
Pavlov, ~F., Tchilingarian, A.A., Sarkissov, R.N., Nartchian, L.P.:
La transplantation de testicules. Une methode pour obtenir une
spermatogenese complete chez les hybrides interspecifiques steriles
de Canards. Genetika, SSSR J...' 61-66 (1966) (in Russian) .
374
Popescu, C.: Particularitati biologice la hibrizii ratei lesesti cu

rata comuna si Peking. Analele stiintifice ale Universitatii "Al.
I. Cura" Din IASI (Serie noua), Sect. II, VI, no. 3, 865-871 (1960).
Riddle, 0., Dunham, H.H., Schooley, J.P.: Genetic hermaphroditism in
a strain of pigeons. Repr. Genetics 27, 165 (1942).
Riddle, 0., Johnson, M.W.: An undescribed type of partial sex-reversal
in dove hybrids from a sub-family cross. Anat. Rec. 75, 509-527
(1939). -
Rigdon, R.H.: Gonads in hermaphroditic ducks: a pathologic study.
Am. 1. Veter. Res. 28, 1125-1134 (1967).
Rigdon, R.H.: Effect of hormones on gonads of sterile hybrid ducks.
Texas Rep. on Biol. and Med. 26, 581-594 (1968).
Seltzer, W., Wineland, N.J.: United States Patent Office ~, 734, 482
(1956).
Stasko, J., Chalupa, I.: Karyotype de Anas platyrhynchos et de Cairina
moschata et de leurs hybrides. Acta Fac. Rerum nat. Univ. Comenianae,
Genet. Tchecosl. 3, 97-105 (1972) (in Czech.).
Taber, E.: Intersexuality in birds. In: Intersexuality in Vertebrates
Van Tienhoven, A.: A method of "controlling sex" by dipping of eggs
in hormone solutions. Poultry Sci. 36, 628-632 (1957).
Wolff, E.: <L'evolution, apres l'eclosion, des Poulets males transformes
en intersexues par l'hormone femelle injectee aux jeunes embryons.
Arch. Anat. Hist. Embryol. 23, 1-38 (1936).
Wolff, E., Haffen, K.: Sur le developpement et la differenciation
sexuelle des gonades embryonnaires d'Oiseau en culture in vitro.
J. Exp. Zool. 119, 381-399 (1952a).
Wolff, E., Haffe~K.: Sur la differenciation sexuelle des gonades
embryonnaires de l' embryon de Poulet en culture in vitro. Ann. Endocr.
13,724-731 (1952b).
Wolff, E., Strudel, G., Wolff, E.: L'action des hormones androgenes
sur la differenciation sexuelle des embryons de Poulet. Arch. Anat.
Hist. Embryol. 31, 237-310 (1948).
Yamashina, M.Y.: Studies on sterility in hybrid birds. III. Cytologica
investigations of the intergeneric hybrids of the muscovy duck if
(Cairina moschata) and the domestic duck'i' (Anas platyrhyncha var. domestica).
Jap. Jour. Genet. 17, 207-228 (1941).
Yamashina, M.Y.: A revised study of the chromosomes of the muscovy
duck, the domestic duck and their hybrid. Cytologia 12, 163-169 (1942).
Zawadowsky, M.: Bisexual nature of the hen and experimental hermaphro-
ditism. Trans. Lab. EXp. Biol. Zoopark. MOscow~, 164-179 (1926).
Some Effects of Embryonic Gonad and Non-Gonadal Grafts
on the Development of Primary Sexual Characteristics in the
Chick
J. Thiebold
A. Introduction
The starting point of this research has been the accomplishment of a

series of coelomic grafts on the chick embryo, in order to study the
sensitivity of the Mullerian ducts to the testis during their edifica-
tion phase. The experiments performed have yielded results which relate
to different problems posed by the genetical organogenesis in the chick.
I shall successively describe here acquisitions in three fields: (1)

The growth of the embryonic gonads; (2) the edification of the l"1ulle-
rian ducts during the "sexually indifferent phase"; and (3) the asym-
metry of the Mullerian duct system in the female embryo.
B. Growth of Embryonic Gonads
Initial experiments by Wolff (1946) have clearly demonstrated the di-

rect action exerted by grafted embryonic gonads on the sexual differ-
entiation in the chick. The testicular grafts suppressed the Mulle-
rian ducts by destruction but did not modify gonadal differentiation
in a female host, while the ovarian grafts had a discreet feminizing
effect on the left gonad and on the Mullerian tract in a male host.
Similar methods of grafting gonads, applied to amphibians, have almost

always revealed inhibitory effects on the gonadal development in the
host (Mikamo and Witschi, 1963).
Experiments have been carried out to see if bird embryonic gonads cause
such an inhibition when grafted onto aohost embryo, prior to its sexual
differentiation (Thiebold and Reichhart, 1974). Nine day gonadal frag-
ments (right and left ovaries, testes) have been implanted into the
coelomic cavity of 50 hrs host embryos. Non-grafted embryos, or embryos
with a neutral tissue (4 day frontal buds), served as controls. Host
embryos were killed when 13 days old and their fixed gonads weighed,
or histologically treated, for a "paper weight" estimation of their
mass. The significance of the differences in weight between homologous
organs of experimental and of control subjects was estimated with the
Student's "t" test.
The following results were recorded:
1. No modification was induced in the host's gonadal development by a

neutral tissue, no matter what the volume of the graft was at the time
of the autopsy.
2. The three types of gonad grafts tested, inhibited the female host's
gonadal development (Figs. 1 and 2). Comparison of 13 day host gonads
376
with control gonads, showed a highly significant (p < 0.01) weight

difference. The degree of this inhibition seems to be independent of
the type (testis, left or right ovary) of the graft, but directly
linked to its volume.
3. The right female gonad is more sensitive to the graft's inhibitory

action than the left one.
4. The histological organization of the inhibited gonads did not differ

from that of the controls. Particularly, the medullo-cortical balance
was not modified.
5. No significant gonad weight modification was noted in male hosts

after ovary or testis implantation.
Fig. 1. Normal aspect of gonads in a control female 13 day embryo

Fig. 2. Reduced (inhibited) gonads in a female 13 day embryo, with a
testicular fragment grafted when 2 days old. Note the presence of the
grafts (arrow)
These results are not easily explainable, even though they are similar
to results from comparable experiments performed in other vertebrates.
Obviously, the inhibitory action is tissue-specific. Moreover, the
extent of the inhibition is seemingly in proportion to the volume of
the implanted foreign gonadal tissue. Thus, these phenomena would be
due to a controlled growth mechanism. This is also the interpretation
proposed by Mikarno and Witschi (1963), after grafting experiments on
Xenopus. It must, however, be stressed that in these latter experiments,
testes grafts (which were alone tested) impaired gonad development of
the male as well as of the female host . In our experiments, the absence
of sensitivity of the testes is another point for discussion, which
justifies further investigations.
377
c. The Edification of the Mullerian Ducts during the "Sexually Indif-

ferent Phase"
Studies on the cartilaginous Scyliorhinus caniculus (Thiebold, 1964),

have shown that in this species (unlike in birds) Mullerian development
is incomplete in male embryos, with destruction occurring precociously,
while the ducts are still far from cloacal connection. I have hypothet-
ically linked this distal agenesis to a precocious functional state
of the Scyliorhinus embryonic testis, which should also have, in addition
to its well known destructive action, an inhibitory effect on the
Mullerian development. I believed that one approach to this problem
would be to test this possible inhibitory action of the testis in the
chick embryo, by means of the embryonic graft method. I have effec-
tively established that early introduction of a testis fragment into
the coelomic cavity of the 2 day old chick embryo, results in a signif-
icant slowing down of the Mullerian ducts development during the
"sexually indifferent" phase (i.e. before the 9th day in the chick
embryo). Therefore (as a rule), in the operated embryos, the latter
have not reached the cloacal membrane as the regression, i.e. destruc-
tion, is beginning (Thiebold, 1962a).
This hypothesis, of an inhibitory action from the embryonic testis on

the progression in the caudal direction of the Hullerian ducts, has
since been supported by a comparative study of the Hullerian develop-
ment in several reptilian species (Raynaud et al., 1969).
This type of experiment allows other conclusions to be made about the

chronology of the Mullerian duct's sensitivity to testicular actions.
If, as shown by ~he present results, the Mullerian ducts during their
edification phase respond to the action of the grafted testis by a
slowing down of their growth, they do not display any sensitivity to
its destructive action. Moreover, in none of the precocious testicular
implantation experiments have we observed even the slightest anticipa-
tion of the well-known necrosis phenomenon, with regard to male control
embryos. The Mullerian cells have, therefore, to attain a sufficient
level of ageing or differentiation before being competent, i.e. respond-
ing by death to the embryonic testis action. This state of competence
is seemingly transitory, since 9 day female ducts, grafted simulta-
neously and side by side with a testicular fragment, fail to degenerate
(Thiebold, 1964).
D. The Origin of the Asymmetry of the Mullerian Duct System in the

Female Embryo
Many researchers who have been concerned with this problem (principally
Wolff, 1953, 1959; Groenendijk-Huijbers, 1962; Hamilton, 1963) have
interpreted the "regression" of the right Hullerian duct from the 11th
day of incubation onward, to be a result of the intervention of a hor-
monal substance of gonadal origin. According to Hamilton (1963) "there
is some evidence that the right duct of the female has a lower thresh-
old for response to the endogenous hormone (presumably the masculin-
izing hormone) than does the left duct".
Several observations and experimental results have led me to an entirely

different interpretation: the Mullerian asymmetry has its origin in
a fall of the growth rate of the right Mullerian duct relative to the
left one. This differential growth might be explained by the different
mechanical state of stretching, in which the two ducts are placed from
378
the eleventh day onwards. While the right duct becomes slack, in con-
sequence of the rupture of its ostial junction to the pleuro-peritoneal
septum, the left one (the anterior anchorage of which is firmer) remains
longitudinally stretched throughout the body cavity and lengthens ap-
proximately in proportion with the general growth of the trunk (Thie-
bold, 1973).
This observation suggests that the cellular proliferation in the Mulle-

rian ducts might be automatically regulated. Following trunk elongation,
the ducts are longitudinally stretched and the cell density per unit
volume is reduced. This density might be automatically restored to the
standard level by an appropriate mitotic rate. It becomes evident that
the left duct, more intensively submitted to stretching, shows a higher
level of cellular proliferation (i.e. growth), than the right one. Such
automatic regulation of mitotic activity has been suggested with regard
to other models (Wigglesworth, 1964).
Thus, the relative loss of stretching of the right Mullerian duct

(probably due to a specific pattern of visceral development in the right
ostial region) would be in my hypothesis, the "primary cause" of its
relative shortening during the first phase of the Mullerian differen-
tiation in the female bird embryo (i.e. between the 11th and the 14th
day of incubation in the chick embryo). It is highly probable that in
a later stage of this differentiation, the secretion of the feminizing
hormone by the embryonic ovary, selectively stimulates the growth of
the left duct, and particularly its cloacal region, which differentiates
into a shell-gland anlage.
Among the observations which may be presented to support this inter-

pretation of the. origin of the right Mullerian duct's rudimentation,
we can select the following: It is not rare to observe the relative
shortening of the right Mullerian duct in 10 day old male chick embryos.
This phenomenon is thus quite independent of the general destruction
of both Mullerian ducts in the male, caused by the testicular hormone.
In this connection, it is of interest to recall the observation of
Hamilton and Teng (1965) that "for female embryos, regardless of the
time of injection or size of the dose of steroid hormone, some invo-
lution of the right duct could not be prevented" (while relatively
low doses are sufficient to prevent the "male type" regression of the
ducts) .
Anomalies in the organogenesis of the Mullerian duct system are ex-

tremely rare. Among several thousand chick embryos I have observed,
only one 13 day old female showed bilateral reduction of its Mullerian
ducts. It was noted that on the left side, the reduction was accompa-
nied by an abnormal extraperitoneal development of the metanephric
tissue at the normal ostial level (Fig. 3). The atypical reduction in
length might thus be explained by a right side type disconnection of
the duct's ostial extremity from the body wall, due to the atypical
extrusion of the metanephros, subsequent loss of stretching and fall
of the growth rate.
The observation by Scheib (1955) that Mullerian ducts, unlike other

embryonic organs, are unable to develop in vitro on standard anhormonal
medium, might also be explained with reference to our interpretation
of the stretching dependent growth of these organs.
Control grafting experiments have fortuitously brought an additional

argument to my hypothesis (Thiebold, 1962b). For reasons of convenience
and in order to standardize the experimental conditions, all implanta-
tions have been performed into the anterior right coelomic space, i.e.
the neighbourhood of the right ostial region. In a great number of oper-
379
Fig. 3. Abnormal retrogression of the

level of the left ostium (arrow) in a
female 13 day embryo. Note the extraperi-
toneal development of the anterior part
of the metanephros. LG Left gonad; MsN
Mesonephros; MtN Metanephros
ated females, retention of the whole right Mullerian duct has been ob-
served after implantation of the embryonic ovary as well as of "sexu-
ally neutral" tissues such as limb buds, frontal buds, mesonephric
fragments, etc. (Fig. 4 and 5). In such 14 day old operated embryos,
the maintained right Mullerian duct showed a typical oviductal differ-
entiation as compared with the left homologue, its cloacal end swelling
into a shell-gland anlage (Fig. 5). In all cases of retention and devel-
opment, the presence of the graft (attached simultaneously to the body
wall and to the ostial extremity of the duct by means of epithelial
tracts), had obviously prevented the anterior disconnection which nor-
mally occurs during the eleventh day of incubation, and which marks
according to my interpretation, the beginning of the relative shorten-
ing. Thus, the fact may be explained that in several experimental cases,
the right ostium was more anteriorly situated than the contralateral
(Fig. 4), since the latter undergoes a slight retrogression relative
to its initial situation after the 10th day of incubation. These re-
sults, and especially the regional differentiation of the maintained
right Mullerian ducts, suggest that these organs must assume a normal
growth and reach a given morphological state that enables them to re-
spond to hormonal stimulation (differentiation of a shell-gland) .
The results of embryonic castrations performed by Wolff (1951) and

Groenendijk-Huijbers (1962), essentially based on which is the "hor-
monal interpretation" of the right Mullerian duct's rudimentation, are
worthy of discussion. Total castration resulted in the complete reten-
tion of the Mullerian ducts in both male and female embryos.
It should first be noted that the number of total castrations, followed

by right Mullerian duct retention, does not seem sufficient to allow
definitive conclusions.
380
Fig. 4. Anterior position - largely in front of the proximal border

(edge) of the mesonephros - of the ostium of a maintained right Mulle-
rian duct. Note the connection of the ostium with the graft (frontal
bud, arrow). 15 day female embryo. RG Right gonad
Fig. 5. Equal development of the shell gland anlage of the left and
maintained right Mullerian ducts. Same embryo as Fig. 4
In Wolff and Wolff's (1951) experiments (X-irradiations of the young

embryo's region giving rise to gonads) it seems if we refere to pic-
tures, that the treatment resulted in deep teratologic effects on the
general organization of the irradiated region.
With regard to Groenendijk-Huijbers's experiments (surgical interven-

tion, destruction of gonadal primordium of four days old chick embryos
by electro-coagulation) no sham operation was mentioned in the author's
paper. Thus, it remains possible to interpret that the surgical inter-
vention alone, disturbing the normal organogenesis of the right ostial
region (in the same way as do the implantations cited above), is re-
sponsible for the retention of the right Mullerian duct. Moreover, in
a more recent paper (Groenendijk-Huijbers, 1965) the same author empha-
sized "genetically determined differences among the strains concerning
the preservation of the right Mullerian duct" (as previously described
by Morgan and Kohlmeyer, 1957; Morgan and Greb, 1959)
Under these conditions, my opinion is that embryonic castration experi-

ments as yet performed, do not allow any conclusion to be reached about
the determination of the Mullerian differentiation in the female chick
embryo. The anomalous androgenic activity of the ovary that is basic
to the "classical" view of the origin of Mullerian asymmetry is not
warranted if my hypothesis is proven to be true.
381
References
Akram, H., Zeiss, A., Weniger, J.-P.: Sur l'activite hormonale des
gonades de l'embryon de Canard femelle hyophysectomise. C.R. Soc.
BioI. 167,340-341 (1973).
Groenendijk-Huijbers, M.M.: Functional characteristics of the testicular
hormone in the chick embryo. Anat. Rec. 111, 237-249 (1960).
Groenendijk-Huijbers, M.M.: The cranio-caudal regression of the right
Mullerian duct in the chick effibryo as studied by castration experi-
ments and oestrogen treatment. Anat. Rec. 142, 9-20 (1962).
Groenendijk-Huijbers, M.M.: The right ovary of the chick embryo after
early sinistral castration. Anat. Rec. 12}, 93-106 (1965).
Hamilton, T.H.: Hormonal control of r,mllerian duct differentiation in
the Chick embryo. Proc. XIII Intern. Ornith. Congr. Ithaca, N.Y.,
pp. 1004-1040 (1963).
Hamilton, T.H., Teng, C.S.: Sexual stabilization of Mullerian ducts in
the Chick embryo. In: Organogenesis (eds. De Haan and Ursprung),
pp. 681-700. New York-Chicago-San Francisco-Toronto-London: Holt,
Rinehart and Winston 1965.
Mikamo, K., Witschi, E.: Functional sex reversal in genetic females
of Xenopus taevis, induced by implanted testes. Genetics 48, 1411-1421
(1963) .
Morgan, W., 'Greb, R.J.: Genetic evolutionary aspects of double oviducts
in GaUus. Poultry science li, 1456-1462 (1959).
Morgan, W., Kohlmeyer, W.: Hens with bilateral oviducts. Nature 180,
98 (1957).
Raynaud, A., Pieau, C., Raynaud, J.: Etude histologique comparative de
l'allongement des canaux de Muller, de l'arret de leur progression
en direction cauda Ie et de leur destruction, chez les embryons males
de diverses especes de Reptiles. Ann. Ernbr. Morph. 3, 21-47 (1969).
Thiebold, J.-J.: Agenesie partielle des canaux de Muller, provoquee
par la greffe de testicules embryonnaires chez l'embryon d'Oiseau.
C.R. Acad. Sci. Paris, 254, 2450-2451 (1962a).
Thiebold, J.-J.: Sur Ie maintien et Ie developpement du canal de Muller
droit de l'embryon de Poulet femelle dans certaines conditions ex-
perimentales. C.R. Acad. Sci. Paris 255, 2665-2666 (1962b).
Thiebold, J.-J.: Contribution a l'etude de l'organogenese urogenitale
et de son determinisme chez un Poisson Elasmobranche: la petite
Roussette Scytiorhinus canicutus (L.). Bull. BioI. France Belgique ~,
253-347 (1964a).
Thiebold, J.-J.: Sur la sensibilite des canaux de Muller aux hormones
genitales chez l'ernbryon d'Oiseau. Ann. Endocr. Paris 24, 136-139
(1964b) • -
Thiebold, J.-J.: Quelques observations sur la developpement du canai
de Muller droit chez l'embryon de Poulet. C.R. Soc. BioI. 167,
338-339 (1973). -
Thiebold, J.-J., Reichart, J.M.: Inhibition du developpement ovarien
par la greffe de gonades embryonnaires chez Ie Poulet. C.R. Soc.
BioI. (1974, in press) .
Wolff, E.: Recherches sur l'intersexualite experimentale produite par
la methode des greffes de gonades a l'embryon de Poulet. Arch. anat.
micro morphol. expo 31, 237-310 (1946).
Wolff, E.: Le determinisme de l'atrophie d'un organe rudimentaire:
Ie canal de Muller des embryons males d'Oiseaux. Experientia 9,
121-133 (1953). -
Wolff, E.: Endocrine function of the gonad in developing vertebrates.
In: Comparative Endocrinology (ed. A. Gorbman), pp. 568-573. New
York: John Wiley and Sons 1959.
Wolff, E., Stoll, R.: Le role de l'hypophyse dans Ie developpernent
embrynonnaire du Poulet, d'apres l'etude des cyclocephales experi-
mentaux. C.R. Soc. BioI. ill, 1215-1217 (1937).
Wolff, Et., Wolff, Ern.: The effects of castration on bird embryos.
J. expo Zool. 11&, 59-98 (1951).
Intersexuality of the Genital System and "Free-Martinism"
in Birds
H. Lutz and Y. Lutz-Ostertag
A. Introduction
The development of the bird embryo and of its genital system is well
known. The male or the female sex develops from a sexually indifferent
stage which is characterized by paired gonads, localized on the median
surface of the mesonephros, and by two Mullerian ducts. In the male
embryo the indifferent gonads differentiate into paired testes, while
the two Mullerian ducts become rudimentary and disappear. In the female
embryo only the left gonad becomes an ovary and increases in size,
whereas the right gonad undergoes degeneration. Correspondingly, the
left Mullerian duct is transforming to an oviduct whereas the right
one becomes rudimentary.
However, many authors have pOinted to deviations from this general rule.
For example, hermaphroditism or the development of two ovaries may oc-
cur. In the male embryo it has been observed that both Mullerian ducts
(or only the right one) may persist, more or less completely. In the
female embryo the right Mullerian duct can be preserved and differen-
tiate into a shell gland (Romanoff, 1960). The causes of such sponta-
neous sexual abnormalities are unknown although they can be experimen-
tally elicited.
Since the classical studies by Wolff and Ginglinger (1935) many authors
have investigated the effect of sex hormones on the genital tract.
Castration produced by X-irradiation (Wolff and Wolff, 1951) can pro-
duce asexual animals. Various types of grafting experiments (Wolff,
1946-1947; Lutz-Ostertag and Negre, 1960; Lutz-Ostertag and Didier,
1961) have demonstrated the effects of testicular and ovarian grafts
on the genital tract of the host embryo. Finally, by the use of organo-
typic culture experiments, the role of sex-hormones could be further
elucidated. It has been possible to observe the development of the sex
organs outside the organism including parabiotic associations of testes
and Mullerian ducts (Lutz-Ostertag, 1954).
It is surErising that for a long time an interesting natural experi-

ment escaped the attention of research workers: the occurrence of free-
martins in birds which originate from eggs with double yolks. We have
studied this for the first time in 1957 and use the term "free-marti-
nism" in comparison to the well known phenomenon in the Bovidae. From
an egg with two yolk masses two embryos may develop, which are either
of the same or of a different sex. Only the latter case has been of
interest to us. In order to assure a normal development the blastoderms
must both be situated on the upper surface of the egg. However, this
is not always the case. More than 500 chick eggs with double yolks and
700 duck eggs have been used for our observations. Only 21 pairs of
chick embryos (each pair consisting of 1 male and 1 female embryo) and
17 pairs of duck embryos developed, from 16 to 19 days in the first
species and 17 to 22 days in the second. Macroscopical and histological
383
studies of these embryos clearly demonstrated masculinization of the

females and feminization of the males (Lutz and Lutz-Ostertag, 1959).
B. Results
I. Masculinization of the Females
In all cases the masculinizing effect is observed on the Mullerian ducts

of the embryos of both species. In extreme cases the left Mullerian
duct is completely missing, but in the majority of embryos it persists
- either as a more or less well developed fragment, or of spare vesicles
interconnected with each other and with the cloaca by the mesothelium.
The right Mullerian duct undergoes a caudo-cephalic involution which
exceeds the classical cephalo-caudal reduction (Figs. 1 and 2) .
Figs. 1 and 2. Genital tracts of chick female embryos. Left Mullerian

duct (lMd) strongly reduced remaining only in the form of vesicles
(vMd) (Fig. 2); right Mullerian duct (rMd) longer as normally;
ov ovary
3M
In the duck embryo, this masculinization also effects the spiralization

of the genital tubercle which qifferentiates towards the male type and
of the syrinx, which develops a very obvious asymmetry which is char-
acteristic of the male sex (Fig. 3).
Fig. 3. Duck female embryo. Right

and left Mullerian ducts remain-
ing on the form of portions and
vesicles. Syrinx (e) strongly
asymmetric. Genital tubercle (gt)
spiralized
The ovary is thin and flat, the thickness of the medulla and especially
that of the cortex are reduced. In the latter case the absence of germ
cells is noteworthy.
II. Feminization of the Males
This phenomenon is less obvious than masculinization of the females.

Macroscopically, one can see the occurrence of two small vesicles near
the cloaca, situated on the right and left side of the bursa Fabricii
(a rudimentary duct of only a reduced mesothelium running laterally
to the mesonephros), (Fig. 4). These are small preserved remnants of
the Mullerian ducts.
The testes, however, seem to be more sensitive to the influence of the

female hormone. Their size is slightly smaller than that in mormal male
embryos of the same stage. The left testis is often flat and has an
irregular and granulated surface. In these cases the histological study
385
reveals the presence of cortical clusters (Fig. 5) or even a cortical

strip. Sometimes the tubules penetrate the tunica albuginea and are
attached directly to the coelom, which indicates a low degree of inter-
sexualiby (Wolff and Hampe, 1950).
Fig. 4. Chick male embryos.

Granulated testes (t). On the
right and left side remains
of !1lillerian ducts
____
alb
~Id:I_-L
Fig. 5. Presence of corti-

cal clumps (e) on the
testes. Well developed
albuginea (alb)
The syrinx of the male duck embryos undoubtedly corresponds to the

male type. The male character is slightly less marked on the genital
tubercles - which show a lesser degree of spiralization.
All these observations were later confirmed by Ruch (1962, 1968).

386
The results which have been reported in this paper agree with those
obtained by the methods of intra-coelomic grafts (Bradley, 1941; Wolff,
1946-1947), or chorio-allantoic grafts of gonads (Lutz-Ostertag and
Negre, 1960; Lutz-Ostertag and Didier, 1961). In order to explain the
intersexuality which has been obtained in the host embryo, in partic-
ular after chorio-allantoic grafts, it is suggested that the hormone
which has been secreted by the male or the female gonad was transported
via the vascular system. An analogous explanation could be applied to
our freemartin cases. Opening of the eggs (as described by Wolff, 1936),
permits the observation of both embryos in each double-egg and espe-
cially their extra-embryonic vascular systems which, in most cases, are
coming in contact with each other at about the third or fourth day of
incubation. At this time it is possible to observe anastomoses. This
phenomenon has also been clearly demonstrated by Ruch (1961). After
injection (at very low pressure) of diluted Chinese ink into the peri-
pheral vessels of one of the embryos, Ruch observed the rapid passage
of the ink into the second twin embryo.
The explanation provided by Keller and Tandler (1916) and Lillie (1916)
seems to be also valid in our case. The transfer of sexual hormones
from one embryo to the other is assured by the circulatory system.
These hormones, however, are effective only when they act very early,
that is before the 7th day of development. This was shown by opening
the eggs at an early stage. When both embryos are situated at a certain
distance or when the yolks are separated by a large vitelline bridge,
we were able to observe that there was no connection between the two
vascular systems and that the genital tract belonging to the two twin
embryos differentiated normally.
These observations have enabled us to understand the relatively dis-

creet effects of the embryonic hormones which act before the 7th day
of development. All these conclusions have also been confirmed by Ruch
(1962). In addition, Ruch has found that the embryo can show symptoms
of feminization when only one embryo develops on two yolks which are
not entirely separated. This occurs frequently (Romanoff and Romanoff,
1949). Ruch estimated that each yolk mass contains 30 - 40 ~g of "fol-
liculine". Consequently, the double amount of this hormone is available
to the embryo and this quantity appears to be large enough to produce
a second proliferation of the germinal epithelium (Wolff and Hampe,
1950) •
All the mentioned authors have used sex hormones in order to obtain
modifications of the genital system. Stoll (1948), however, has re-
ported that a rise of the incubation temperature to 40,5 0 C during the
early stage of development, stabilizes the Mullerian ducts of the male
embryo until the end of the incubation time. The right Mullerian duct
of the female embryo can develop well under these conditions. These
results have been confirmed in the quail embryo by Lutz-Ostertag (1966).
By the use of teratogenic substances such as 1162 F (para-aminobenzene
sulfamide), albucid (para-aminobenzene sodium sulfonacetylamide),
disodium methyl arsenate, sodium dimethyl arsenate, arrhenal and so-
dium cacodylate, saponin and eserine salicylate and non-teratogenic
compounds like saccharin and diphteria toxin, Stoll (1948) obtained
the same effects as with sex hormones in the female embryo. However,
these chemical products have no effect in the male embryo.
From these facts it can be seen that a teratogenic agent can stabilize
the development of organs which are normally transitory in the embryo's
life.
387
III. Effects of Pesticides on the Bird's Genital Tract
For the last 20 years increasing amounts of pesticides have been widely
spread in the natural environment. The question arises whether these
products might also have an effect on the viability of the bird embryo
and its ability to hatch, as on the development of its genital system.
We are briefly discussing only this latter aspect. If we intend to
study the influence of these products on the development of embryos
of wild birds and their genital tract we will have to work under con-
ditions which are similar to those found in nature. Therefore, commer-
cial products would have to be employed at doses which are indicated
by the manufacturer. In our laboratory experiments all the eggs were
similarly treated, which is hardly the case under natural conditions.
The following pesticides have been used:

Organo-phosphoric insecticides:
Parathion (chick, quail)
Azinphos (pheasant, grey partridge, red-legged partridge)
Organo-chlorinated insecticides:
Aldrine (chick, quail, pheasant)
Endosu.lfan or Thiodan (chick, quail)
DDT (chick, quail)
Lindane (quail)
Herbicides and anti-brushwood sprays:
2,4-D (pheasant, grey partridge, red-legged partridge, quail)
2,4,S-T (chick, quail)
Simazine (chick, quail)
Synergized Pyrethrines: (chick, quail)

All these pesticides which have been tested up to now have an effect
on the genital tract of the various species. In quail, the Lindane
acts at slightly higher doses only.
The male embryos present symptoms of feminization: testes with more

or less important cortical clusters; incomplete (or no) regression of
the Mlillerian ducts; rudiments persisting as vesicles. In certain
cases, the Mlillerian tract is similar to that of a normal female em-
bryo (Fig. 6) - namely, a structure with a shell gland and a rudimen-
tary rigilt duct.
In the female embryo (except the red-legged partridge treated with

Azinphos and the quail sprayed with low doses of Lindane) the right
duct shows a reduced cephalo-caudal regression, but it very often
reaches a length corresponding to 1/3 up to 3/4 of the contra-lateral
left duct. In some cases, the posterior part of this duct develops
into a shell gland, often similar to that of the normal female embryo
at the same stage (Fig. 7). The right gonad regresses to a small ex-
tent only and develops into a gonad, analogous to a true ovary (Fig.8)
Whereas sometimes the size of the left ovary remains unchanged, in
most cases the shape of this gonad is abnormal - according to the
pesticide used. It has to be pointed out that these modifications are
often accompanied by an important diminution of the number of germ
cells which is associated with a reduced fertility of the embryos.
This observation was confirmed by David (1973), on 6 days old chick
embryos that were treated with DDT before incubation.
We found that after spraying of quail eggs with 2,4-D the individuals
which hatched from such eggs show a reduced fertility: delay of the
laying-time; initial diminution of the egg-size; and a smaller ovary.
388
Fig. 6. Pheasant male embryo. Effects of 2,4-0. Mulle rian ducts weakly
regres~ed
Fig. 7. Pheasant f e male embryo. Effects of 2,4-0. Right Mullerian duct

partially regressed but with a well d eve lop e d shell gland
From experiments with the organo-typic culture-technique on a medium

which contained Parathion, 2,4,5-T or Simazine, it became evide nt that
these products have a selective effect on the embryonic germ c e lls,
but they can also influe nce the Sertoli c e lls and induce a vacuolar
appearance of the cultured gonad.
The problem arises, in which way the partial regression or the p e r-

sistence of the Mullerian ducts (which is symptomatic for the femi-
nizing effect in t h e male embryo), can be explained. In order to e lu-
cidate this phenomenon one of us, in collaboration with David (1974),
has cultivated in vitro (1) Mullerian ducts from 9 days old normal chick
embryos on a medium containing DDT; and (2) Mullerian ducts from t reated
male embryos on the standard medium of Wolff et al. Under these c on-
ditions the ducts which normally regress in culture (Lutz-Ostertag,
1954) did not involute, or regressed only partially. Since the regres-
sion is normally caused by the action of a proteolytic enzyme (Sch e ib-
389
Pfleger, 1955i Lutz and Lutz-Ostertag, 1956) it seems that DDT has a
direct action on this enzyme. This assumption is confirmed when Mlil-
lerian ducts, which had been taken from DDT-treated embryos, were as-
sociated by the same method with testes from normal embryos. In 60%
of the cases, despite the presence of androgenic hormones, the ducts
did not regress. In similar experiments, Didier and Lutz-Ostertag
(unpubl.) have also demonstrated that the 2,4,5-T and Simazine have
a direct influence on the proteolytic enzyme. In addition, both authors
have provided evidence that these two pesticides also elicit a change
of steroid production. When male gonads from treated embryos are grafted
on a normal host-embryo the gonads of the latter become granular. This
corresponds, as shown by histological investigations, to a low femini-
zation - as can be seen from the presence of cortical clusters which
are colonized by germ cells.
Fig. 8. Grey Partridge female

embryo. Effects of Azinphos.
Presence of a right ovary
All these results suggest that some of the pesticides have a complex
effect: either they provoke an inhibition of the proteolytic enzyme
or a change of the hormonal secretion of the testes.
C. Conclusion
Our experiments with the use of pesticides reveal their hazardous char-
acter, with regard to the development of the genital tract in birds.
These products can mimic the effects of estrogenic sex hormones: with
feminizing actions in the malei and a "superfeminizing" influence in
the female embryos. The most spectacular effect is exerted on the level
of the primary germ cells and leads to sterility, or reduced fertility.
390
Such facts demonstrate the resulting hazards when these products are
spread in large quantities in our natural environment. One should be
careful, however, not to draw more general conclusions from the obser-
vations in one species only, since some are more sensitive than others
and at present, there are no particular criteria to judge the vulnera-
bility of a -given species. Only careful experiments can provide us
with the information that we need. At any rate, these products should
be used in nature very cautiously.
References
Bradley, E.M.: Sex differentiation of Chick and Duck gonads as studied

in homoplastic and heteroplastic host-graft combination. Anat. Rec.
79,507-530 (1941).
David, D.: Action du DDT sur Ie peuplement en gonocytes des ebauches
gonadiques de l'embryon de Poulet age de 6 jours. Experientia ~,
81 5- 81 6 (1 973) .
lingstrachtigkeit des Rindes. Wien. tierarztl. Wschr. 2, 513-526
(1916).
Lillie, F.R.: The theory of free-martin. Science il, 611-613 (1916).
Lutz, H., Lutz-Ostertag, Y.: Contribution a l'etude du maintien des
canaux de Muller de l'embryon de Poulet male par l'action des ultra-
sons. Arch. Anat. micro Morph. expo 46, 218-235 (1956).
Lutz, H., Lutz-Ostertag, Y.: Free-Martinisme spontane chez les Oiseaux.
Devel. Biol. I, 364-376 (1959).
Lutz-Ostertag, Y~: Contribution a l'etude du developpement et de la
regression des canaux de Muller chez l'embryon d'Oiseau. Bull. BioI.
France et Belgique ~, 333-412 (1954).
Lutz-Ostertag, Y.: Action de la chaleur sur Ie developpement de l'ap-
pareil genital de Caille (Coturnix coturnix japonica) . C.R. Acad. Sci.
262, 133-135 (1966).
Lutz-Ostertag, Y., Didier, E.: A propos du Free-Martinisme spontane
chez les Oiseaux. Action des testicules sur l'embryon de Poulet
femelle. C.R. Soc. Biol. 155, 2232-2234 (1961).
Lutz-Ostertag, Y., Lutz, H. :-sexualite et Pesticides. Annee Biologique,
4e Serie 13,173-185 (1974).
Lutz-Osterta~ Y., Negre, Ch.: Contribution a l'etude du Free-Marti-
nisme chez les Oiseaux. C.R. Soc. BioI. 154, 2284-2286 (1960).
Romanoff, A.L.: The avian embryo. New York: MacMillan Company 1960.
Romanoff, A.L., Romanoff, A.G.: The avian egg. New-York: John Wiley
and Son 1949.
Ruch, J.V.: Contribution a l'etude des modifications du systeme geni-
tal des embryons issu d' oeufs doubles de Poule (Gallus domesticus) .
Arch. Anat. Hist. Embr. 45, 63-129 (1962).
Ruch, J.V.: Modifications dU-systeme genital observees chez des em-
bryons issus d'oeufs doubles de Poule. Arch. Anat. Hist. Embr. 21,
607-612 (1968).
Scheib-Pfleger, D.: Contribution biochimique a l'etude des processus
de differenciation des canaux de Muller chez l'embryon de Poulet.
Bull. BioI. France et Belgique 89, 404-490 (1956).
Stoll, R.: Action de quelques agents physiques et chimiques sur les
canaux de Muller de l'embryon de Poulet. Arch. Anat. micro Morph.
expo 37, 333-346 (1948).
Wolff, Et:: Les bases de la teratogenese experimentale des Vertebres
Amniotes d'apres les resultats des methodes directes. Arch. Anat.
Hist. Embr. il, 1-382 (1936).
391
Wolff, Et.: Recherches sur l'intersexualite experimentale produite

par la methode des greffes de gonades a l'embryon de Poulet. Arch.
Anat. micro Morph. expo 36, 69-90 (1946-1947).
Wolff, Et., Ginglinger, A.:-Sur la production experimentale d'inter-
sexues par l'injection de folliculine a l'embryon de Poulet. C.R.
Acad, Sci. Paris 200,2118-2121 (1935).
Wolff, Et., Hampe, A~Sur la reprise de la poussee des cordons sexuels
males sous l'influence d'injections d'une hormone femelle a l'embryon
d'Oiseau. C.R. Soc. Biol. 144, 1100-1102 (1950).
Wolff, Et., Wollf, Em.: The effects of castration on bird embryos.
J. expo Zool. 116, 59-98 (1951).
The Prenatal Development of Bovine Freemartins
A. Jost, J. P. Perchellet, J. Prepin, and B. Vigier
A. Introduction
Freemartins in cattle have long been a classical example of naturally

occurring intersexuality in a cornmon animal species. The beautiful
studies of Frank Lillie (1916, 1917, 1923) and of his pupils (Chapin,
1917; Willier, 1921; Bissonnette, 1924, 1926) and those of Keller and
Tandler (1916) are well known. The freemartin condition inspired theo-
ries of sex differentiation and many brilliant experiments in amphib-
ians and other vertebrates. It suffices to recall that a freemartin
is a genetic female calf born as a twin to a bull calf, and afflicted
with more or less severe abnormalities of the internal reproductive
organs. The prerequisite for the appearance of these anomalies is that
the chorionic blood vessels of the male and female twins are fused and
permit exchange of blood. According to Lillie this also permits trans-
fer of testicular hormone into the female fetus. In the postnatal con-
dition, the freemartin gonads seldom display any ovarian characteris-
tics (Chapin, 1917; Rajakoski and Hafez, 1963); they may be extremely
reduced ("asexual type" of Witschi, 1939) or contain sterile cords very
similar to testicular seminiferous cords; clusters of interstitial
cells may be recognized (e.g. Willier, 1921; Bissonnette, 1926; Short
et al., 1969). Release of testosterone into the gonadal vein has been
established in one especially strongly masculinized adult freemartin
(Short et al., 1969).
The condition of the sex ducts also differs from one animal to the
other. The upper (tubal) segment of the female ducts (the derivatives
of the embryonic Mullerian ducts) has usually disappeared; in the lower
segment the uterine part of the Mullerian ducts may have retrogressed
more or less completely. Two male structures, derived from the upper
and lower extremities of the Wolffian ducts, namely the epididymis and
seminal vesicles, are usually present, whereas in some cases 'the vasa
deferentia persist. A prostate has rarely been observed (Willier, 1921;
Fraser-Roberts and Greenwood, 1928) and the external genitalia are
exceptionally masculinized (Buyse, 1936; Kieffer and Sorensen, 1971).
It is worthwhile to recall some of the assumptions accepted by Lillie

and his school concerning the differentiation of the sex glands (for
pertinent literature see Jost et al., 1972, 1973a,b). They adhered to
the theory of the sex cords ingrown from the superficial "germinal
epithelium"; according to this theory the primary sex cords which dif-
ferentiate into seminiferous tubules in the developing testes remain
underdeveloped in the "medulla" of the developing ovary. "Secondary
sex cords" or "cords of Pfluger" will later constitute the ovarian
cortex. The first freemartin effect was interpreted as an inhibition
of the proliferation or the ingrowth of the Pfluger's cords. However
it is very significant that there is in no paper of this school any
readable picture showing the ovarian medullary cords. On the contrary,
in the description of a 3,75 cm-long freemartin and a control female,
Lillie (1923, p. 67) states that in the female "the cords of Pfluger
393
are definitely indicated but that the primary sex cords are not sharply
differentiated from the stroma of the medulla"; the "medulla" of the
freemartin did not much differ from that of the female. Lillie also
noted that the inhibition in gonadal growth is accompanied by inhibi-
tion of the growth of the cortex "though apparently not as yet to any
marked degree, by stimulation of the male homologs in the ovary"
(Lillie, 1923, p. 65). In the drawings accompanying the descriptions,
nothing like "primary cords" can be seen in the medulla.
Nevertheless Lillie accepted the concept that primary sex cords were
present and that under hormonal stimulation could be converted into
seminiferous tubules. Witschi's (1931, 1957, 1967) theory of the dual
inductor system in the early gonadal primordium contributed to the
belief that the seminiferous tubules seen in the postnatal freemartin
gonads were simply the primary sex cords laid down at an early stage
in the ovarian primordium.
Another point of interest concerns the time of onset of hormonal ac-

tions and the role of hormones in sex differentiation. According to
Lillie (Lillie and Bascom, 1922; Lillie, 1923) male hormone was pro-
duced only after the differentiation of the testes and did not control
the initial gonadal differenti'ation. Others accepted the view that the
ini tial diff,erentiation of the gonads is controlled by inductors
(Witschi, 1957, 1967) or by hormones (Wolff and Ginglinger, 1935).
Finally the role of the testicular factor in producing the freemartin

effect was diversely interpreted. According to Lillie (1923) the male
hormone had successively several effects, namely inhibition of the
differentiation of the ovarian cortex, degeneration of the MUllerian
ducts and later "stimulating effects on the sex cords and rete of the
gonad and on the epididymis and Wolffian duct". The same male hormone
was held to inhibit heterologous sex characters and to stimulate the
homologous ones (Lillie, 1923, p. 71). Witschi (1965) proposed that
only a "medullary antagonist" (a gonadal inductor) is passed from the
male to the female fetus, and that it inhibits the cortex of the free-
martin gonad, thus permitting "compensatory testicular development".
The modified gonads would then masculinize the genital tract.
During the last decade we have studied several aspects of sex differ-
entiation in freemartin and in control fetuses and we encountered
several of these fundamental questions concerning normal and abnormal
sex differentiation.
B. Material and Techniques
Up to the present time we have studied histologically 82 freemartin

fetuses, 111 control males and 81 control females at ages ranging from
39 to 250 days. In a series of animals XX/XY chromosomal chimaerism
has been examined in the liver (Vigier et al., 1972) or other tissues
(Vigier et al., 1973).
The 39 to 77-day-old freemartin fetuses were obtained experimentally,

superovulation being induced in cows with gonadotropic hormones, as
already reported (Jost et al., 1972). The post-insemination age of
these fetuses was known exactly. Those at older stages were collected
at the slaughter-house and their age was estimated according to their
size.
394
The only point in need of comment in the present paper concerns the
criteria used to diagnose presumptive freemartins at early stages.
As was previously reported (Jost et al., 1972), the chorions and their
blood vessels were carefully examined for vascular connections; in
freemartins over 52 days of age such connections are constantly ac-
companied by genital alterations observable macroscopically or micro-
scopically; in addition XX/XY mosaicism can be detected in the liver.
In younger pairs, when freemartin effects cannot yet be recognized,
it was admitted that females united to males were presumptive free-
martins when vascular anastomosis was clearly seen and when XX/XY chi-
maerism could be detected.
C. Sex Differentiation in Con,trol Fetuses
The chronology of sexual differentiation in normal males and in normal

females is summarized in Table 1.
Table 1. Developmental chronology of normal calf fetuses and of free-

martins (age in days)
Age Males Females Freemartins
40 Seminiferous cords Undifferentiated Undifferentiated

gonad gonad
Interstitial cells
Ano-genital dist." Gonadal growth stopped
50 Upper MUll. ducts \ Upper MUll. ducts \
Germ cells: no multi-
Scrotum, prostate plication
60 Male extern. genit.
70 Coiled epididymis
Branched seminal
vesicles First meiosis
Exceptional meiosis
80 Wolff. ducts inter- Wolff. ducts ± main-

rupted tained
90 Gonads resume growing

Earliest seminiferous
cords
Eventual masculiniza-
100 tion (semin. vesicles
etc.)
First ovarian
follicles
110
$5
I. Males
Testes become histologically recognizable on about day 40, when first

clear cells adhere to one another, eventually encompass germ cells
and thus delineate the primordia of the future seminiferous tubules.
Interstitial cells become recognizable a few days later. It should be
noticed that testicular organogenesis proceeds for several weeks; tu-
buli recti contacting a rete testis made of open canaliculi differen-
tiate only after day 70.
Concerning the other sex structures it is noteworthy that the earliest

trend towards maleness is migration of the phallus towards the umbili-
cus. Already on day 47 the anogenital distance is greater in males
than in females; ten days later the penis has its definitive location
under the belly wall and the scrotum is well developed. Retrogression
of the upper (tubal) part of the MUllerian ducts starts after day 50
(Fig. 3); the posterior parts of the uterine horns are still present
until approximately day 70; the MUllerian ducts have almost completely
disappeared on day 80.
The Wolffian ducts are to a large extent similar in both sexes until
approximately day 70. Their diameter similarly diminishes between
days 60 and '70, but afterwards they persist in males while in females
they slowly retrogress. In the female genital tract important segments
of the Wolffian ducts still can be seen 40 or 50 days later.
Seminal vesicles differentiate in males in two steps: (1) smallout-

growths of the posterior Wolffian ducts appear between day 56 and 58;
(2) after day 70 these simple diverticles become branched and grow
rapidly. In females, small Wolffian diverticles, homologous to seminal
vesicles appear (after day 70) later than in males in our material,
although Bissonnette (1924) and Krehbiel (1963) noticed them earlier.
On day 110 the branched seminal vesicles were 7 mm in length in males,
and the unbranched homologous structures in females only 0.5 mm.
The epididymis results ftom the coiling of the anterior part of the
Wolffian duct starting between days 70 and 80. No structure really
similar to the epididymis differentiates in females.
II. Females
During the first week following testicular differentiation in males,

the presumptive ovaries are undifferentiated. They are made of a super-
ficial epithelium, a central blastema of cells and, between these, a
looser layer of mesenchymal tissue. Germ cells are present throughout
the primordium. Later, as the number of germ cells increases, nests
of germ cells extend beneath the epithelium. Meiosis begins in the
deepest located germ cells on day 75. The role in folliculogenesis
of the central blastema and of the cellular strands which differen-
tiate in it (van Beek, 1924; Gropp and Ohno, 1966) still has to be
studied further. In all cases, true ovarian organogenesis resulting
in the appearance of ovarian follicles does not begin before day 100.
The uterine part of the MUllerian ducts becomes enlarged after day 60,
while the Wolffian ducts are still present. Short sections of the
Wolffian ducts have disappeared in a 79-day-old fetus, at the level
of the mesonephros. But anterior and posterior remnants still persist
until day 110. The prolonged persistance of the Wolffian ducts in fe-
males is an important feature in the explanation of the freemartin.
396
D. Sexual Organogenesis in Freemartins
Our earlier observations on the developmental history of freemartin

fetuses have already been reported (Jost et al., 1972, 1973a,b). Four
periods can be distinguished, more or less arbitrarily.
40 - 48 days. During the week following the onset of testicular organo-

genesis in males, freemartins were similar to normal females. The go-
nadal volume (Fig. 1) and the number of germ cells (Fig. 2) were some-
what reduced, but this result should be confirmed on a larger number
of cases.
I
I
t2
3-0 2~ __ " .....01
t , ,.j/
' "31( i
II
I
"
1 i
25
i
I ""
l As
i
I
/ 1>1' .
J
Il,
2.0 ,
T / ,/
/
/ : "
.r~'
:.
/1 "
1.5
to
40 42 45 49 52 55 60 62 age
Fig. 1. Variations in gonadal volume according to gestational age in

males ( 0 ) , normal females (0) and freemartins (.) between 40 and
62 days. Each point on the curves gives the mean volume of both gonads
in all fetuses of the same age group. The numbers indicate the number
of fetuses for each age and the vertical lines the range of individual
values
49 - 60 days. This period is of utmost importance in male fetuses,

since it covers the differentiation of most of the male sex characters
- except the derivatives of the Wolff ian ducts - and the retrogression
397
of the tubal part of the Mullerian ducts. During the same period the
first anomalies occur in freemartins, but it should be immediately
emphasized that these anomalies do not include any masculinization of
either the gonads or of the genital tract or of the external genitalia.
In freemartins, the upper Mullerian ducts become inhibited and retro-
gress from day 52 on as they do in males (Fig. 3). In the meantime,
the growth - i.e. the increase in volume - of the gonads is stopped
(Fig. 1). The number of germ cells no longer increases (Fig. 2) - in
control females there is a 90-fold increase in the number of germ
cells between 50 and 70 days (Fig. 2). As a result the so-called
"cortical cords" do not differentiate or are stopped.
20.0.1
10.0.
50.
40.
30.
20.
...........'"u
10.
....
:E
e
5
4
3
2
Fig. 2. Average number of germ
~ cells (Chalkley's technique)
for one gonad in normal females
( 0) and in freemartins (.)
0.50. between 40 and 80 days. Numbers
indicate the number of fetuses
0.25 studied at the same age (log.
40. 50 60. 70. 80. DAYS
plot)
60 - 90 days. This period is rather arbitrarily distinguished from the

following one. The major reason for considering it by its'elf is that
it precedes the differentiation of follicles in females, when the num-
ber of germ cells reaches a peak value (Erickson, 1966). Meiosis starts
on day 75. In males, testicular organogenesis proceeds, and the seminal
vesicles and epididymides clearly differentiate.
In freemartins, we observed meiosis exceptionally in a localized super-

ficial area of the gonad of a 77-day-old freemartin. As a rule, growth
of the gonad remained arrested and the superficial epithelium vanished;
a testicular-like albuginea differentiated after day 70.
Small unbranched Wolffian diverticuli (homologous to seminal vesicles)

may be present from day 62 on, but as a whole, even between 70 and 88
days of age, signs of masculinization, when present, were slight.
Over 90 days. This period includes ovarian differentiation in normal

females and completion of the differentiation of the testicular duc-
tuli in males.
In freemartins the gonads resume growth during the next three months,
and this growth concerns the rete tubules as well as the cellular
398
blastema surrounding the rete below the albuginea. In 15 out of 33

fetuses studied between days 93 and 250, structures similar to fetal
testicular seminiferous cords were observed in the blastema (Fig. 4)
Sometimes similar cords were connected with more or less irregular
clusters of clear cells aggregated together, occasionally with germ
cells. These cellular organizations resembled seminiferous cords in
the stage of formation. Since they were found in freemartins at ages
ranging from 93 to 222 days, it would appear that seminiferous cords
can differentiate in the blastema at various ages. Details concerning
these formations were given in a previous paper (Jost et al., 1973b)
and will not be reported again. Some conclusions will be quoted.
60
__ ~ o,
:~---
; :' ...
50 A~
,01# :
.~ . ..ol
-;<>--1-5 -
40
30
\
\ 3
10
6,
40 45 49 52 55 60 62 oge
Fig. 3. Variations in average diameter of the upper Mullerian duct

wi th gestational age in males (0), in normal females (0) and in
freemartins (.). Numbers indicate the number of fetuses for each age.
The vertical lines indicate the range of individual values.
Among the 15 freemartins showing cords in their gonads, the number

and size of the cords were not clearly related to the age of the fetus.
The best-developed cords were seen in a 97-day-old freemartin, but
usually the cords were present in the largest gonads and in those in
which the rete was the largest. Germ cells were present in these cords
(or in the rete) in the younger cases; after day 150 the germ cells
became scarce. It seems that even if present, they disappear after
399
the fifth month of pregnancy. It should also be noticed that "forming

cords" deprived of germ cells were seen in a 222-day-old freemartin.
This would suggest that germ cells do not play an indispensable role
in the organogenesis of "seminiferous cords".
Fig. 4 A and B. Histological sections through (A) the gonad of a

97-day-old freemartin showing a seminiferous cord with germ cells
and clusters of interstitial cells (x 600); (B) the epididymis of
the same animal (a part of the rete is also seen). (x 48)
It should be emphasized that "interstitial cells" could be histologi-

cally recognized with certainty only in the three most "masculinized"
fetuses, i.e. those which contained the greatest number of well dif-
ferentiated cords out of the 15 freemartins showing cords. Of course
the possibility cannot be excluded that physiologically differentiated
interstitial cells failed to be recognized in histological sections.
The condition of the genital tract deserves some consideration (Table 2).
In freemartins aged from 93 to 125 days, the seminal vesicles were
longer than in control females and branched as in males; epididymides
were present in only 6 animals. An epididymis was lacking in three
93- and 108-day-old freemartins; it cannot be excluded that an epidi-
dymis can still develop after the age of 108 days. Prostatic buds
were present on the urogenital sinus of 7 of these fetuses. The degree
of inhibition and retrogression of the MUllerian parts was variable.
A few pOints are of interest:

1. There is no correlation between the degree of inhibition of the
MUllerian ducts and the degree of masculinization of the Wolffian
ducts (Fig. 5) and of the prostate (Table 2) .
400
Fig. 5. Schematic recon-

struction of the inter-
~'
108 ~E
nal genital organs of
GO 110 freemartin fetuses;
G :" their age is printed in
E~", G , the figures (97, 108,
, '.
100 and 125 days). In
MD the upper part of the
~--,- MD figure are two freemar-
tins with no seminifer-
WD ous cords or intersti-
I - -WD
,, tial cells in their
sv sv
gonads (G) (fine punc-
"Vsv
tate); the Wolffian
~
ducts (W D) are inter-
rupted. In the lower
part of the figure are
~
~ two freemartins, the
gonads of which con-
tained many seminifer-
ous cords (heavy punc-
tate) (see Table 2).
The four animals hav e
epididymides (E) and
97 seminal vesicles (S V)
The condition of the
Mullerian ducts (MD)
varies
- WD
MD
sv
Ilmm
2. In those cases in which the Wolffian ducts are present and contin-
uous after day 80, one can assume that they were submitted to andro-
gens before day 77 or 78 (the age when they become interrupted, near
the mesonephros, in normal f e males) .
3. The relation between the condition of the gonads and that of the
genital tract is not unequivocal.
The two fetuses (97- and 125-day-old) which have the most "masculinized"
gonads according to the degree of differentiation of cords and inter-
stitial cells, also have the most strongly masculinized Wolffian deriv-
atives.
In the 115-day-old freemartin the seminiferous cords had diff e rentiated

and seminal vesicles were well branched, but the Wolffian ducts were
interrupt e d. This means that the level of androgens circulating in the
fetus was not very high between days 75 to 80. The reverse applies to
the 93- and 109-day-old fetuses in which the Wolffian ducts were con-
tinuous although the gonads contained only small testicular structures.
Table 2. Condition of the genital tract in 11 freemartins aged 77 to 125 days
Estimated age (days) 77 88 93 97 108 108 108 109 110 11 5 125

Size (rom) 106 154 175 186 226 228 228 223 223 254 292
Seminiferous cords 0 0 + +++ 0 0 + + 0 ++ ++
Interstitial cells 0 0 0 +++ 0 0 0 0 0 0 ++
Continuous Wolff. ducts yes no yes yes no no no yes no no yes
Epididymis + 0 0 +++ 0 + 0 ++ + + ++
Seminal vesicles:
- length 2.5 21 18 47 33 10 17 26 35 56 63
- branched no no yes yes yes yes yes yes yes yes yes
Uterine horns + vagina yes a
no E: no yes yes yes yes no no yes
Prostate 0 0 + ++ + 0 + + 0 + +++
Seminiferous cords, epididymis and prostate are scored according to the number respectively of cords
present in the gonads, of sections of the coiled Wollfian duct in the largest epididymal section and
of prostatic buds in serial sections.
Length of the seminal vesicles is expressed in 10 2 \lm.
aSmal1 remnants only.
9
402
In the four freemartins whose gonads had no masculine structure, the

Wolff ian ducts were interrupted but the seminal vesicles were branched
and stimulated.
E. Discussion and Interpretations
Earlier studies on sex differentiation have strongly supported the

view that the fetal testis produces two kinds of factors governing
differentiation of the genital tract, an androgenic stimulating hor-
mone the effects of which can be duplicated by androgens, and a Mul-
lerian inhibitor (Jost, 1953, 1965).
In vitro the calf fetal testis inhibits the rat Mullerian duct (Jost
et al., 1972; Josso, 1972; Jost, 1974) whereas testosterone and other
androgens do not (Jost and Bergerard, 1949; Josso, 1971); in vivo mater-
nally administered testosterone or other androgens masculinize the
genital tract of the female calf fetus without inhibiting the Mulle-
rian ducts (Jost et al., 1963; Jost, 1965). These facts strongly sup-
port the view that in the calf fetus also, different hormones inhibit
the Mullerian ducts and stimulate the masculine characters. The con-
dition of the genital tract in freemartins can easily be interpreted
along the same line.
1. During a first period extending between days 50 and 60 the genital

tract seems to be submitted exclusively to a Mullerian inhibiting ac-
tivity with no indication of any masculinizing activity. The proof
that the genital tract of the female fetus responds to the influence
of androgens during that period was given by the incipient masculini-
zation observed in fetuses whose mother was injected androgens and
which were killed on day 49, 56 (Jost, 1966) or 60 (Jost et al., 1973a).
It is striking that in freemartins during the same period, multipli-

cation of the germ cells ceases and growth of the gonad stops. The
concomitance and the quantitative correlation (Jost et al., 1972)
between Mullerian and gonadal inhibition suggests that the Mullerian
inhibiting factor is also acting on the presumptive ovary and germ
cells. But it does not induce any visible testicular organogenesis in
the gonadal anlage.
During the next month, between days 60 and 90, the freemartin fetus
still seems to be submitted to the Mullerian inhibiting factor since
ovarian or germinal inhibition persists and Mullerian retrogression
proceeds, at least in some fetuses. In others the Mullerian inhibitor
ceases rather early judging from the size of the Mullerian remnants.
In addition to effects on the Mullerian ducts some very slight mascu-
linizing influence was revealed by the appearance of small buds of
seminal vesicles on day 62, i.e. earlier than in normal females. Simi-
larly in one 77-day-old freemartin some coiling of the upper Wolffian
duct indicated the formation of an epididymis.
Some trend toward a masculinization of the external genitalia revealed

by an increase of the anogenital distance was seen only twice in our
whole collection: in one 60- and one 125-day-old animal. Both had
vascular connections with two male fetuses.
2. In a subsequent second part of prenatal life, after approximately

day 90, a definite masculinization of certain parts of the genital
tract in some freemartins sets in. At the developmental stage only
some parts of the genital tract remain susceptible to masculinization,
403
namely those parts of the Wolffian ducts which are still present, and
the gonads which have not yet differentiated either ovarian or testic-
ular structures. The external genitalia can no longer be virilized
(Jost et al., 1973a).
Several important questions remain unsolved: (a) How is the differen-

tiation of "seminiferous cords" and of interstitial cells triggered
in the freemartin gonad? (b) Why do the germ cells not survive in
these gonads? (c) Does the hormone that masculinizes the Wolffian de-
rivatives come from the testes of the male twin or from the gonads of
the freemartin itself or from both? Some comments can be made on these
questions:
a) Structures resembling seminiferous cords were not seen in freemar-

tins younger than 93 days. We do not know how long the differentiation
of an extreme degree of masculinization would take as seen in the 97-
day-old freemartin, or whether th~s kind of animal could have shown
seminiferous cords earlier than day 90. It is true that we had only
a limited number of freemartins between 70 and 90 days at our disposal
(one 77-, one 84- and one 88-day-old animal). Therefore we cannot ex-
clude the possibility that SOme masculine structures could differen-
tiate in the gonads shortly before day 90.
The condition of the sex ducts in freemartins gives indications as to

the hormonal condition prevailing in these animals between days 70
and 90. In normal females the Wolffian ducts become interrupted be-
tween days 75 and 80; it can be assumed that if these ducts are con-
tinuous in freemartins the level of androgens circulating at the age
of 75 days was already rather high. Similar calculations concerning
the Mullerian inhibitor are not as clear-cut since the retrogression
of these ducts can probably be somewhat delayed in freemartins, but
one can guess that in those animals in which the uterine horns and
vagina persist, the Mullerian inhibitor was probably not very active
after day 70.
In the ten freemartins studied between days 88 and 125, no clear cor-
relation appeared between the degree of inhibition of the posterior
part of the Mullerian ducts and the testicular differentiation of the
gonads. It would appear that the presence of an intense activity of
the l1ullerian inhibitor is not an absolute requirement at the moment
when cords differentiate (Table 2) .
On the other hand, concerning the masculinizing hormone, it should be

recalled that in the four freemartins with continuous Wolffian ducts
(androgenization before day 75), the gonads contain some testicular-
like cords; however there is one opposite case showing some "semini-
ferous cords" and interstitial cells on day 115, despite interrupted
Wolffian ducts.
It should also be remembered that no masculinization of the gonads

was seen in female calf fetuses submitted to maternally administered
androgens (Jost, 1965). These females had however not been exposed
to the Mullerian inhibiting hormone.
During normal testicular differentiation the interstitial cells dif-

ferentiate later than the seminiferous cords; in modified gonads of
freemartin fetuses under 250 days of age the same might well occur,
since seminiferous cords are more frequent than recognizable inter-
stitial cells. Could the seminiferous cords induce in one way or an-
other the differentiation of interstitial cells? (Kithara, 1923).
In any case one should distinguish two phases in the masculinization
of gonad, first the differentiation and aggregation of the Sertoli
404
cells which build up the seminiferous cords (Jost, 1972) and second
the differentiation of the interstitial cells.
b) The absence of germ cells in adult freemartins has long been recog-
nized (Willier, 1921; Ohno, 1969; Short et al., 1969). The freemartin
fetus described by Hay (1950) has long been considered exceptional for
the great number of germ cells present. The estimated age of this 25-
cm-Iong freemartin fetus is 115 days; at that stage the gonads of most
of the freemartins contain germ cells. These cells become more and
more scarce as the freemartin fetus becomes older and are rare after
day 150. The cause for their disappearance is not yet clear. Short
(1971) discussed the same problem in different situations.
c) So far no definite evidence permits an assessment of the origin of

the androgens which masculinize the Wolffian derivatives and prostate.
Two kinds of observations are relevant. On one hand in the three fe-
tuses in which the gonads contained the greatest number of seminiferous
cords and clusters of interstitial cells, the genital tract was also
strongly masculinized. But on the other hand the seminal vesicles,
epididymis and prostate may show androgenic stimulation in animals
whose gonads contain no recognizable testicular structure.
The available data could be understood by assuming that freemartins

are submitted to an extraneous androgen, and that at least in some
of them, their own modified gonads participate in the stimulation of
the genital tract. The reason why during the early period of prenatal
life only the MUllerian inhibiting hormone should pass from the male
to the freemartin, whereas later also androgens might pass is not
known. Quantitative differences in testicular activity, perhaps also
of pituitary activity, might be involved.
References
Beek, W.F. van: Die Entwicklung des Eierstockes beim Rinde. Z. Anat.
EntwickI.-Gesch. 11, 458-558 (1924).
Bissonnette, T.H.: The development of the reproductive ducts and ca-
nals in the freemartin with comparison of the normal. Amer. J. Anat.
TI, 267-345 (1924).
Bissonnette, T.H.: The "high flanker" testis in cattle, with its bear-
ings of the problem of the scrotum and on that of the freemartin
testis. Anat. Rec. 33,47-58 (1926).
Buyse, H.: A case of extreme sex modification in an adult bovine free-
martin. Anat. Rec. 66, 43 (1936).
Chapin, C.L.: A microscopic study of the reproductive system of foetal
freemartins. J. expo Zool. ~, 453-482 (1917).
Erickson, B.H.: Development and radio-response of the prenatal bovine
ovary. J. Reprod. Fert. 11, 97-105 (1966).
Fraser-Roberts, J.A., Greenwood, A.W.: An extreme freemartin and free-
martin-like condition in the sheep. J. Anat. 21, 87-94 (1928).
Gropp, A., Ohno, S.: The presence of a cornmon embryonic blastema for
ovarian and testicular parenchymal (follicular, interstitial and
tubular) cells in cattle, Bos taUf'us. Z. Zellforsch. li, 505-528
(1966) .
Hay, D.: Etude d'un freemartin spontane chez les ruminants. Arch. Anat.
Histol. Embryol. (Strasbourg) 33, 53-80 (1950).
Josso, N.: Effect of testosterone-and of some of its 17-hydroxylated
metabolites on the MUllerian duct of the foetal rat, in organ cul-
ture. Rev. europ. Etudes clin. bioI. l§., 694-697 (1971).
405
Josso, N.: Activite inhibitrice du testicule de foetus de veau sur

Ie canal de Mliller de foetus de rat, en culture organotypique:
role des tubes seminiferes. C.R. Acad. Sci. (Paris) 274, 3573-3576
(1972) .
Jost, A.: Problems of fetal endocrinology: the gonadal and hypophyseal
hormones. Rec. Progr. Horm. Res. 8, 379-418 (1953).
Jost, A.: Gonadal hormones in the sex differentiation of the mammalian
fetus. In: Organogenesis (eds. R.L. de Haan and H. Ursprung), pp.
611-628. New York: Holt, Rinehart and Winston 1965.
Jost, A.: Steroids and sex differentiation of the mammalian foetus.
Proc. 2nd into Congr. Hormonal Steroids, Hilan 1966. Excerpta 11e-
dica Int. Congr. Ser. 132, 74 (1966).
Jost, A.: Hormonal factors in the sex differentiation of the mammalian
foetus. In: Discussion on Determination of Sex. Phil. Trans. Roy.
Soc. Lond. B 259, 119-130 (1970).
Jost, A.: Donnees preliminaires sur les stades initiaux de la diffe-
renciation du testicule chez Ie rat. Arch. Anat. micro Morphol.
exp . ..§j, 415-438 (1972).
Jost, A.: Mechanisms of normal and abnormal sex differentiation in
the fetus. In: Birth Defects and Fetal Development; Endocrine and
11etabolic Factors (ed. K.S .. Moghissi). Springfield, USA: Ch.C.
Thomas 1974.
Jost, A., Bergerard, Y.: Culture in vitro d'ebauches du tractus genital
du foetus de rat. C.R. Soc. Bioi. (Paris) 143, 608-609 (1949).
Jost, A., Chodkiewicz, M., Mauleon, P.: Intersexualite du foetus de
veau produite par les androgenes. Comparaison entre l'hormone foe-
tale responsable du free-martinisme et l'hormone testiculaire adulte.
C.R. Acad. Sci. (Paris) 256, 274-276 (1963).
Jost, A., Vigier, B., Prepi~J.: Freemartins in cattle: the first
steps of sexudl organogenesis. J. Reprod. Fert. ~, 349-379 (1972).
Jost, A., Vigier, B., Prepin, J., Perchellet, J.P.: Studies on sex
differentiation in mammals. Rec. Progr. Horm. Res. 29, 1-41 (1973a)
Jost, A., Vigier, B., Prepin, J., Perchellet, J.P.: Le developpement
de la gonade des freemartins. Ann. Bioi. Anim. Bioch. Biophys. 21,
415-437 (1973b).
lingstrachtigkeit des Rindes. Untersuchungen liber die Entstehungs-
ursache der geschlechtlichen Unterentwicklung von weiblichen Zwil-
lingskalbern welche neben einem mannlichen Kalbe zur Entwicklung
gelangen. Wien. Tierartzl. Wschr. 3, 513-526 (1916).
Kieffer, N.M., Sorensen, A.M.: Some cytogenetic aspects of intersexu-
ality in the bovine. J. animo Sci. 32, 1219-1228 (1971).
Kithara, Y.: tiber die Entstehung der Zwischenzellen der Keimdrlisen
des Menschen und der Saugetiere und liber deren physiologische Be-
deutung. Arch. Entwickl.-I1echan. 52, 571 (1923).
Krehbiel, E.B.: Differentiation of gonads in the bovine embryo. Dis-
sertation Ph.D., Kansas State University, USA, 91 p. (1963).
Lillie, F.: The theory of the free-martin. Science 43, 611-613 (1916).
Lillie, F.: The free-martin, a study of the action of sex hormones
in fetal life of cattle. J. expo Zool. 23,371-451 (1917).
Lillie, F.R.: Supplementary notes on twins~n cattle. Bioi. Bull. 44,
47-48 (1923).
Lillie, F.R., Bascom, K.F.: An early stage of the freemartin and the
parallel history of interstitial cells. Science 22, 624-625 (1922)
Ohno, S.: The problem of the bovine freemartin. J. Reprod. Fert.
Suppl. 7, 53-61 (1969).
Rajakoski,-E., Hafez, E.S.E.: Derivatives of cortical cords in adult
free-martin gonads of bovine quintuplets. Anat. Rec. lil, 457 (1963)
Short, R.V.: Germ cell sex. In: Proc. Int. Symp. The Genetics of the
Spermatozoon (eds. R.A. Beatty and S. Gluecksohn-Waelsch), pp. 325
- 345. Edinburgh-New York 1971.
406
Short, R.V., Smith, J., Mann, T., Evans, E.P., Hallett, J., Fryer, A.,
Hamerton, J.L.: Cytogenetic and endocrine studies of a freemartin
heifer and its bull co-twin. Cytogenetic 8, 369-388 (1969).
Vigier, B., Prepin, J., Jost, A.: Absence de-correlation entre le
chimerisme XX/XY dans le foie et les premier signes de freemartinisme
chez le foetus de veau. Cytogenetics 11, 81-101 (1972).
Vigier, 0., Prepin, J., Jost, A.: Absence de chimerisme XXjXY dans
les tis sus somatiques des foetus de veau freemartins et jumeaux
males. Ann. Genet. li, 149-155 (1973).
Willier, B.H.: Structures and homologies of freemartin gonads. J. expo
Zool. 33, 601-611 (1921).
Witschi, ~: Range of the cortex-medulla antagonism in parabiotic
twins of Ranidae and Hylidae. J. expo Zool. ~, 113-145 (1931).
Witschi, E.: Modification of the development of sex in lower verte-
brates and in mammals. In: Sex and Internal Secretions (eds. E.
Allen, C. Danforth and E.A. Doisy), 2nd ed., Vol. 2, Chapt. IV,
pp. 145. Baltimore: Williams and Wilkins 1939.
Hokkaido Univ. Ser. VI, Zool. 13, 428-439 (1957).
Witschi, E.: Hormones and embryonic induction. Arch. Anat. micro
Morphol. expo 54, 601-611 (1965).
Witschi, E.: Biochemistry of sex differentiation in vertebrate embryos.
In: The Biochemistry of Animal Development (ed. R. Weber), Vol. II,
pp. 193-225. New York: Academic Press 1967.
Wolff, E., Ginglinger, A.: Sur la transformation des Poulets males en
intersexues par injection d'hormone femelle (folliculine) aux em-
bryons. Arch. Anat. Histol. Embryol. (Strasbourg) 20, 219-278 (1935)
The Role of Androgens in Sexual Differentiation of Mammals
F. Neumann, W. Eiger, H. Steinbeck, and K.-J. Grat
Numerous investigations have been undertaken by such pioneers, as

Dantchakoff, Wolff, Wiesner, Witschi, Burnes, Moore, Jost, and Bouin
and Ancel (1903), to clarify the mechanism of sexual differentiation
and it was discovered that steroid hormones, especially androgens,
play an important role, although some processes are still unknown.
As early as 1903 Bouin and Ancel expressed the belief that the gonads
of the male fetus influenced sexual differentiation. They postulated
this hypothesis after discovering that the testes of 30 mm long pig
fetuses contained well-developed interstitial cells. Further funda-
mental experiments were performed, which finally led to the present
theories about the hormonal regulation of sexual differentiation in
mammals (Jost, 1947 a-c). He proved the principle of "basic femaleness"
(postulated by Wiesner, 1934, 1935) by castrating rabbit fetuses in
utero. Irrespective of their gonadal sex, nearly complete female dif-
ferentiation occurred. That meant the somatic sex underwent female
differentiation even in genetically male fetuses.
Three essential conclusions were made:
1. In the fetal testes inductors are formed which cause male somatic
differentiation;
2. The ovaries or ovarian hormones do not playa part in sexual dif-
ferentiation under physiological conditions;
3. irrespective of the genetic sex there is a tendency to female so-
matic development, confirming Wiesner's hypothesis of the so-called
"basic femaleness".
In castration experiments, Jost was able to counteract the feminiza-

tion of male fetuses (to a very great extent, but not completely) by
substitution treatment with androgens. Some female structures always
remained, e.g. the Mullerian ducts, which normally later form the
Fallopian tubes~and the uteri. From this it was deduced that, at least
two factors formed in the fetal testes must be involved in somatic
sexual differentiation. These are: (1) androgens; (2) one or several
other factors which are responsible at least for the regression of
the Mullerian ducts.
Thus, irrespective of the genetic, chromosomal or gonadal sex, somatic

differentiation can be male or female. This is because primary struc-
tures for both male and female development are present and because the
differentiation takes place using identical primary structures. The
former is true of the internal genitalia, where structures are present
for both male and female. These are the so-called Wolffian and Mulle-
rian ducts.
Fig. 1 shows cross sections at the level of the gonads at about the
30th day of development in a genetic female and male dog fetus. At
408
this stage the differentiation of the internal genitalia is not com-

plete in this species, so that both duct systems are still present
in both the male and female fetuses. If the development follows the
normal pattern in accordance with the gonadal sex - the epididymides
deferent ducts and the seminal vesicles are formed from the Wolffian
ducts. When normal female development takes place these anlagen de-
generate. In females, the MUllerian ducts normally develop into the
Fallopian tubes and the uteri. All other structures of the genital
tract, including the external genitalia, develop primarily from the
same anlagen in both sexes.
Fig. 1 a and b. Internal genitalia of dog fetuses before differentia-

tion of the gonoducts (approximately day 30 of fetal development) .
(a) Normal female fetus. (b) Male fetus. C cortex ovarii, K kidney,
M medulla ovarii, MD MUllerian duct, PN mesonephros, T testis,
WD Wolffian duct. (x 22)
At the 17th day of embryonic development in rat fetuses, when the go-
nads are already differentiated, the structures of the external geni-
talia are still almost identical in males and females. There are no
differences at this stage of development, e.g. in the formation of
the sinus urogenitalis or the anogenital distance (Fig. 2).
As the experiments of Dantchakoff, Jost and others have shown, andro-

gens play an important role in the somatic differentiation of sexual
organs. When androgens are given to females during the "critical"
phase of differentiation in embryonic development it is possible to
induce a nearly complete male genital tract, in addition to female
genital structures such as the Fallopian tubes and the uteri (Fig. 2)
From these classical experiments it was then known what happened if

androgens became active in females and in males. However, it was still
unknown what happened in the male when no androgens became active but
the other factor(s), sometimes called the X-factor, became active?
The discovery of the highly effective antiandrogen, cyproterone acetate 1

(Hamada et al., 1963; Neumann and Hamada, 1964), provided a further
IThis compound was synthesized by Dr. Wiechert and his group in the
laboratories of SCHERING AG, Berlin-Bergkamen!Germany.
409
useful tool for investigating this problem. Several experiments have

been carried out using this substance in rats, dogs, mice, hamsters,
guinea pigs and sheep (Neumann and Steinbeck, in press). The results
obtained were not uniform with all the species investigated. There
were big differences, above all with respect to the differentiation
of the internal genitalia. Thus, in rats and mice which were exposed
to this anti androgen during the critical phase of differentiation,
the degeneration of the Wolffian ducts, which was expected, could
never be observed (Elger et al., 1973i Neumann et al., 1970a). The
form of intersexuality in male fetuses which was achieved in some of
the other species did, hOvlever, correspond to the theoretical concept.
This was true, for instance, in the dog (Steinbeck et al., 1970) and
in rabbits (Elger, 1966).
Epididymis _ _ _ _ _+ __......,
Gonad -----------4---+*~
TUbc-------------1---~
Ductus deferens -------1I-....l~~~
Uterus ------------1-------".......,;..,""\"-.
Vrina ry bladder -----t------------~,.:=:I"..;c,;!/.L:..:

Seminal vesicle -----i-----------, ........:-N
Coagulating g land ..-::;="""''''-
Prosbte-----------+--------~r_~~
Canalis urogeniblls ---/------------=---".1
Penis ------------- l--------------1}

Derived Crom the
- -l____"=
Wolff ian duel ----{]
Derived from thc ____
Mullerian duel
Derived from the ----+----1
urogenib l sinus
Androgens 1. Stabilization of the Wolffian ducts

active (Differentiation of epididymes,
ductuli deferentes, seminal vesicles)
2. Differentiation of the accessory
sexual glands deriving
from the urogenital sinus
(Prostate, coagulating glands,
bulbourethral glands)
3. Differentiation of male external
genitalia (Penis, scrotum)
4. No differentiation of a vagina
Factor X 1. Stabilization of the MUllerian ducts

not active (Differentiation of tubes and uterus) Fig. 2. Schematic
illustration of the
2. (No descent of the gonads?) internal genitalia
3. (Differentiation of ovaries?) of a virilized female
fetus
410
Fig. 3
411
In this paper, the dog has been used as an example to show how male
differentiation proceeds when the effect of androgens is blocked, but
all the possible other factors which contribute to sexual differentia-
tion are still present and active (Figs. 3 and 4) .
These figures present semidiagramatic reconstructions of a normal male

and a feminized genetically male fetus, whose mother had been treated
with cyproterone acetate. Below these reconstructions are histological
cross-sections of different levels of the genital tract of the same
animal. In normal males the mesogenitalia with the Wolffian duct (sec-
tion 1) lies between the testis and the remnants of the pronephros.
The Mullerian duct to the right has almost completely retrogressed in
this phase of embryonic development. More posteriorally are the bi-
lateral mesogenitalia with the Wolffian ducts to the left and right
of the rectum (section 2). After the mesogenitalia has become joined
(section 3), the Wolffian ducts approach the urethra (sections 4 and 5)
in the caudal part of the genital cord. They then open separately into
the sinus urogenitalis (sections 6 and 7). The development of prostatic
buds can be seen in sections 5 and 6.
In the region of the external genitalia, the sinus urogenitalis is

nearly completely surrounded by the glans lamella ring (section 8) .
In feminized male dog fetuses the mesogenitalia are empty in the re-
gion of the genital fold as well as in the genital cord. In some of
the sections shown, dense mesenchymal cell clusters can be seen clearly
at those points where the Wolffian ducts are normally located (see
sections 4 to 7). In a more caudal position an anlage of a vagina can
be seen (sections 8 to 10) joining the urethra to form the sinus uro-
genitalis further caudally (sections 11 to 13). As in female animals
the sinus urogenitalis is located excentrally and the glans lamella
ring is not closed, but sickle-shaped or like a horse shoe.
On the basis of Jost's experiments and our investigations, as well as

those of numerous other workers, it is now known fairly precisely which
of the processes of somatic sexual differentiation are controlled by
androgens and which are not.
The regression of the Mullerian ducts, the descent of the testes and,
of course, the differentiation of the gonads themselves are not con-
trolled by androgens. In Fig. 5 the four basic possibilities are sum-
marized assuming that, besides androgens, another factor, which is
formed in the gonads of male individuals, plays a part in sexual dif-
ferentiation. Diagrams are given for the situations where both factors
are active and where both, or only one, are not active.
Only one of these four possibilities corresponds to normal differen-

tiation, that is no. 3 for normal female and no. 4 for normal male
development. Cases 1 and 2 always represent a form of intersexuality.
Besides the effect of androgens on these structures, androgens have

other effects in the sexual differentiation. The mammary glands also
belong to the secondary sexual characteristics. In as far as differ-
ences occur between males and females in the development of the mam-
mary glands, they are cuased by androgens. The principle of the "basic
femaleness" is also true with regard to the differentiation of the
~ Fig. 3. Serial cross-sections through a normal male dog fetus, day 44

of fetal development. For description see text. WD Wolffian duct(s),
T testis, M mesogenitalia, R rectum, B bladder, PB prostate buds,
cu urogenital canal, LG lamina glandaris. (x 9)
412
Fig. 4. Serial cross sections through a "feminized" dog fetus, day 44

of fetal development. The mother was treated daily from day 23 to day
44 with 10 mg of cyproterone acetate per kilogram, s.c. For description
see text
mammary glands. Unless an androgenic influence is present during the

so-called critical phase of differentiation, feminine characteristics
develop.
There are three events on which hormones have an effect:
1. The development of the nipples.
2. The development of primary mammary gland buds.
3. Quantitative development of the mammary gland tissue or sensitivity
of the glandular tissue in adulthood to hormonal influences, for in-
stance progestogens and/or estrogens.
413
Fig. 4 (cant.). RN mesonephros, RWD rudiments of the Wolffian duct,

T testis, M mesogenitalia (without any duct system), R rectum, B blad-
der, U urethra (note the absence of prostate buds in this region) ,
V vagina, SU urogenital sinus, LG lamina glandaris. (x 9)
Thus, male rats and mice do not develop nipples, but females do. If,
in male rats and mice no androgens become active during the critical
phase of differentiation, even male rat fetuses develop nipples (Ray-
naud, 1961; Raynaud and Frilley, 1947).
In addition, in mice but not in rat fetuses, androgens cause the de-
struction of the primary glandular buds (Neumann and Elger, 1969;
Raynaud, 1961). This process of destruction does not take place when
androgens are not active, e.g. when the pregnant mother is treated
with cyproterone acetate (Neumann and Elger, 1966a). The reverse is
true of female mice fetuses in which destruction of the primary glan-
dular buds follows androgen treatment of the mothers.
414
In two other species, guinea-pigs and in man, wide species differences

occur, nipple anlagen and glandular buds are not influenced by andro-
gens. This means that these structures do not show any morphological
difference between males and females at the fetal stage. This applies
also to various other species, e.g. to cattle, pigs, dogs and cats,
where nipples are normally developed not only in females but also in
males. Nevertheless, stimulation of mammary glands in adulthood with
estrogens or progesterone is more effective in females than in males,
as far as the tubulo-alveolar growth is concerned.
IndiHerent
stage
WoUfian duct
Miillerian duct
o Factor X active, androgcns not ac tive

e Androgens active, factor X not active Fig. 5. Principal
o Both components not active possibilities for the
o Both components active development of the
internal genitalia
415
We had supposed at first that under the influence of androgens in the

fetal stages, less glandular tissue would be found so that in adult-
hood there would be simply less substrate to react to further adequate
hormonal stimuli (Neumann and Elger, 1967). This hypothesis was found
to be unsatisfactory because there were no differences in the amount
of mammary tissue in female and in male rat. fetuses, neither at the
time of birth nor up to the 30th day of age (i.e. before puberty) ,
(Dlisterberg, 1973; Rieser et al., 1970).
In man too, mammary gland development does not differ in males or fe-
males, at least until puberty (see Figs. 6 and 7) .
"'C • ; ••• ;
.~. "'. -:#'1
Fig. 6 a and b. Mammary gland anlagen of human fetuses (crown-rump

length 11 cm).·(a) Female fetus. (b) Male fetus. (x 126)
416
Fig. 7 a and b. Mammary gland anlagen of newborn children. (a) Female.

(b) Male
Besides the influence of androgens during fetal and early postnatal

stages, other hormones influence the development of the mammary gland,
especially during later stages. One of these hormones, especially im-
portant for the tubulo-alveolar growth of the mammary glands in rodents,
is prolactin. A positive feedback mechanism exists between estrogens
and prolactin secretion. At , least in rodents, the mechanism of action
of estrogens takes place at two levels.
Firstly, it seems to occur at the hypothalamic level by inhibiting the

prolactin-inhibiting-factor (PIF) secretion or synthesis and secondly,
at the level of the pituitary, by direct stimulation of the synthesis
and/or secretion of prolactin.
One of the hypotheses which has been formed is based on the assumption
that in fundamental masculinity (when androgens became active in a
certain phase of differentiation during fetal development) this mecha-
nism is less sensitive than in fundamental feminity. This means that
with the same stimulus the effect of prolactin production is less in
males or masculinized females, than in females.
There is some evidence for this assumption. Thus, in spayed female rats
treated with estrogens the hypophyseal weight became greater in com-
parison with males. The weight increase of the pituitary was caused by
an increase of those cells which secrete prolactin (Fig. 8).
Observations by other authors (Hayashi, 1967a,b, 1969; Kurcz et al.,

1967) also seem to confirm this hypothesis. It has not been possible
417
up to the present time, to prove that the secretion of prolactin in

females is higher in comparison to males after the same stimulus, and
furthermore it is also not known what prolactin in males really does.
Pituitary weights
Treatment:
DE = estradiol
HE = hypophysectomy
FEM= pre· and postnatally feminized
K untreated controls
mg pituitary/1 00 g B.W.
5
Fig. 8. DNA-content of
4 mammary glands of
adult rats after daily
treatment with 10 ~g
3 estradiol for 21 days.
The male and female
2 animals of all 5
groups have been cas-
trated in adulthood
before treatment.
OE estradiol treat-
1. 2. 3. 4. 5. groups ment, FEM cyproterone
acetate treated pre-
d ~ d ~ d and postnatally (femi-
nized), K controls, no
K K DE DE FEM treatment after cas-
DE tration
A second possibility for the sex specific differences in the formation

of mammary glands is that of a sex-specific difference in the respon-
siveness of the mammary gland tissue itself. It is possible, due to
this difference in responsiveness, that the female mammary gland struc-
tures have a more sensitive growth reaction to an adequate hormonal
stimulus. One possibility does not exclude the other and, on the basis
of the facts we know, other explanations or interpretations might be
possible.
In addition, androgens act also in the differentiation of neural cen-

ters which regulate. the mode of gonadal secretion and certain kinds
of behavior according to sex. This is valid, above all for sexual be-
havior (for review see Barraclough, 1967; Davidson and Levine, 1972;
Dorner, 1972; Gorski, 1971; Goy, 1970i Goy and Phoenix, 1971; Harris,
1964; Neumann et al., 1970b).
418
The characterization of neural centers always takes place subsequent

to the somatic differentiation. It is thus possible to cause psychic
intersexes by relevant hormonal manipulation. As early as 1938, Dant-
chakoff observed that the female offspring of guinea-pigs, which had
Diagrammatic representation of events at the hypothalamic,

pituitary, and gonadal level, which result in steroid secretion
and ovulation in normal females.
CYCLIC TONIC
Preoptic-
suprachiasmatic
area
Ovum
Diagrammatic representation of events at the hypothalamic, pituitary,

and gonadal level in normal males or in females after destruction
of the preoptic area ur neonatal androgen treatmen~.
TONIC
....
Higher Preoptic-
neu tral I suprachiasmatic ...
control area
E E
No ova Sperm
[modif. from c.A. Barraclough,

Rec. Progr. Hormone Res. 22 (1966) 503]
Fig. 9. Schematic illustration of the correlations between neural cen-

ters and the gonads
419
been given testosterone during pregnancy, showed masculine sexual be-

havior after treatment with androgens in adulthood. Two years before
this, Pfeiffer (1936) had made an interesting observation in rats.
He implanted testes in newborn female rats. ,Vhen the animals were ma-
ture he discovered that they did not have cycles, but were permanently
in estrus and did not ovulate. The ovaries were small and pale. In
histological sections, follicles could be observed, but no corpora
lutea. It was concluded from these investigations that the hormones
from the implanted testes caused characteristic masculine functioning
of the pituitary. All these conclusions are now considered to be cor-
rect, except that it is not the pituitary which is subjected to sex
specific funtional differentiation, but hypothalamic centers. It has
since been demonstrated, in numerous investigations, that a single
dose of 10 ~g or even 5 ~g of testosterone propionate given to newborn
female rats, is sufficient to cause the sterility syndrome of permanent
estrus without ovulation, which has first observed by Pfeiffer. This
syndrome is known as the "early androgen syndrome".
It is now assumed that there are two hypothalamic centers which are
responsible for the secretion of LH/ICSH. One center is situated in
the median hypothalamus and is responsible for the tonic formation of
LH. It is sufficient in male individuals to maintain the secretory and
generative functions of the testes. Apart from this, there is a yet
higher center in the pre-optic region which induces the cyclic secre-
tion of LH (Fig. 9).
It seems quite likely that differentiation of the cyclic center is

inhibited when androgens become active in a certain phase of differen-
tiation.
It is possible to induce this situation in mature females by destroying

the pre-optic center using sterotactic methods. This manipulation in
males has no influence on hormonal biosynthesis and on spermatogenesis
in the testes, at least in rats.
In contrast, female characterization of these centers in male animals

can be induced by preventing androgens being active in the "critical"
phase, which in rats and mice is the first few days after birth and
in guinea-pigs during the 30th to the 40th day of fetal development.
In our own investigations we were able to induce this situation by
using the antiandrogen cyproterone acetate. If one castrates male rats
in adulthood (which have been treated with cyproterone acetate during
the first days of life), and implants ovaries after castration, then
hypophyseal LH is secreted cyclically, as in female animals. These
animals also ovulate and show female sexual behavior (Neumann and
Elger, 1966b).
It is considered that these experiments with antiandrogens have added

significantly to our knowledge of the role of androgens in the sexual
differentiation of a number of animal species. The earlier results
have been confirmed and a greater understanding of the whole process
is obtained. However, a number of regulatory mechanisms still remain
unexplained.
420
References
Barraclough, C.A.: Modifications in reproductive function after expo-

sure to hormones during the prenatal and early postnatal period.
In: Neuroendocrinology (eds. L. Martini and W.F. Ganong), pp. 62-99.
Bouin, P., Ancel, P.: Sur la signification de la glande interstitielle
du testicule embryonnaire. C.R. Soc. Biol. (Paris) 55, 1682 (1903).
Dantchakoff, V.: Role des hormones dans la manifestation des instincts
sexuels. C.R. Aoad Sci. (Paris) 206, 945-947 (1938).
Dantchakoff, V.: Der Aufbau des Geschlechts beim hoheren Wirbeltier.
Jena: Gustav Fischer 1941.
Davidson, J.M., Levine, S.: Endocrine regulation of behavior. Ann. Rev.
Physiol. li, 375-408 (1972).
Dorner, G. (ed.): Sexualhormonabhangige Gehirndifferenzierung und
Sexualitat. Wien: Springer 1972.
Dtisterberg, B.: untersuchungen an Meerschweinchen tiber die Empfindlich-
keit neuraler Zentren, die die Prolactinsekretion regulieren.
Diplomarbeit - Freie Universitat Berlin, 1973.
Elger, W.: Die Rolle der fetalen Androgene in der Sexualdifferenzierung
des Kaninchens und ihre Abgrenzung gegen andere hormonale und soma-
tische Faktoren durch Anwendung eines starken Antiandrogens. Arch.
Anat. micro Morph. expo 55, 657-743 (1,966).
Elger, W., Graf, K.-J., Steinbeck, H., Neumann, F.: Hormonal control
of sexual development. 3rd European Congress of Anatomy, Manchester,
September 1973. Advances in the Biosciences. Oxford-Braunschweig:
Pergamon Press - Vieweg (in press) •
Goy, R.W.: Experimental control of psychosexuality. Phil. Trans. Roy
Soc. B. 259, 149-162 (1970).
Goy, R.W., Phoenix, C.H.: The effects of testosterone propionate ad-
ministered before birth on the development of behavior in genetic
female rhesus monkeys. In: Steroid Hormones and Brain Function
(eds. C.H. Sawyer and R.A. Gorski), pp. 193-202. Los Angeles: Uni-
University of California Press. U.C.L.A. Forum Med. Sci. no. 15,
1971.
Hamada, H., Neumann, F., Junkmann, K.: Intrauterine antimaskuline
Beeinflussung von Rattenfeten durch ein stark gestagen wirkendes
Steroid. Acta endocr. (Kbh.) 44, 380-388 (1963).
Harris, G.W.: Sex hormones, brain development and brain function. The
Upjohn lecture of the Endocrine Society. Endocrinology 75, 627-648
(1964). -
Hayashi, S.: Difference in responsiveness to estrogen of the anterior
hypophysis between normal and neonatally estrogenized rats. J. Fac.
Sci. (Tokyo) 11, 227-234 (1967a).
Hayashi, S.: Difference in luteinization in intrasplenic ovarian grafts
between normal and persistent-estrous rats following ovariectomy.
J. Fac. Sci. (Tokyo) 11, 235-242 (1 967b) .
Hayashi, S.: suppression~y estrogen injections of luteinization in
intrasplenic ovarian grafts in ovariectomized "normal" and neona-
tally estrogenized rats. Annot. zool. jap. 42, 12-20 (1969).
Jost, A.: Recherches sur la differenciation sexuelle de l'embryon de
Lapin. I. Introduction et embryologie genitale normale. Arch. Anat.
micro Morph. expo 36, 151-200 (1947a).
Lapin. II. Action des androgenes de synthese sur l'histogenese
genitale. Arch. Anat. micro Morph. expo 36, 242-270 (1947b).
Lapin. III. Role des gonades foetales dans la differenciation sexu-
elle somatique. Arch. Anat. micro Morph. expo 36, 271-315 (1947c).
Kurcz, M., Kovacs, K., Tiboldi, T., Orosz, A.: Effect of androgenisa-
tion on adenohypophyseal prolactin content in rats. Acta endocr.
(Kbh.) 54, 663-667 (1967).
421
Neumann, F., Elger, W.: The effect of the anti-androgen 1,2u-methylene-

6-chloro-A4,6-pregnadiene-17u-ol-3,20-dione-17u-acetate (cyproterone
acetate) on the development of the mammary glands of male foetal
rats. J. Endocr. 1£, 347-352 (1966a).
Neumann, F., Elger, W.: Permanent changes.in gonadal function and
sexual behavior as a result of eraly feminization of male rats by
treatment with an antiandrogenic steroid. Endokrinologie 50, 209-
225 (1966b). -
Neumann, R., Elger, W.: Steroidal stimulation of mammary glands in
prenatally feminized male rats. Europ. J. Phamacol. 1, 120-123 (1967)
Neumann, R., Elger, W.: Der EinfluB von Antiandrogenen auf Differen-
zierungsvorgange. I. Schering-Symposium on Endocrinology: Advances
in the Biosciences, Berlin 1967, pp. 80-107. Oxford: Pergamon Press
1969.
Neumann, F., Elger, W., Steinbeck, H.: Antiandrogens and reproductive
development. Phil. Trans. Roy. Soc., Lond. B 259, 179-184 (1970).
Neumann, F., Hamada, H.: Intrauterine Feminisierung mannlicher Ratten-
feten durch das stark gestagen wirksame 6-chlor-A6-1,2-methylen-
17u-hydroxy-progesteronacetat. 10. Symp. Dtsch. Ges. f. Endokrino-
logie, Wien 1963. Berlin~Gottingen-Heidelberg-New York: Springer
1964.
Neumann, F., Steinbeck, H.: Antiandrogens. In: Handbuch der experimen-
tellen Pharmakologie (eds. O. Eichler, A. Farah, H. Herken and A.D.
Welch). Berlin-Heidelberg-New York: Springer (in press) •
Neumann, F., Steinbeck, H., Hahn, J.D.: Hormones and brain differen-
tiation. Workshop Conference on Integration of Endocrine and Non-
Endocrine Mechanisms in the Hypothalamus, Stresa 1969. In: The
Hypothalamus (eds. L. Martini, M. Motta and F. Fraschini), pp. 569-
603. New York-London: Academic Press 1970.
Pfeiffer, C.A.: Sexual differences of the hypophyses and their deter-
mination by the gonads. Amer. J. Anat. 58, 195-221 (1936).
Raynaud, A.: Morphogenesis of the mammary gland. In: Milk - the Mammary
Gland and its Secretion (eds. S.K. Kon and A.T. Cowie), vol. 1,
Raynaud, A., Frilley, M.: Destruction des glandes genitales de l'em-
bryon de souris par une irradation au moyen des rayons X, a l'age
de 13 jours. Ann. Endocr. (Paris) 8, 400 (1947).
Rieser, U., Schulz, U., Neumann, F.: Ursachen der geschlechtsunter-
schiedlichen Milchdrlisengewebsentwicklung bei Ratten. Experientia
(Basel) 26, 1148-1149 (1970).
Steinbeck, H:, Neumann, F., Elger, W.: Effect of an antiandrogen on
the differentiation of the internal genital organ in dogs. J. Reprod.
Fertil. 23, 223-227 (1970).
Wiesner, B.~: The postnatal development of the genital organs in the
albino rat - with a discussion of a new theory of sexual differen-
tiation. J. Obstet. Gynaec. Brit. Emp. 41, 867-922 (1934).
Wiesner, B.P.: The postnatal development o~the genital organs in the
albino rat with a discussion of a new theory of sexual differentia-
tion. VI. Effects of sex hormones in the heteronomous sex. J. Obstet.
Gynaec. Brit. Emp. ~, 8-78 (1935).
Recent Studies on the Intersexual Programming of the
Genetic Rat Male Pseudohermaphrodite *
A. S. Goldman
During the last few years our development of highly specific and potent
inhibitors of testosterone and corticosterone biosynthetic enzymes and
more recently, our introduction of the use of antibodies to testoster-
one, and antibodies to LH:FSH, have provided us with highly selective
experimental agents for hormone deprivation or interference directly
within the mammalian fetus (Goldman, 1974b). These agents have produced
reversible chemical castration of the fetus by ablation of testosterone
at the level of its synthesis, circulation, or uptake. From these stu-
dies, intersexes have been pronuced in rats, which are exact copies of
those occurring naturally in the human genetic defects of congenital
adrenal hyperplasia and testicular feminization. These diseases are
examples of testosterone deprivation of the fetus and neonate achieved
by g.enetic castration, at the level of testosterone synthesis (in the
case of certain forms of congenital adrenal hyperplasia) or uptake,
in the case of testicular feminization. The use of these experimental
and genetic testosterone depriving agents, taken together with the
classical work of Jost (1972) and Neumann et al. (1970), has consider-
ably substantiated the hypothesis that testosterone is the organizer
of mammalian androgen-dependent differentiation.
Review in depth of all of these experimental studies is not possible

here, but they have been presented in detail elsewhere (Bongiovanni
et al., 1967; Goldman and Bongiovanni, 1967; Goldman, 1970, 1972,
1974a,b). I have examined at several levels of biologic organization,
sex differentiation in animals rendered testosterone unresponsive dur-
ing critical periods of masculine differentiation, either genetically
or experimentally. The genetic animals are the Stanley-Gumbreck genetic
male rat pseudohermaphrodites while the experimental male pseudoher-
maphrodites have been produced by Dr. Neumann (Schering, Berlin) by
prenatal and prepubertal treatment with cyproterone acetate.
A. Genotype and Phenotype
The hereditary form of testicular feminization which occurs in the rat

has been described in detail by Stanley et al. (1973). This animal is
a genetic male with an XY karyotype and a chromatin negative nuclear
sex (Table 1). Consistent with their XY karyotype these rats have in-
guinal testes with well-developed Leydig cells and no Mullerian duct
derivatives. However, the testicular tubules demonstrate germinal ar-
rest and no male reproductive structures of Holffian duct origin exist.
Consistent with a female phenotype are the animal's short vagina end-
ing in a blind cuI de sac, and a feminine number and positioning of
nipples. Furthermore, the pseudohermaphrodite has a feminine-quanti-
*This work was supported by research grants from the U.S.P.H.S. (HD-
4683), the Rockefeller Foundation and the NATO Research Programme.
423
tative pattern of urinary corticosteroid excretion and exhibits no

sexual behavior. The pseudohermaphrodite also has a feminine-qualita-
tive pattern of urinary steroidal metabolites of adrenal and gonadal
origin (Gustafsson and Goldman, 1972; Goldman et al., 1973) due to a
phenotypically female pattern of hepatic steroid metabilizing enzymes
(Einarsson et al., 1972; Gustafsson et al., 1974).
Table 1. Sex of the pseudohermaphrodite
Level of organization Sex

Female Pseudo Male
Karyotype xx Xy Xy
Nuclear chromatin positive negative negative
Organs
Wolffian ducts absent absent present
Mullerian ducts present absent absent
Nipples and nipple lines present present absent
Vagina present present absent
Physiology
Behavior prefers prefers prefers
saccharin saccharin water
Urinary corticosteroids F F M
Biochemistry - hepatic enzymes
7a-hydroxylase F F M
16a- hydroxylase F F M
3a-hydroxysteroid dehydrogenase F F M
3S-hydroxysteroid dehydrogenase F F M
Sa-reductase F F M
Corticosterone sulfurylase F F M
This rat has feminine differentiation of that part of the brain regu-
lating taste preference, in that it prefers sweet solutions, e.g. 1%
saccharin, as does its littermate female, in contradistinction to the
preference of water by its littermate males (Shapiro and Goldman, 1973)
We have attributed the development of these phenotypically female
characteristics in the pseudohermaphrodite to its genetically absent
androgen-dependent differentiation. The primary genetic defect of the
male rat pseudohermaphrodite is believed, by most investigators, to be
apparently due to its inherited end organ insensitivity to androgens,
reflected at the cellular level by deficient target organ androgen
binding proteins (Ritzen et al., 1972; Bardin et al., 1973; Milin and
Roy, 1973).
In certain respects this rat may properly be classified as a phenotypic

intersex. The weights of its body, thyroid and preputial glands, and
kidneys are intermediate between those of its normal male and female
littermates. Its growth rate is also intermediate between that of the
two normal sexes (Fig. 1). When relative gland weights are calculated
on the basis of body weight, we find that the relative weights of the
pituitary, adrenal and preputial glands of the pseudohermaphrodite
become phenotypically female (Fig. 2). However, the relative weight
424
. .~·I·
420 ..
.'
aa.···
360 ~
t
11
MALE ........
...• PSEUDO ............... .
t. "'f. ' ..,. . . . ..... .. . . . . . . .

11 ..... •...... FEMALE
~.
:-- ................... .
BODY 11 •• ' +.····.t···· 11
HEIGHT leO f
G. .{.'f~""
t'·.l··· t
..:i··· + VAG I NA OPENS IN PSEUDOS
.:;':"1 VAG I NA OPENS IN FEMALES
60f
~;":
1.. ....
.,. ... ,... ;·0·.. ·,·· 4·0.. ·.. ·· ·60·· .. ··· s·o.. ·.. · 'i'o'o '" "lio'" "i',j 0·.... i'60··· .. i's'o'!' '2 O
!O··'
.. 1 .1 , I I I I I , I I , I I I , It
ONE STANDARD AGE (DAYS)

DEVIA TI ON
t P<,05 VERSUS PSEUDOS 11 : P<,Ol /I /I
Fig. 1. Intersexual growth curve of the pseudohermaphrodite as compared

to littermate male and female. Significant differences appear in three
sexes after onset of puberty which is significantly d e layed in pseudo-
hermaphrodites, as compared to females, and measured by time of vaginal
opening
PnUllARY THYROI AOREIIAL PREPUTIAL KIDNEY

I1GI 00 IIGl100 IIGI100
GlAIlDS
Gil .'~. GM B.W. Gil B. W.
....I...
,I
: :
: i
: :
~ i
! f
P 1\ M M
I: ± ON( STAN AR ("IATI r FE~1ALE
P VERSUS '~ LE OR FEI1ALE .i . . ... .. - •• ~
to ..... 1- . .: ..
PSEUOO
.; . .... 01- ... 1 ,~ MALE
P VERSUS PSEUOO Qo • • .. .(: - . ",1)
i • . • . ' • •2
'Y " ':":'1-. "' _
Fig. 2. Re lative weights (in mg/l00 g B.W.) of pituitary, thyroid,

adrenal, preputial glands, and kidneys of pseudohermaphrodite and lit-
termate females and males. Relative weights of the pituitary, adrenals,
and preputial glands of the ps e udohermaphrodites are phe notypically
female, but those of the thyroid are phenotypically male. Note Greek
and Roman letter abbreviations for statistical probability for this
and Figs. 4, 5 and 8
425
of the thyroid glands of the pseudohermaphrodite is significantly dif-

ferent from that of the female but is almost identical to that of the
male. Thus, the weights of the pituitary, adrenal and preputial glands
are probably phenotypically female, and the intermediate absolute
weights are most likely a reflection of the intersexual body weight
and growth of the animal. The thyroid relative weights should be clas-
sified as male.
B. Female Preputial Gland Biochemical Sex
The preputial glands are bilateral apocrine glands at the terminal por-
tion of the urinary tract in both male and female rats. These glands
are androgen-responsive but not androgen-dependent, in that they will
enlarge in both sexes under the influence of androgens, but do not
require androgens for maintenance of their weight. As mentioned above,
the phenotypic sex of this rat's preputial gland size is most probably
female.
c. Preputial Gland Androgen Metabolism
We became interested in the possibility of sex differences in androgen

metabolism by the preputial glands after the report of Bottiglioni et
al. (1971), that the genital skin of adult male offspring "feminized"
by in utero treatment of pregnant rats with cyproterone acetate had a
deficient aciivity of steroid Sa-reductase as measured by a diminished
conversion of testosterone to Sa-dihydrotestosterone. Our first ap-
proach to the problem of sex differentiation of target organ androgen
metabolizing enzymes was to study changes in activity of genital skin
Sa-reductase in the term rat fetus produced by maternal treatment with
cyproterone acetate or cyanoketone, the specific inhibitor of 3S-hy-
droxysteroid dehydrogenase, ~5-4-3-ketosteroid isomerase in the fetal
testis. This inhibitor produces an exact copy of the male-pseudoherm-
aphrodite which occurs in human congenital adrenal hyperplasia due to
a genetic deficiency in the dehydrogenase-isomerase enzyme system
(Goldman, 1970). The genital skin contains the anlagen of the preputial
glands. We were able to confirm the finding of Bottiglioni et al.
(1971) that the genitalia of male fetuses experimentally "feminized"
by treatment with either specific androgen depriving agent had a sig-
nificantly lower level of activity of steroid-Sa-reductase than the
corresponding male control tissues (Goldman, 1973). We had determined
that this was not a direct inhibition of the enzyme in vivo by these
agents, but rather an immediate consequence of the blocking of mascu-
line differentiation which they produced in the fetus and which per-
sists to adult life.
We then sought possible sex differences in the metabolism of testos-

terone and androstenedione by the preputial glands of the rat pseudo-
hermaphrodite, as compared to that of its littermate male and female
(Goldman and Klingele, 1974b). The metabolism of androstenedione and
testosterone by the preputial glands involves conversion of the sub-
strates through the activity of four enzymes (Fig. 3). The enzymes
include steroid-Sa-reductase which catalyzes irreversibly the reduc-
tion of the ~4 bond of androstenedione and testosterone to the Sa-
hydrogen. The 3-ketone is converted either to the 3a-hydroxyl group
by 3a-hydroxysteroid dehydrogenase, or to the corresponding 3S-hydroxyl
group by 3S-hydroxysteroid dehydrogenase. The 17-ketone and 17S-hydroxyl
426
groups are interconverted by 17B-hydroxysteroid dehydrogenase. We also

looked at the metabolism of androstenedione and testosterone by pre-
putial glands of 6 months old male offspring whose mothers had been
treated throughout gestation and the first 3 weeks of life with cypro-
terone acetate by Dr. Neumann and co-workers.
.'- ..", . .Y\\>
•("
.
.........-.! ......
. 1
<.o
"
;
.,..............J......... . .
/y(J" ,J... .. . . ... ;
HO .. ··l.... . .......
........ . i ....f.. i
...... .
; i :. 0
(' "'r" ·····1······ _ ··__ _ _t> ." ..............,' ,: ...... ; .........,,. ........:.........:/ 3.J-HS AilDROSTEROilE ''''. i .:.~
~/' ·· ·r···· . >.
~. l.............)
.;..., ... 5"-1\[ UCTA
O-;:::::...· .............. if' . . ..... ... )
- ---it>
.; ., .
ANDRO TE ED lONE 5"-1\: DR TI,'IE I IE: 3B-HSD [> ( ............ ( .... .., ... ( ............ /
HO/ ·····.._.····~;···'v·······
118-HS EP IAI DROSTEROIIL
(.,.j/r:;J • .·_·._:.'_:__ ,/(,j;': ~., ~,~:; :;~:~DJ: ;l:ir: ';~

o (,..
o~< ~·~· . . . . .,. . .) 5 ~RE UCTASE ~.••...../ ....~ .......•...•.) --_-e>- ( ............ +. . . ...- -:. .... . . . . .
TESTOSTERQ;IE DIHYDROTESTOSTERO: E 3s-HS HQ..- '··............ t . . . . /)
5.-AHDROSTA1I[ -36-17e- 0IOL
Fig. 3·. metabolism of androstenedione and testosterone by preputial

glands of the three sexes. This metabolism involves activity of pri-
marily three enzymes: Steroid 5~-reductase, 3~-hydroxysteroid dehydro-
genase (3~-HSD), 3B-hydroxysteroid dehydrogenase (3B-HSD), and 17B-
hydroxysteroid dehydrogenase (17B-HSD)
Wi~h androstenedione as a substrate (Fig. 4) we found a significantly

greater amount of 5~-androstane-3,17-dione made by preputial glands
of the male than that made by those of the pseudohermaphrodite and
female, while the glands of both the pseudohermaphrodite and female
made significantly greater amounts of the 3~-hydroxylated product,
androsterone. In other words, there is a sex-dependent difference in
the pattern of metabolites formed by the preputial glands. The pattern
of metabolites produced by the glands of the pseudohermaphrodite sig-
nificantly differed from that of its littermate male, and was nearly
the same as its littermate female. The difference in pattern consists
primarily of an increased level of activity of 3~-hydroxysteroid de-
hydrogenase in the glands of the female and of the pseudohermaphrodite.
Virtually identical results are obtained in the metabolites produced
by the experimental male pseudohermaphrodites produced by prenatal
and neonatal treatment with cyproterone acetate.
427
S. -ANDROSTANED l ONE
;C
ANDROSTERONE
%
SUBSTRATE
M P M P" P II
Fig. 4. Primary products of preputial gland metabolism of androstene-

dione by genetic male pseudohermaphrodites (left panel) and by experi-
mental male pseudohermaphrodites (right panel), produced by daily in-
jections of cyproterone acetate to the mother at 30 mg/kg from day 12
to term and postnatally at 10 mg/kg to male rat offspring from day one
to three weeks of age. Metabolic studies were performed when male off-
spring were 6 months of age_ Note phenotypically female pattern of
products produced by maternal and experimental pseudoherrnaphrodites
With testosterone as a substrate (Fig. 5) the genetic male pseudoher-

maphrodite has similar levels of dihydrotestosterone, 5a-androstane-
3a,17S-diol and androsterone as those produced by females, and these
levels are significantly different from those produced by the litter-
mate males. The metabolism of testosterone by preputial glands of cy-
proterone acetate-induced pseudohermaphrodites has the same pattern
with respect to dihydrotestosterone and 5a-androstane,3a-17S-diol as
does that of the female. Thus, it seems reasonable to conclude that
the preputial glands of the genetic and experimentally induced male
pseudohermaphrodi tes are programmed in utero as a female on at least two
levels of biological organization, namely, in growth and in the sex-
dependent differentiation in the level of 3a-hydroxysteroid dehydro-
genase.
GEl! TIC ~.ALE PSEUDOIlERl'JlPHROD ITE EXPERIKEI TAL MAL[ PSEUDOHERMAPHRODITE

01 HY ROTESTOSTEROllE DIHYDROIESTOSTERO E
5.-1\1 ROSTANE-3-.l7s-DIOL So-AIIDROSTANE-.l .l7s-D IOL
y- AIIDROSTEROtlE
It
...1....
%
,...I..., ....1...
! ;
Ii
UBSIRATE
i: ~
ro
~ ~:
i
Sa
i
l······!
,~
:
i, ,. L.: rL'lb1
.
:,, ,, i ~ ~ ! i y~ ~
:
:
!
,,j
•
'::, •i,.
! l
i : ....... .
i! i I, t, :
: : -'_~-',:.-.: I: ! , , : ~
F M P M F M P M
Fig. 5. Primary products of preputial gland metabolism of testosterone

by genetic (left panel) and experimental (right panel) pseudoherrnaphro-
dites under the same conditions as Fig. 4. Note phenotypically female
pattern of products by both types of pseudoherrnaphrodites
428
D. Isoelectrofocusing of Preputial Gland Receptor Proteins
It is generally believed that androgen action in the adult animals is

mediated in ventral prostate and other androgen-dependent target organs
6f the rat by the Sa-reduced metabolite of testosterone, Sa-dihydro-
testosterone by way of the cytosol dihydrotestosterone receptor pro-
teins. The uptake of these dihydrotestosterone-receptor complexes by
nuclear chromatin has been assumed to be the event which initiates
nuclear responses to androgen in the adult animal. It is conceivable
that the irreversible events produced by testosterone or its Sa-reduced
metabolite in programming male sexual differentiation during critical
developmental periods have the same molecular chain of events. Thus,
the complex of androgen-receptor protein after incorporation into
target organ nuclei would cause DNA and RNA directed synthesis of male
specific proteins.
The demonstration that the preputial gland cytosol of the pseudoher-

maphrodite lacks significant dihydrotestosterone receptor proteins have
been made mostly by ultracentrifugal techniques (Ritzen et al., 1972;
Bardin et al., 1973; Milin and Roy, 1973). We became interested in
looking for sex differences in the presence of the nuclear chromatin
uptake of dihydrotestosterone receptor proteins in the preputial glands
of the pseudohermaphrodite, by the micromethod of isoelectrofocusing
which we have deve],oped to study the receptor proteins in .the fetus
(Katsurnata and Goldman, 1974a), Isoelectrofocusing is one of the best
ways to separate proteins, because this method can resolve finer dif-
ferences in isoelectric points than other methods due to the fact that
diffusion artefacts (size, shape, charge) are eliminated. The receptor
proteins studied.primarily in the rat prostate appear to be acidic and
recently, Mainwaring and Irving (1973) applied electrofocusing to the
8s binding proteins of the rat prostate which have been partially puri-
fied by DNA-cellulose chromatography. They found a single peak of di-
hydrotestosterone labelled protein at an isoelectric point (pI) of
5.8 in the cytosol. We have reported the separation of multiple dihydro-
testosterone receptors in rat ventral prostate cytosol and nuclei with
the use of this micromethod of isoelectrofocusing (Katsumata and Gold-
man, 1974). Four dihydrotestosterone binding proteins fulfill the cri-
teria for dihydrotestosterone receptor proteins - on the basis that
the receptors are taken up by prostate nuclear chromatin and the bind-
ing of dihydrotestosterone is blocked by cyproterone acetate. The four
peaks are pI 5.8, Y2; 6.8, 03; 7.8, E3; and 8.3, ~l. We applied this
technique to the cytosol of the preputial glands of the three different
sexes (Fig. 6). The male preputial cytosol contains most of the dihydro-
testosterone binding proteins present in prostate cytosol, including
the 4 prostate cytosol dihydrotestosterone receptors. However, only
the male cytosol yz is taken up by the male preputial gland nuclear
chromatin (Table 2). The female cytosol also contains most of the same
binding proteins as prostate cytosol, except those with pI values
above 8 (Fig. 6). Of these binding proteins in the female cytosol only
yz is taken up by the female, male, and pseudohermaphrodite chromatin
(Table 2). The cytosol of the pseudohermaphrodite has all of the bind-
ing proteins except the Y2 binding protein (Fig. 6). It can be seen
that the male cytosol Y2 is taken up by the nuclear chromatin of both
the male and the pseudohermaphrodite, but the female chromatin recog-
nizes the sexually different male yz receptor and fails to take it up
(Table 2). Consequently, this study shows that the preputial gland
cytosol of the genetic male pseudohermaphrodite lacks the Y2 dihydro-
testosterone receptor, the presumed receptor protein which is respon-
sive to testosterone via its Sa-reduced metabolite of testosterone.
However, since the nuclear chromatin of the pseudohermaphrodite has
the capability of taking up the Y2 receptor from either the male or
429
female cytosol and does not distinguish the male receptor as does the
female chromatin, it would appear that the nuclear chromatin uptake
of the dihydrotestosterone receptors by the pseudohermaphrodite's pre-
putial is phenotypically male. This work confirms the indications of
previous workers - that the major defect in the preputial gland cyto-
sol of the pseudohermaphrodite is absence of the cytosol Y2 dihydro-
testosterone receptor, which most probably explains the expression of
the female phenotype of the preputial glands over several levels of
biological organization. The nuclear chromatin of the preputial gland
of the pseudohermaphrodite has the capacity to take up this receptor
from the cytosol of either sex in the phenotypically male pattern and
thus, would be expected to exhibit male differentiation if a Y2 recep-
tor were available.
"w
CD e
=>
.....
"E
.g- 4
o
o o
.....
I
o
o e
z Fig. 6. Isoelectrofocused
=> patterns of dihydrotestos-
o terone binding proteins
m
4 of preputial gland cytosol
o of three sexes. Pattern of
pseudohermaphrodite dif-
fers from those of male
4 5 6 7 8 9 and female by absence of
dihydrotestosterone recep-
pJ VALUE tor protein Y2, pI 5.8
We also examined the dihydrotestosterone binding proteins in the Wolf-

fian ducts of fetuses of 19 days of gestation, after treatment of preg-
nant rats with either cyproterone acetate at 30 mg/kg in dimethyl-sulf-
oxide or vehicle alone on days 14 to 19 (Goldman and Katsumata, 1974).
The Wolffian ducts contain the anlagen of the preputial glands. The
cytosol of the Holffian ducts during this critical period of masculine
differentiation contains the same dihydrotestosterone binding proteins
as the adult preputial glands including the Y2 receptor (Fig. 7).
Cyproterone acetate blocks the binding of this receptor with dihydro-
testosterone (Fig. 7). Thus, in fetuses rendered unresponsive to testos-
430
terone, either genetically or with cyproterone acetate, the dihydro-

testosterone receptor protein, Y2' is made either naturally deficient
or is experimentally blocked.
Table 2. Dihydrotestosterone receptors present in preputial gland

cytosol and uptake by nuclear chromatin
Cytosol receptor
Male Female Pseudohermaphrodite
Presence of
Y2 (pI 5.8) receptor + +
Uptake by nuclear
chromatin of
Male + +
Female +
Pseudohermaphrodite + +
+ Presence Or uptake.
- Absence or failure of uptake.
~ 5r---------------------------~
lU
.....
o
a:
..~
Cl.
co
e
"-
lU
CD
Fig. 7. Isoelectrofo-
:;)
..... cused patterns of dihy-
"- drotestosterone binding
..,eCo proteins of the Wolffian
o ducts of fetuses of 19
Q days of festation from
..... pregnant rats treated
I
o wi th vehicle (0······0)
o or cyproterone acetate
z
:;)
at 30 mg/kg from days
oco 14 to 19 (. • ).
o~--~----~----~--~~--~------~ Note that Y2 receptor
3 4 5 6 7 8 9 protein has been exper-
imentally blocked by
pI VALUE the anti androgen
E. Masculine Differentiation of the Sexual Zentrum
At this pOint, it would seem reasonable to conclude that the genetic

male rat pseudohermaphrodite is programmed i n utero not as a male, but
as a female, with the exception of its intersexual body weight and
growth, and its phenotypically masculine relative weight of its thyroid
glands. The female phenotype of the brain centers controlling taste
preference in the pseudohermaphrodite made us wonder whether its hypo-
thalamic-pituitary axis might be female also. Neumann et al. (1970)
had already observed that adult male experimental pseudohermaphrodites
"feminized" by prenatal and neonatal cyproterone acetate have a female
431
programming of the hypothalamic centers controlling cyclicity of gonad-

otropin secretion. Thus, we expected to find female patterns of pitu-
itary or serum levels of the various pituitary hormones and a femini-
zation of the hypothalamic-pituitary axis in the genetic pseudohermaph-
rodite. Pituitary and serum levels of luteinizing hormones (LH) and
follicle-stimulating hormones (FSH) have been previously shown to be
elevated in pseudohermaphrodites as compared with values in normal
littermate males (Bardin et al., 1973). Thus, we have compared pitu-
itary and serum levels of LH, FSH, prolactin and growth hormone col-
lected at the same time from genetic pseudohermaphrodites and those
of normal littermate males and females - as determined by radioimmuno-
assay (Goldman et al., 1975). With respect to sex differences, total
pituitary content and pituitary concentration of these hormones indi-
cate an intersexual character of the levels of pituitary hormones.
The serum concentration of F8H of the pseudohermaphrodite (Fig. 8) is
clearly masculine.
I
LH FSH GROWTH HORMONE
F P M F P M F P M F P M
Fig. 8. Serum concentration of pituitary hormones in three sexes.
Note phenotypically masculine levels of serum FSH in pseudohermaphro-
dite
F. Phenotype of the Hypothalamic-Pituitary Axis
These intersexual levels of the various pituitary hormones and pheno-

typically male serum levels of FSH in the pseudohermaphrodite, prompted
us to examine the sex of its hypothalamic pituitary axis by studying
its vaginal cytology, response to ovarian transplants and (in collab-
oration with Dr. Gustafsson) its urinary steroidal excretion patterns
after ovarian transplants (Shapiro et al., 1974). More than half the
pseudohermaphrodites studied were in constant vaginal estrus, while
the remaining rats displayed either persistent diestrus or irregular
cyclicity, tending towards lengthened estrus (Table 3).
Five to six weeks after transplantation, donor ovaries implanted in

female rats contain numerous follicles, sparse interstitium and the
bulk of the ovary was composed of corpora lutea. Ovarian grafts from
males and pseudohermaphrodites contain small follicles, virtually no
interstitial glandular tissue, no corpora lute a and the bulk of the
432
ovary is composed of large cystic follicles. Histochemical activity

of 3S-hydroxysteroid dehydrogenase also demonstrates that the response
of ovarian grafts in pseudohermaphrodites are phenotypically different
from that of females, but similar to that of males.
Table 3. Vaginal cycles of control females and male rat pseudohermaph-

rodites
Total Type of vaginal cycle

no. of b
rats 5 days Irregular C constalft Persistent
estrus diestrus e
Females 26 20 6 0 0 0
Pseudos 20 o 0 4 11 5
aFour day cycle composed of 1 day proestrus, day estrus and 2 days
diestrus.
bPive day cycle, the same as 4 day cycle, with an additional day of
estrus or diestrus.
CIrregular cycles were 3 to 12 days, with 2 to 7 days of estrus and
1 to 4 days of diestrus.
dVaginal smears composed solely of cornified cells.
eoaily vaginal smears varying numbers of cornified cells, but leuko-
cytes predominate.
Urinary sex steroids from male rats with ovarian grafts are undetect-
able because the highly polar steroids produced by the male liver are
too polar to be detected by the chromatography-mass spectometry method
used. Female rats with ovarian transplants excrete both Cl9 and C21
steroids of gonadal origin, while pseudohermaphrodites excrete only
C l 9 steroids originating from ovarian grafts (Table 4). Both females
and pseudohermaphrodites excrete metabolites of corticosterone and
these are the only hormones detectable in the urine two weeks after
the removal of the grafts. Thus, the ovarian grafts in pseudohermaph-
rodites do not excrete progesterone metabolites but only metabolites
of Cl9 estrogen precursors.
The occurrence of constant estrus in one-half the pseudohermaphrodites

is indicative of a phenotypically male type of continuous, tonic gonad-
otropin secretion causing estrogen production by the testis. It is
unlikely that four out of the 20 pseudohermaphrodites showing "cyclic
vaginal cytology" have phenotypically female hypothalamic-pituitary
axes because, although it is possible for tonic estrogen secretion to
result in vaginal cyclicity, the presence of progesterone is required
(Maekawa, 1959). As mentioned above, progesterone metabolites have not
been found in the urine of intact pseudohermaphrodites. Nevertheless,
masculinization of the hypothalamic-pituitary axes is also indicated
by the constant vaginal estrus and cystic ovarian' transplants in the
two "cyclic" pseudohermaphrodites with functioning ovarian grafts.
Thus, these results provide evidence that the phenotype of hypothala-
mic-pituitary axis of the pseudohermaphrodite may not be female.
433
Table 4. Major urinary steroid metabolites from castrated females and

male rat pseudohermaphrodites bearing functional ovarian transplants
Females Pseudos
Ovarian origin
C19 metabolites
3a,7a-dihydroxy-5a-androstan-17-one +
5a-androstan-3a,7a,17S-triol + +
3a,11S-dihydroxy-5a-androstan-17-one + +
C21 metabolites
3a,15a-dihydroxy-5a-pregnan-20-one +
3a,16a-dihydroxy-5a-pregnan-20-one +
Adrenal origin
15-hydroxylated corticosterone metabolites + +
apresence (+) or absence (-) or steroid in urine.
G. Masculine Phenotype of Vaginal Physiology of Pseudohermaphrodite
We examined the phenotypic sex of one other physiologic parameter in

the genetic male pseudohermaphrodite - that of vaginal opening and
response of vaginal patency to hormone deprivation and replacement
(Shapiro and Goldman, 1975). The vagina in littermate females opens
at 57 ± 4.7 (s.d.) days of age, but the vaginal opening of the pseudo-
hermaphrodite is significantly delayed to 78 ± 17.4 (p < 05) (Fig. 2)
The vaginae of littermate females remain patent after adrenalectomy

and ovariectomy (Fig. 9). However, ovariectomy converts the cyclicity
of female vaginae to peristent diestrus unless estradiol benzoate at
a dose of 0.5 ~g/day is administered (Fig. 9). Diestrus returns with
lower doses of estradiol-17-benzoate. However, the vaginae of all the
females remain patent after complete endocrinectomy. At the time of
adrenalectomy of the 15 pseudohermaphrodites with patent vaginae,
6 were in estrus, 2 were in diestrus and 2 had irregular cyclicity
(Fig. 9). Adrenalectomy does not affect vaginal patency or smears.
Within thwo days following testectomy, the vaginae of all the pseudo-
hermaphrodites displayed constant diestrus and after one week the va-
ginae closed. The vaginae could not be reopened with large daily doses
of progesterone or androsterone, the primary steroid produced by testes
of these animals (Goldman and Klingele, 1974a). After a large daily
dose of testosterone propionate the vaginae of two of the 15 pseudo-
hermaphrodites reopened and remained in diestrus. The vaginae of the
remaining pseudohermaphrodites were reopened by daily injections of
0.5 ~g/estradiol-17-benzoate and the same response of vaginal smears
to estradiol-17-benzoate were obtained as those of their littermate
females.
The vaginae of 5 pseudohermaphrodites never opened, nor could they be

opened by any of the procedures or any hormone including estrogen
(Fig. 9). This observation and the closing of the vaginae after gonad-
ectomy (unlike the vaginae of littermate females) suggest that the
perineum of the pseudohermaphrodite may have been partially masculin-
434
ized - because the vaginae of normal females transplanted to intact

males remain patent and respond in similar fashion to estradiol treat-
ment of the males, as do the vaginae of intact females (Takewaki, 1956;
Yazaki, 1959). These results suggest that the failure of opening may
be a defect in the vaginae themselves rather than in the hypothalamic-
pituitary axis of the pseudohermaphrodite. At any rate, the failure of
opening and the closing of the vaginae of the pseudohermaphrodite in-
dicates that the phenotypic sex of the vaginae is not female, but per-
haps masculine or, at least, intersexual. Thus, it appears the genetic
defect in th i s animal prevents androgen-dependent differentiation in
most tissues allowing expression of the female phenotype except in its
hypothalamic-pituitary axis, thyroid relative weights and responses of
vaginal patency to hormone deprivation and replacement, wherein the
masculine phenotype may be expressed.
! I I
AAIMALS '0 , VAGINAL
FEMLES 10
STATE
OPEN ,.1I1I..lIlIlIlIlIliilllll..II..II..IIIIII....II::..D.I~ESTRUS
OIESTRUS ~~TR~S7
CYCLlC 10
10137 10
ESTRUS
PSEUOOS 15
ESTRUS 6 1 • • • • ....
' 1 .... ' • • 1
.. .. ' .... .... 1•• • ...... 'or . 'r
01 ESTRUS 2 ,3 10
"CYCLIC" 2
(LO ED
!' ::l : ';; 'l' ;i;:!;: ;;;t:; '; ' ;; '::; 'l~';.~
EVER
5
~~"
OPENED
'1" ..
PROCEDURE; AoREX TesTEx OR

OVARIEX
VEHICLE
0.1 ML/o
PROGES-~- TESTOSTER-~TRA
TERONE STERONE ONE-PRO-
10L
BENZOAT t _
2 MG/o 5 MG/D PRIONATE -0 . 5~~ " G/o
4 MG/D
Fig. 9. Response of vaginal patency to hormone deprivation and replace-

ment with various hormones. Afte r adrenalectomy and gonadectomy note
persiste nce of vaginal patency in littermate females but phenotypically
non-f e ma le vaginal closure in 15 out of 20 pseudohermaphrodites. Also,
note phenotypically non-female occurrence of 5 out of 15 pseudoher-
maphrodites failure of vaginal opening even after estrogen treatment
Although it is difficult to explain the phenotypically masculine pro-

gramming of the hypothalamic-pituitary axis and vagina of the pseudo-
hermaphrodite, there are thre e possible explanations. The programming
of th e se parameters may be more sensitive to androgen imprinting than
other target organs. Indeed, th.e androgen insensitivity of the male
rat pseudohermaphrodite is not absolute, as it will r e spond to large
doses of androgen (Stanley et al., 1973; Bardin et al., 1973; Goldman
and Klingele, 1974a-c). These target organs may be programmed before
the androgen insensitivity d e v e lopes. Masculine differe ntiation of the
pseudohe rmaphrodite appears normal up to day 17in uter o when the anlagen
for the ductal and accessory systems begin to regress (Stanley et al.,
1973). In this regard, the patt e rn of gonadotropin release in the neo-
natally castrated male rat may be partially masculinized because of
possible prenatal androgen programming (van der Schoot and Zeilmaker,
1972). Lastly, the programming of these target organs may involve other
435
receptors (e.g., testosterone) than the deficient dihydrotestosterone

receptor. Unlike testosterone, dihydrotestosterone cannot induce mascu-
linization of the sexual centers of the hypothalamus nor can it program
male behavioral patterns (Luttge and Whalen, 1970; McDonald and Doughty,
1972; Johnson, 1972).
In summary, we have perhaps raised more questions concerning the pheno-

typic sex of the rat pseudohermaphrodite than we have answered, but it
seems clear that the preputial glands of the pseudohermaphrodite have
expressed its inborn female program associated with its genetic defi-
ciency of the Y2 dihydrotestosterone cytosol receptor. The absence of
the receptor also imparts to the adult gland its phenotypically female
enzyme pattern of androgen metabolism. The phenotypic differentiation
of the hypothalamic-pituitary axis in this animals is apparently mas-
culine in certain respects and the molecular basis of this difference
of sex differentiation remains to be clarified.
References
Bardin, C.W., Bullock, L.P., Sherins, R.J., Movszovicz, I., Blackburn,

W.R.: Part II. Androgen metabolism and mechanism of action in male
pseudohermaphroditism: a study of testicular feminization. Rec.
Progr. Horm. Res. 29, 65-109 (1973).
Bottiglioni, I.F., Eberlein, W.R., Goldman, A.S., New, H.: Disorders
of adrenal steroid biogenesis. Rec. Progr. Horm. Res. 23, 375-499
(1967). -
Bottiglione, I.F., Collins, H.P., Flamigni, C., Neumann, F., Sommer-
ville, I.F.: Studies on androgen metabolism in experimentally femi-
nized rats. Endocrinology 89, 553-559 (1971).
Einarsson, K., Gustafsson, J.A., Goldman, A.S.: Metabolism of steroid
hormones, sterols, and bile acids in liver microsomes from male,
female and male pseudohermaphroditic rats. Eur. J. Biochem. 11,
345-353 (1972).
Goldman, A.S.: Animal models of inborn errors of steroidogenesis and
steroid action (eds. Gibian and Plotz). Mammalian reproduction,
pp. 389-436. Berlin-Heidelberg-New York: Springer 1970.
Goldman, A.S.: Inhibitors of steroidogenesis: Biochemical and pharma-
cological implications. Excerpta Medica Int. Congr. Ser. 219, 839-
848 (1972a).
Goldman, A.S.: Inhibition of 3S-hydroxY-6 5 -steroid dehydrogenase.
Hormones and antagonists. Gynec. Invest. 2, 213-238 (1972b).
Goldman, A.S.: Rat fetal target organ 6 5 ,3S-hydroxysteroid dehydro-
genase: Effect of cyanoketone and cyproterone acetate. Endocrinol-
ogy 92, 1043-1050 (1973).
Goldman~A.S.: Influence of hormones on sex differentiation. Contem-
porary OB/GYN 3, 69-80 (1974a).
Goldman, A.S.: Sexual programming of the fetus and neonate studied by
selective biochemical testosterone depriving agents. Adv. Biosci.
Hormones and Embryonic Development 13, 17-40. Braunschweig: Pergamon
Press - Vieweq 1974b.
Goldman, A.S., Bongiovanni, A.M.: Induced genital anomalies. Ann. N.Y.
Acad. Sci. 142, 755-767 (1967).
Goldman, A.S., Gustafsson, J.A., Gustafsson, 5.: Female pattern of
metabolism of 4-14C corticosterone in male pseudohermaphroditic
rats. Proc. Soc. expo BioI. Med. ill, 691-696 (1973).
Goldman, A.S., Katsumata, M.: Separation of androgen binding proteins
in rat fetal wolffian duct and genital tissues: Fetal effects of
cyproterone acetate administered to the mother. Biochim. Biophys.
Acta (submitted, 1974).
436
Goldman, A.S., Klingele, D.A.: Developmental defects of testicular

morphology and steroidogenesis in the male rat pseudohermaphrodite
and response to testosterone and dihydrotestosterone. Endocrinology
94,1-16 (1974a).
Goldman, A.S., Klingele, D.A.: Persistent postpubertal elevation of
activity of steroid 5a-reductase in the adrenal of rat pseudoher-
maphrodites and correction by large doses of testosterone or DHT.
Endocrinology ~, 1232-1240 (1974b).
Goldman, A.S., Klingele, D.A.: Feminine patterns of androgen metabolism
and presence of 6 5 ,3S-hydroxysteroid oxidoreductase in target organs
of genetic rat male pseudohermaphrodites. Proc. Soc. expo Biol. Med.
ill, 539-545 (1974c).
Goldman, A.S., Root, A.W., Duckett, G., Shapiro, B.H.: Pituitary hormone
secretion in the genetically male rat pseudohermaphrodite. J. Endo-
crinol. (1975, in press) .
Gustafsson, J.A., Goldman, A.S.: Steroid monosulfates in urine from
male pseudohermaphroditic rats. Endocrinology 90, 931-935 (1972).
Gustafsson, J.A., Carlstedt-Duke, J., Goldman, A.S.: On the hepatic
sulfurylating activity in male pseudohermaphroditic rats. Proc.
Soc. expo Biol. Med. 145, 908-911 (1974).
Johnson, D.C.: Sexual differentiation of gonadotropin patterns. Amer.
Zool. 12, 193-204 (1972).
Jost, A.::Embryonic sexual differentiation (morphology, physiology,
abnormalities). In: Hermaphroditism, Genital Anomalies, and Related
Disorders (eds. Jones and Scott), 2nd ed., pp. 16-62. Baltimore:
Williams and Wilkins 1972.
Katsumata, M., Goldman, A.S.: Separation of multiple dihydrotestosterone
receptors in rat ventral prostate by a novel micromethod of electro-
docusing: Blocking action of cyproterone acetate and uptake by nu-
clear chromatin. Biochim. Biophys. Acta 359, 112-129 (1974).
Luttge, W.G., Whalen, R.E.: Dihydrotestosterone androstenedione,
testosterone: Comparative effectiveness in masculinizing and defe-
minizing reproductive systems in male and female rats. Horm. Behav .
..1-, 265-281 (1970).
Maekawa, K.: Counteraction between estrogenic and progestational prop-
erties of 17a-ethinyl-19-nor-testosterone. Endocrinology ~, 161-165
(1959) .
Mainwaring, W.I.P., Irving, R.: The use of deoxyribonucleic acid -
cellulose chromatography and isoelectric focusing for the charac-
terization and partial purification of steroid-receptor complexes.
Biochem. J .. 134, 113-127 (1973).
McDonald, P.G., :DOughty, C.: Comparison of the effect of neonatal ad-
ministration of testosterone and dihydrotestosterone in the female.
Reprod. Fertil. 30, 55-62 (1972).
Milin, B., Roy, A.K-:-;- Androgen receptor in rat liver cytosol, "defi-
ciency" in pseudohermaphrodite male rats. Nature (New Biol.) 242,
248-250 (1973). ---
Neumann, F., Berswordt-Wallrabe, R. von, Elger, W., Steinbeck, H.,
Hahn, J.D., Kramer, M.: Aspects of androgen-dependent events as
studied by antiandrogens. Rec. Progr. Horm. Res. 26, 337-410 (1970)
Ritzen, E.M., Nayfeh, S.N., French, F.S., Aronin, P.~: Deficient nu-
clear uptake of testosterone in the androgen-insensitive (Stanley-
Gumbreck) pseudohermaphroditic male rat. Endocrinology 21, 116-124
(1972) .
Shapiro, B.H., Goldman, A.S.: Feminine saccharin preference in the
male rat pseudohermaphrodite. Horm. Behav. 4, 371-375 (1973).
Shapiro, B.H., Goldman, A.S., Gustafsson, J.A.~ Masculinization of
the hypothalamic-pituitary axis in the androgen insensitive genetic
male rat pseudohermaphrodite. Endocrinology (1975, in press).
Shapiro, B.H., Goldman, A.S.: Testicular maintenance and masculinized
development of the vagina in the androgen insensitive genetically
male rat pseudohermaphrodite. Endocrinology (1975, in press).
437
Stanley, A.J., Gumbreck, L.G., Allison, J.E.: Part I. Male pseudoher-

maphroditism in the laboratory Norway rat. Rec. Progr. Horm. Res.
~, 43-64 (1973).
Takewaki, K.: Difference in response to injections of estrogen between
intratesticular and subcutaneous vaginal transplants in male rats.
J. Pac. Sci. U. Tokyo Sec. IV 2, 641-653 (1956).
van der Schoot, P., Zeilmaker, G.H.: Aspects of the function of ovarian
grafts in neonatally castrated male rats. Endocrinology 91, 389-395
(1972). -
Yazaki, I.: Effects of adrenalectomy, injections of hormonic steroids
or gonadotropins and subjection of stressful stimuli or subcutaneous
ovarian grafts in castrated male rats as studied by daily examina-
tions of vaginal smears. Jap. J. Zool. ~, 267-277 (1959).
Chromosomes and Sex Differentiation *
U. Mittwoch
The age-old search to discover the cause underlying the development

of the two sexes seemed to be crowned with success at the onset of
the twentieth century. The discovery of an unlike pair of chromosomes
in the spermatocytes of different species of insects led McClung (1902)
to suggest that the two types of spermatozoa might give rise to male
and female individuals respectively. Like many new ideas, McClung's
was not immediately accepted. Even three quarters of a century later,
when the chromosomal mechanism of sex differentiation is firmly estab-
lished in many classes of anJmals, some of the early problems still
remain unsolved.
To begin with, the chromosomal mechanisms which bring about sexual

differentiation appear to be extraordinarily variable. When the male
has an x- and a Y-chromosome, sex is determined by the type of sper-
matozoon which fertilizes the egg. The male is therefore said to be
the heterogametic sex, while the female is homogametic. This system
was first discovered in the meal worm, Tenebrio molitor, and was subse-
quently found to occur also in the fruitfly, Drosophila melanogaster, as
well as in man and other mammals (Mittwoch, 1973). However, although
man and Drosophila share the same formal sex chromosome mechanism, the
details of sex determination are clearly different; since in man and
other mammals, the Y-chromosome is strongly male determining, whereas
in Drosophila, sex is ordinarily determined by the number of X-chromo-
somes present. In some insects with male heterogamety, a Y-chromosome
is absent, which means that male determining spermatozoa lack a sex
chromosome.
In birds as well as in Lepidoptera, the female is the heterogametic

sex. This means that the eggs fall into two classes regarding their
sex chromosome constitution, while the spermatozoa are all the same
in this respect. The sex chromosomes in animals with female heteroga-
mety are often called ZW in the female and ZZ in the male, although
obviously the letters assigned to these chromosomes are arbitrary.
The fact that mammals and birds have sex chromosome mechanisms which
appear to be diametrically opposed to each other indicates that the
sex determining mechanism failed to be stabilized at an early stage
of vertebrate evolution. In lower vertebrates (fish as well as in am-
phibia), both male and female heterogamety occur; indeed, in the platy-
fish, Xiphophorus maculatus, male and female heterogamety are found wi thin
a single species (Kallman, 1965). Among reptiles, there is definite
evidence of female heterogamety in different species of snakes (Ray-
Chaudhuri et al., 1971). By contrast, male heterogamety has been demon-
strated in the chromosomes of some species of lizards (Pennock et al.,
1969; Cole, 1974). In a crocodile, Crocodylus palustris, no evidence of
heteromorphic sex chromosomes was found (Singh and Ray-Chaudhuri, 1973)
"'This work was supported by a grant from the -Science Research Council.
439
and the same applies to three species of turtles (Singh, 1972). Evi-
dently, sex chromosome mechanisms must have arisen repeatedly during
the evolution of vertebrates, suggesting that the switch mechanism,
which results in male or female development respectively, can be brought
about by different chromosomal regions.
The absence of heteromorphic sex chromosomes does not necessarily mean

that a pair of chromosomes may not function as sex chromosomes. In sev-
eral species of fish, genetic markers giving rise to characteristic
color patterns, are closely linked to sex differentiating segments of
chromosomes, which are not known to be cytologically distinct. The
earliest example of sex linkage in a fish was described by Aida (1921)
in the medaka, Oryzias latipes, in which a color factor is responsible
for the production of a red pigment. The inheritance of this factor
suggested that in this species, males are normally XY and females XX.
Exceptional XY females do occur, however, and these give rise to XX
daughters as well as to XY and YY sons (Aida, 1936).
The term "sex reversal" is often applied to individuals whose phenotypic
sex has developed in opposition to that indicated by the sex chromosome
constitution. However, the. term "sex inversion" is more specific, since
"sex reversal" has also been applied to situations in which changes in
the sexual phenotype occur secondarily, after the original process of
sex differentiation has been accomplished (Reinboth, 1970). The fact
that complete sex inversion, leading to the development of fertile fe-
males, is possible in the medaka; as well as the viability of YY males,
suggests that the sex chromosome mechanism is still in a relatively
early stage of evolution and that the X- and Y-chromosomes do not differ
markedly from each other.
Sex determination in the medaka has more recently been studied by
Yamamoto (1969), who succeeded in obtaining complete sex inversions
by means of mammalian steroids. Estrogens, either estrone or stilbes-
terol, when administered to young while the gonads were still in the
indifferent stage, caused the chromosomal males to develop into females.
These XY-females became the mothers of YY-males. However, by further
treatment with estrogens, they could be made to develop as females
which gave rise to all male offspring. Yamamoto also succeeded in pro-
ducing XX-males by the administration of methyltestosterone.
In the platyf ish, XiphophoY'us maculatus, color factors which are respon-
sible for the production of macromelanophore patterns, are closely
linked to the sex differentiating segments of the sex chromosomes.
In the Jamapa strain, males are XY and females XX. This fish is a live-
bearer and Anders et al. (1969) succeeded in obtaining sex inverted
XY-females by X-irradiation of the mothers. In turn, the XY-females
produced YY-males. By further irradiation of embryos, YY-females were
produced but these did not prove to be fertile.
The oviparous fish Rivulus marmoY'atus is unusual in being a hermaphrodite

which normally reproduces by self-fertilization. Sex chromsomes are
clearly not involved in this species. Therefore, it must be assumed
that the formation of ovarian and testicular regions is under the con-
trol of those largely unknown mechanisms, which normally bring about
the differentiation of organs within individuals. Harrington (1974)
found that by lowering the temperature in which these fish were raised,
the gonads developed into testes instead of ovotestes. Although the
production of males rather than hermaphrodites should probably not be
regarded as sex inversion in the strict sense, the change in gonadal
differentiation in response to a change in temperature would seem to
be prima facie evidence that the process is affected by the metabolic
rate. Although the role of steroid hormones in the process of sex dif-
ferentiation is far from being settled, these hormones are known to
~o
affect the mitotic cycle in target organs (Epifanova, 1971). Moreover,

while the action of X-irradiation on cells and tissues is certainly
complex, its effect as a mitotic inhibitor is widespread. Results of
experimentally induced sex inversion in fish therefore pOint to a
possible relationship between sex differentiation and growth.
In mammals, the sex chromosomes have undergone many evolutionary chan-

ges. At the cytological level this is seen, first, in the marked asym-
metry of the sex chromosome pair. Although we must assume that the sex
chromosomes have originated from an ordinary homomorphic pair, in most
mammalian species including man, the Y-chromosome is considerably
smaller than the X, which means that it contains less chromosomal ma-
terial. At least one X-chromosome is necessary for the survival of the
cell and the YY-constitution, which as we have seen is viable in fish,
is unknown in mammalian cytogenetics.
The differential behavior of the two X-chromosomes in female mammals

may be seen as a further consequence of the unequal amount of genetic
material which is present in XX compared with XY karyotypes. One of
the two X-chromosomes of female mammals replicates its DNA late in the
mitotic cycle and, in interphase cells, condenses to form the sex chro-
matin body (Mittwoch, 1973). Most of the genes on this X-chromosome
are inactive .. Therefore, female mammals produce the same amounts of
the products of genes located on the X-chromosomes as are found in
male mammals (Lyon, 1972). This process is known as "dosage compensa-
tion".
Although dosage compensation is a well established phenomenon in mam-

mals, it should be pointed out that it does not seem to occur in birds
or reptiles. In chickens, the small unpaired sex chromosome occurs in
the female, but neither of the two large sex chromosomes in males is
late replicating (Schmid, 1962) and it appears that dosage compensation
does not take place (Ohno, 1967). In snakes, the two X-(Z-)chromosomes
label synchronously with the autosomes (Ray-Chaudhuri and Singh, 1972).
There is evidence that late replicating heterochromatic chromosomal
regions require less time for DNA synthesis than do euchromatic regions,
which replicate at the normal time (Schmid and Leppert, 1968). Thus,
apart from its effect on dosage compensation, the heterochromatiniza-
tion of one of the X-chromosomes of female mammals may be a device to
minimize the difference in time required to complete the mitotic cycle
in males and females. It seems likely that the time-table for cell
division is more stringent in mammals than in birds and reptiles.
Irrespective of the number of X-chromosomes, the sex of a mammal is

determined principally by the presence or absence of a Y-chromosome.
If a Y-chromosome is present, the embryo will develop into a male and
in the human species there is ample evidence that this will occur even
in the presence of two, three or four X-chromosomes. Conversely, in
the absence of a Y-chromosome, the embryo will develop into a female,
even if only a single X-chromosome is present (Mittwoch, 1973). In
spite of their undoubted masculinity, patients with Klinefelter's syn-
drome and 47,XXY-chromosomes have small testes, which. in adults lack
spermatogenesis. Adult patients with Turner's syndrome have, in addi-
tion to various somatic abnormalities, streak gonads which lack ovarian
follicles.
The XO-chromosome constitution is also known in the mouse and in this

species leads to a more normal female development. Mice with this chro-
mosome constitution are phenotypically normal females, but they produce
fewer offspring than do normal females (Morris, 1968) and their repro-
ductive life is shorter. It is therefore of interest to note that in
human fetuses with XO sex chromosomes, ovaries containing germ cells are
441
present, which are not detectable different from normal (Singh and
Carr, 1966). However, the germ cells become abnormally few in number
after the third month of pregnancy and few, if any, remain at the time
of birth. It seems therefore, that the rate of germ cell atresia, which
normally takes place in female mammals, is increased in the XO-chromo-
some constitution. This probably happens both in the human species and
in the mouse and the difference in fertility may be related, at least
in part, to the difference in time of life span of the two species.
An XO mouse can have her young at an age when her human counterpart
is still in her mother's womb.
While abnormal numbers of X-chromosomes will bring about various ab-

normalities of sexual development, in mammals the decision - male or
female - is under the control of the Y-chromosome. In the human spe-
cies, the male differentiating region of the Y-chromosome is located
on the short arm (Jacobs and Ross, 1966). We know from castration ex-
periments (Jost, 1970), that the fetal testis plays a decisive role in
the development of the secondary sexual characteristics. If testes are
present, the HUllerian ducts regress and the ~volffian ducts develop
into vasa deferentia and seminal vesicles. If testes are absent, the
MUllerian ducts develop into Fallopian tubes and uterus, while the
Wolffian ducts disappear. These developments are controlled by secre-
tions from the fetal testes; fetal ovaries do not play any role at
this stage. We may conclude therefore, that the crucial task of the
mammalian Y-chromosome is to induce the indifferent gonad to develop
into a testis; and one of the major tasks in developmental cytogenetics
is to elucidate the mechanism underlying this process.
In eutherian mammals it has not been possible to achieve gonadal sex

inversion by experimental means, but certain natural experiments of
abnormal sex differentiation may help to throw some light on this prob-
lem. Nothwithstanding the strong male determining power of the Y-chro-
mosome, a number of examples are known in which testicular differentia-
tion occurs apparently in the absence of a Y-chromosome. A very high
degree of sex inversion is achieved by the Sex reversed factor in the
mouse (Cattanach et al., 1971). In this condition, an autosomally in-
herited dominant factor causes XX animals to develop into males. The
XX animals which carry the Sxr factor have the appearance of normal
males, though their testes are small and in adult mice lack germ cells.
Germ cells, however, are present in embryos and in newborn mice, but
they become progressively fewer and none are left by the time the ani-
mals are ten days of age.
In the embryo, the testes of XX Sxr mice are histologically indistin-

guishable from those of their XY litter mates. However, measurements
from serial sections have shown that already on days 15 and 16, the
XX testes are smaller than testes from XY embryos (Hittwoch and Buehr,
1973; Table 1) the difference is significant at the one per cent level.
These results suggest that the markedly small size of the testes of
adult XX, Sxr mice is already foreshadowed by a subnormal growth rate
in the embryo.
One of the 15 day old XX, Sxr embryos included in Table 1 had bilateral
ovotestes and was, therefore, a hermaphrodite. The mean gonadal volume
of the two ovotestes was 820 ~m x 10 5 - the lowest value found in any
embryo with testicular tissue.
Hermaphroditism is found as a rare abnormality in other mammalian spe-

cies and is most fully documented in man. Of 108 true hermaphrodites
listed by Polani (1970), 59 had apparently only 46,XX cells, 21 were
46,XY while 28 had more than one cell line. It is possible, however,
that at least some of the patients with the 46,XX cells had a hidden
442
cell line containing a Y-chromosome and that this cell line had become
rare. Hermaphrodites having an ovary on one side and a testis on the
other are said to be lateral, while bilateral hermaphrodites have ovo-
testes on both sides. Unilateral hermaphrodites have an ovotestis on
one side and an ovary or a testis on the other one. There is marked
lateral asymmetry with respect to type of gonad (Table 2). As pointed
out by Polani (1970), ovaries are more often on the left and testes on
the right side, in both xx and XY individuals.
Table 1. l1ean gonadal volumes in mouse embryos ()lm 3 x 10 5 )
Chromo- Gonad Age of embryos 15 days Age of embryos 16 days

somes
No. of Mean S.E. No. of Hean S.E.
embryos gonadal embryos gonadal
volume volume
XY Testes 8 933 28.8 21 1333 28.2

XX Testes 4 846 17 .5 9 1152 71 .6
XX Ovaries 5 303 21.0 8 328 17 .2
Table 2. Lateral asymmetry of gonads in human hermaphrodites. (From

Polani, 1970)
Chromosome Type of gonad No. on

constitution right left
46,XX Ovary 28 43
Testis or ovotestis 56 41
46,XY Testis 25 14
Ovary or ovotestis 23 34
In this connection it is of interest that Jost et al. (1972) found the

volumes of the right fetal gonad in bovine fetuses to be slightly lar-
ger, on average, than the volume of the left gonad (Table 3). This
tendency was most pronounced in freemartin fetuses and among normal
embryos, was more marked -in females than in males.
Table 3. Lateral asymmetry of fetal gonadal volumes in cattle. (Data

from Jost et al., 1972)
Gonads Sex No. of Right Left Differ- S.E. Diff. / P

fetuses ence S.E.
Testes Male 28 1 .46 1 .43 0.029 ±0.026 1 . 12 0.30
Ovaries Female 21 1 .62 1. 55 0.082 ±0.034 2.41 0.05
Ovaries Freemartin 29 0.66 0.59 0.068 ±0.014 4.86 0.001
Further evidence on lateral asymmetry is provided by the total ridge

count. This is a quantitative estimate of the finger-tip pattern size,
which is affected by the number of X- and Y-chromosomes present (Pen-
rose, 1967). It is thought that the ridges are laid down during the
Table 4. Lateral asymmetry in total ridge counts. (Data from Holt, 1954 )
No. of Mean total ridge count Difference S.E. Diff./S.E. P

individuals
Right Left
Fathers 50 69.5 65.4 4.12 ± 1.42 2.90 0.01

Mothers 50 64.4 60.0 4.40 ± 1.37 3.21 0.01
Sons 53 68.5 67.0 1. 51 ± 1.31 1.15 0.20
Daughters 53 61 .4 57.8 3.57 ± 1.22 2.93 0.01
Additional sibs ~ 65 74.2 71.8 2.40 ± 1.29 1 .86 0.10
Additional sibs ~ 46 57.6 53.2 4.41 ± 1.39 3.17 0.01
Table 5. Lateral asymmetry of gonadal volumes in chick embryos. (From Mittwoch et al., 1971 )
Age (days) n Males S.E. n Females S.E.
Mean gonadal volume Mean gonadal volume
Left/right Left/right
5 2 2.12 0.09 3 1 .74 0.08
6 3 1. 46 0.09 1 1 .82
7 2 1.72 0.23 4 1.72 0.10
8 3 1. 29 0.10 8 2.18 0.13
8.75 6 1. 35 0.07 4 2.95 0.30
9 4 1. 32 0.10 3 2.62 0.37
10 6 1.32 0.09 3 3.41 0.37
£
444
formation of the hands, from the eight to the tenth week of fetal life.
The ridge count is likely to be related to the size of the finger pads
at this time. It was shown by Holt (1954) that the ridge count tends
to be higher on the right than on the left side. Moreover, the data in
Table 4 show that this difference is more marked in females than in
males. It would seem therefore, that the asymmetry of gonad in true
hermaphrodites can be related to a difference in growth rate between
right and left sides. We may assume that the homeostatic mechanism,
which normally regulates gonadal differentiation in mammals, has broken
down in the development of hermaphrodites. In this labile situation,
the gonad on the faster growing right side is more likely to develop
into a testis, while the gonad on the slower growing left side tends
to develop into an ovary.
Gonadal asymmetry is far more marked in birds than in mammals. With

few exceptions, only the left gonad of female birds develops into an
ovary. The right one normally remains in a rudimentary state but, if
it does develop, tends to develop into a testis (Lillie, 1952). The
relationship of ovary on the left and testis on the right appears to
be the same as that in human hermaphrodites. I would like to suggest,
however, that this is for two opposing reasons. Measurements of gonadal
volumes in chick embryos have shown that the left gonad is larger than
the right one in both male and female embryos (Mittwoch et al., 1971)
(Table 5). Until about day 7, while the gonads are in the indifferent
state, the degree of asymmetry is similar in both sexes; after this
time, however, when the gonads become sexually differentiated, the
difference in size between left and right gonads becomes less in males
and greater in females. At the same time the incipient ovary, which
previously had been somewhat smaller than the incipient testes, over-
takes its male counterpart in size and henceforth becomes the larger
gonad. In contrast to the situation in mammals, the ovary of chick
embryos is a secretor of sex hormones (Wolff and Wolff, 1951). Thus,
birds with a sex chromosome mechanism which appears to be the opposite
of that in mammals differ from mammals also in that the ovary is the
dominant embryonic gonad and that organs on the left side seem to have
a faster growth rate than those on the right.
During the course of vertebrate evolution, the sex chromosomes have

become progressively more powerful in controlling the process of go-
nadal differentiation, which originally must have been under environ-
mental control. The chromosomal control of sex differentiation would
seem to have become most efficient in mammals, where the Y-chromosome
is able to induce testicular differentiation under almost any condi-
tions, including most chromosome abnormalities. Nevertheless, the oc-
currence of hermaphroditism, especially in individuals with only a
single cell line, is a reminder that even in mammals the gonad is orig-
inally bipotential. Lateral asymmetry is but one of the pointers re-
lating gonadal differentiation with differential rates of growth.
Testes need to produce large numbers of small gametes while ovaries

have to manufacture a small number of large gametes. Since the two
basic function of gonads are essentially complementary, it is under-
standable that in the early evolutionary period of sex chromosomes in
vertebrates, two opposite systems, i.e. male and female heterogamety,
were tried out. However, in addition to their function as organs of
gametogenesis, the gonads are also producers of sex hormones and it
may well be that for this reason, mammals had no choice but to adopt
the system of male heterogamety. The mammalian embryo develops in its
mother's womb, where embryos of both sexes find themselves in an en-
vironment rich in estrogen. In order to develop a male phenotype, an
embryo must counter the maternal sex hormones by secreting sufficient
amounts of androgens at an early stage of development. Therefore, the
445
function of the Y-chromosome may be to induce an early and rapid de-

velopment of the embryonic gonad, thus enabling it to secrete male
sex hormones before the embryo becomes feminized. By contrast, the
avian embryo does not find itself in this predicament and so sex de-
termination by female heterogamety is possible, where rapid growth
leads to ovarian differentiation.
It would clearly be most desirable to obtain further evidence in dif-

ferent vertebrates, particularly in species which are livebearers. It
would seem however, that the evidence available so far gives little
support to the idea that sex differentiation is under the control of
structural genes, but suggests that the sex differentiating segments
of the sex chromosomes act by affecting the rate of growth and devel-
opment, perhaps by increasing the metabolic rate of the bipotential
gonad at a critical stage in development.
References
Aida, T.: On the inheritance of color in a fresh-water fish, Aplocheilus

latipes Temmick and Schlegel, with special reference to sex-linked
inheritance. Genetics 6, 554-573 (1921).
Aida, T.: Sex reversal in- Aplocheilus latipes and a new explanation of
sex differentiation. Genetics 21,136-153 (1936).
Anders, A., Anders, F., Rase, S.:-XY females caused by X-irradiation.
Experientia 25, 871 (1969).
Cattanach, B.M.~Pollard, C.E., Hawkes, S.G.: Sex-reversed mice: XX
and XO males. Cytogenetics 10, 318-337 (1971).
Cole, C.J.: Evolution of parthenogenetic species of reptiles. (This
volume) .
Epifanova, 0.1.: Effects of hormones on the cell cycle. In: The Cell
Cycle and Cancer (ed. R. Baserga), pp. 145-196. New York: Marcel
Dekker 1971.
Harrington, R.W., Jr.: Sex determination and differentiation in fishes.
In: Control of Sex in Fishes (ed. C.B. Schreck), pp. 4-12. Sea Grand
and V.P.I. & S.U. Press 1974.
Holt, S.B.: Genetics of dermal ridges: bilateral asymmetry in finger-
ridge counts. Ann. Eugen. 18, 211-231 (1954).
Jacobs, P.A., Ross, A.: Structural abnormalities of the Y chromosome
in man. Nature (Lond.) 210, 352-354 (1966).
Jost, A.: Hormonal factors in the sex differentiation of the mammalian
foetus. Phil. Trans. Roy. Soc. B 259, 119-130 (1970).
Jost, A., Vigier, B., Prepin, J.: Freemartins in cattle: the first
step of sexual organogenesis. J. Reprod. Fert. 29, 349-379 (1972)
Kallman, K.: Genetics and geography of sex determination in the poe-
ciliid fish, Xiphophorus maculatus. Zoologica 50, 151-190 (1965).
Lillie, F.R.: Development of the Chick. 3rd ed-.-(revised by H.L.
Hamilton). New York: Henry Holt 1952.
Lyon, M.F.: X-chromosome inactivation and developmental patterns in
mammals. BioI. Rev. 47, 1-35 (1972).
McClung, C.E.: The accessory chromosome - sex determinant? BioI. Bull.
3, 43-84 (1902).
Mittwoch, v.: Genetics of Sex Differentiation. New York-London: Aca-
demic Press 1973.
Mittwoch, V., Buehr, M.L.: Gonadal growth in sex reversed mice. Differ-
enciation 1, 219-224 (1973).
Mittwoch, V.,-Narayanan, T.L., Delhanty, J.D.A., Smith, C.A.B.: Gonadal
growth in chick embryos. Nature (New BioI.) 231, 197-200 (1971).
Morris, T.: The XO and OY chromosome constitution-in the mouse. Genet.
Res. 1.1., 125-137 (1968).
446
Ohno, S.: Sex Chromosomes and Sex-linked Genes. Berlin-Heidelberg-

New York: Springer 1966.
Penrose, L.S.: Finger-print pattern and the sex chromosomes. Lancet 1,
298-300 (1967).
Pinnock, L.A., Tinkle, D.W., Shaw, M.W.: Minute Y chromosome in the
lizard genus Uta (family Iguanidae). Cytogenetics (Basel) .§., 9-19
(1969) .
Polani, P.E.: Hormonal and clinical aspects of hermaphroditism and the
testicular feminizing syndrome in man. Phil. Trans. Roy. Soc. Lond.B
259, 187-204 (1970).
Ray-Chaudhuri, S.P., Singh, L.: DNA replication pattern in sex chromo-
somes of snakes. Nucleus 12, 200-210 (1972).
Ray-Chaudhuri, S.P., Singh, L., Sharma, T.: Evolution of sex-chromo-
somes and formation of W-chromatin in snakes. Chromosoma 33, 239-
251 (1971). -
Reinboth, R.: Intersexuality in fishes. Soc. Mem. Soc. Endocrinol. ~,
515-543 (1970).
Schmid, W.: DNA replication patterns of the hetero-chromosomes in
GaUus domesticus. Cytogenetics 1, 344-352 (1962).
Schmid, W., Leppert, M.F.: Karyotyp, Heterochromatin und DNS-Werte bei
13 Arten von Wlihlmausen (Microtinae, Mammalia-Rodentia). Arch.
Julius Klaus-Stift. Vererbungsforsch. Sozialantropol. Rassenhyg. 40,
35-49 (1968). -
Singh, L., Ray-Chaudhuri, S.P.: DNA replication pattern in the chromo-
somes of Crocodylus palustris (Lesson). Nucleus 16, 33-37 (1973).
Singh, R.P., Carr, D.H.: The anatomy and histology of XO human embryos
and fetuses. Anat. Rec. 155, 369-383 (1966).
Wolff, Et., Wolff, Ern.: The effect of castration on bird embryos. J.
expo Zool. 116, 59-67 (1951).
and D.J. Randall), vol. III, pp. 117-175. New York: Academic Press
1969.
Subject Index
Ablation experiments, in Chromosomal constitution of sex

birds 365ff., 379 f. (s. also Ploidy), in
crustaceans 99ff. amphibia 311ff.
gastropods 164, 174ff. crustaceans 92f., 102, 112f.
mammals 433-435 insects 146ff.
nemertines 34ff. invertebrates 438ff.
oligochaetes 73ff. mammals 393f., 422f.
planarians 20ff. reptiles 340ff.
polychaetes 50ff. teleosts 239, 243ff.
teleosts 236ff. vertebrates 438ff.
Actinomycin D 318ff. Coelenterates 1ff.
Adenohypophysis, in Crustaceans 91ff., 106fr.
mammals 416ff., 430f. Crustecdysone 99
teleosts 207, 210ff. Cyproterone acetate 408ff.
Amphibia 311ff., 318ff.
Androgenic hormone, in
crustaceans 91ff., 117f.
gastropods 164f. Diandry in teleosts 252, 263ff.,
Androgens, s. Sex steroid 305f.
hormones Dichromatism in teleosts 265ff.
Annelids 41-83 Distribution in natural habitat,
Anti-androgen, s. Cyproterone in
acetate Anthias 284ff.
Arrhenogenous specimens, in Rivulus 252
amphibia 311ff. scarids 272ff.
crustaceans 108ff. Dorsal bodies in gastropods 162
Asexual reproduction, in
coelenterates 1ff.
planarians 14ff., 20ff.
Echinoderms 188ff.
Echiurids 84ff.
Estrogens, s. Sex steroid
Behavior, reproductive, in hormones
Anthias 286ff. Evolutionary opportunism among
scarids 278ff. hermaphrodites 263f., 278ff.,
Birds 356ff., 375ff., 382ff. 306f.
Brain, s. Neurosecretion Eye-stalk factors in crustaceans
99f.
Castration, s. Ablation
experiments Fishes 201-310
Castration, para$itic 121ff. Fissiparous reproduction in
Catecholmanines,effects in planarians 14ff., 23ff.
Oligochaetes 80ff. Freemartins, bovine 392ff.
Chimeras, in Free-martinism, in
coelenterates 6ff. birds 382ff.
nemertines 30ff. nemertines 33ff.
448
Gastropods 158ff., 170ff. Nemertines 30ff.

Grafting experiments, in Nervous system, s. Neurosecretion
amphibia 312f. Neurosecretion, in
birds 375ff., 386 coelenterates 11
coelenterates 6ff. crustaceans 99ff.
crustaceans 100ff., 114ff. gastropods 162, 167, 171ff.
echinoderms 192ff. lamellibranchs 186
gastropods 165f., 174ff. nemertines 34ff.
mammals 431-433 oligochaetes 57ff., 68ff.,
nemertines 30ff. 79ff .
oligochaetes 73ff. planarians 19, 20ff.
planarians 15ff.
polychaetes 53ff.
teleosts 236ff., 253
Gonadotropins, s. Adenohypophysis Oligochaetes 57ff., 64ff., 72ff.
Gynandromorphism, in Organ culture in vitro, in
crustaceans 101f. echinoderms 162
nemertines 32f., 36, 38f. gastropods 164f., 170ff.
lamellibranchs 182, 185f.
oligochaetes 64ff.
Oxygen supply, effects in
Heterochromatinization 140f., 440 gastropods 166f.
Heterochromosomes, s. Chromosomal
constitution of sex
Heterogamety, s. Chromosomal
constitution of sex Parabiosis, in
Homogamety, s. Chromosomal coelenterates 6ff.
constitution of sex nemertines 30ff.
Hybridization Parasite, influence on sex of
influence on sexual development host in
in birds 356ff. crustaceans 115
role for evolution of partheno- insects 121ff.
genesis in reptiles 348ff. lamellibranchs 183, 185
Hypothalamus in teleosts 242
mammals 416-419, 430ff. Parthenogenesis in reptiles
teleosts 215-217 340ff.
Pesticides, effects on sexual
development in birds 387ff.
Phenotypical sex determination
Illumination effects on sexual in BoneUia 84ff.
development in teleosts 250 Pheromones in polychaetes 55
Inhibitors of steroid hormone Phylogenetic relationship of
biosynthesis 422 hermaphroditic fishes 295ff.
Insects 121ff., 135ff., 146ff. Pituitary, s. Adenohypophysis
Interstitial cells, in Planarians 14ff., 20ff.
coelenterates lff. Ploidy, relation to sex devel-
mammals 399, 403 opment in
teleosts 211, 215, 237 amphibia 312ff.
insects 140ff.
teleosts 247f.
Polychaetes 41ff., 48ff.
Lamellibranchs 179ff. Polyspermy 141
Preputial gland 425ff.
Prolactin in mammals 416ff.
Protandry, in
Mammals 392ff., 407ff., 422ff. asteroids 189ff.
Mammary gland 412ff. crustaceans 94ff.
Molluscs 171-200 gastropods 170ff.
Monandry in teleosts 263ff. polychaetes 42ff., 48ff.
Mullerian duct inhibiting factor Protogyny, in
in mammals 402f., 408f., 414 polychaetes 42ff.
449
Protogyny, in Effects on sex development in

teleosts 201ff., 222ff., amphibia 318ff.
263ff., 284ff. birds 358, 364ff.
Pseudohermaphrodite, rat 422ff. mammals 402ff., 407ff.
Puromycin, effects on sexual teleosts 206ff., 241, 291f.
development in amphibia 318ff. Social interactions, influence
on development of sex, in
BoneUia 84ff.
polychaetes 49ff.
Receptor proteins of preputial teleosts 289-291
gland 428-430 Symbionts, influence on sexual
Regeneration experiments, in development in insects 138ff.
coelenterates 6ff.
j1emertines 30ff.
planarians 20ff.
polychaetes 47 Teleosts 201ff., 222ff., 228ff.,
teleosts 236ff. 236ff., 243ff., 249ff.,
Reptiles 332ff., 340ff. 263ff., 284ff., 295ff.
Temperature, effects in
chelonians 332ff.
crustaceans 108ff.
Self-fertilization in gastropods 165f.
insects 137 polychaetes 44ff.
teleosts 249f. teleosts 250ff.
Sex center, hypothalamic 417ff., Tentacular humoral factors in
430f. gastropods 163ff., 171ff.
Sex chromatin 243ff., 440 Testis ova in teleosts 228ff.,
Sex chromosomes, s. Chromosomal 256f.
constitution of sex Thelygenous specimens, in
Sex inversion amphibia 311ff.
definition of 439 crustaceans 108ff.
occurrence in Transplantation, s.- Grafting
amphibia 318ff. experiments
coelenterates 5ff. Trophic factors, effects on
gastropods 170ff. development of sex in
lamellibranchs 180 oligochaetes 77ff.
mammals 441f.
oligochaetes 62, 65ff., 74ff.
polychaetes 41ff., 48ff.
teleosts 201ff., 236ff., X-factor, s. Mullerian duct
263ff., 284ff. inhibiting factor
Sex steroid hormones X-irradiation, effects in
Biosynthesis in coelenterates 9ff.
birds 361ff. echinoderms 192
mammals 422ff. teleosts 439f.
teleosts 203ff.
Theoretical and Applied Genetics
Internationale Zeitschrift ffir Theoretische und
Angewandte Genetik Continuation of "Der Ziichter"
founded in 1929
Managing Editor: H. Stubbe

Editorial Board: H. Abplanalp, L. AlfOldi, R.W. Allard,
E. Andersson, B.L. Astaurov, S. Barbacki, 1.S.F. Barker
D.K. Belyaev, A. Gustafsson, R. Hagemann, H. Kappert,
H. Kihara, H. Kuckuck, R.C. Lewontin, W.l. Libby,
H.F. Linskens, 1. Mac Key, F. Mechelke, G. Me1chers,
B.R. Murty, R. Riley, A. Robertson, W. Seyffert,
H. Skjerv~ld, H. Stubbe, L.D. Van Vleck
Breeding genetics, with the aid of chemistry and mathe-

matics, has become considerably more fundamental and
general. This development has moved from the genetics
of the individual to that of the group and, in turn, to the
study of the evolution and origin of domesticated species.
Improved mathematical models, which can be quantita-:
tively solved or simulated by the computer, allow the
new science of molecular genetics to study gene-enzyme
interaction and the regulation of inherited characteristics.
TAG fills a vital need for detailed research reports in this
field and serves as an international vehicle for the
exchange of scientific information.
The journal publishes original papers and reviews on
the following:
Mathematical Genetics. Analysis of Genetic Models
Evolutionary Genetics. Genetics and Biochemistry of
Development. Population Genetics. Breeding Methods
'Physiology of Qualitative and Quantitative Characteristics
Physiology of the Gene. Cytogenetics
Sample copies as well as subscription and back-volume
information available upon request.
Please address: or Springer-Verlag

Springer-Verlag Springer-Verlag New York Inc.
Berlin Werbeabteilung 402 I Promotion Department
Heidelberg D 1000 Berlin 33 175 Fifth Avenue
New York Heidelberger Platz 3 New York, N Y 10010
Monographs Vol. 1: S. Ohno Vol. 5: J. Muller
Regulation of Aldosterone
Sex Chromosomes
on Endocrinology and Sex-linked Genes Biosynthesis
Editors: F. Gross, 33 figures. X, 192 pages. 1967' 19 figures. VII, 137 pages
A. Labhart, M.B. Lipsett, ISBN 3-540-03934-1 1971
Cloth DM 46,- ISBN 3-540-05213-5
T. Mann, L.T. Samuels,
ISBN 0-387-03934-1 Cloth DM 44,-
J. Zander ISBN 0-387-05213-5
(North America) Cloth
Cloth $14.80 (North America)
Cloth $13.80
Vol. 2: K.B. Eik-Nes,

E.C. Horning Vol. 6: K. Federlin
Gas Phase Chromatography Immunopathology of Insulin
of Steroids Clinical and Experimental
85 figures. XV, 382 pages Studies
1968 53 figu res. X III, 185 pages
ISBN 3-540-04277-6 1971
Cloth DM 46,- ISBN 3-540-05408-1
ISBN 0-387-04277-6 Cloth DM 60,-
(North America) ISBN 0-387-05408-1
Cloth $14.80 (North America)
Cloth $19.90
Vol. 3: F.G. Sulman

Hypothalamic Control Vol. 7: E.W. Horton
of Lactation Prostaglandins
I n collaboration with 97 figures. XI, 197 pages
M. Ben-David, A. Danon, 1972
S. Dikstein, Y. Givant, ISBN 3-540-05571-1
K. Khazen, J. Mishkinsky- Cloth DM 64,-
Shani, I. Nir, C.P. Weller ISBN 0-387-05571-1
58 figures. X II, 235 pages (North America)
1970 Cloth $23.80
ISBN 3-540-04973-8
Cloth DM 64,-
ISBN 0-387-04973-8 Vol. 8: E. Gurpide
(North America) Tracer Methods in Hormone
Cloth $19.60 Research
35 figures. XI, 188 pages
1975
Vol. 4: U. Westphal ISBN 3-540-07039-7
Steroid-Protein Interactions Cloth DM 72,-
144 figures. XIII, 567 pages ISBN 0-387-07039-7
1971 (North America)
ISBN 3-540-05312-3 Cloth $29.60
Springer-Verlag Cloth DM 105,-
Berlin ISBN 0-387-05312-3
(North America)
Heidelberg Cloth $29.90 Prices are subject to change
New York without notice

1975 Book IntersexualityInTheAnimalKingd

Uploaded by

Copyright:

Available Formats

1975 Book IntersexualityInTheAnimalKingd

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1975 Book IntersexualityInTheAnimalKingd

Uploaded by

Copyright:

Available Formats

Intersexuality

in the Animal Kingdom

With 221 Figures

ISBN-13 978-3-642-66071-9 e-ISBN-13: 978-3-642-66069-6

Library of Congress Cataloging in Publication Data. Symposium on Inter-

When Richard Goldschmidt' coined the term "intersexuality" in 1915,

Among all the anatomical and physiological features of living organisms

Recent developments in current biological research have brought the

Whatever is understood by intersexuality, the term always implies

'Biol. Zbl. 35, 565 (1915).

and objectives of current research on intersexuality. No claim is

At first sight the heterogeneity of the contributions is rather

Despite such unevenness, there was unanimous agreement among the

No single theme is likely to emerge as the "essence" of this book,

3Studia Mendeliana Brlinn, 203 (1923).

The Symposium on Intersexuality in the Animal Kingdom was held in the

To all persons and institutions who lent us their support I am deeply

March 1975 R. Reinboth

Sex and Sex Determination in Coelenterates •••.•..•.••••.••....••

Sexual Differentiation in the Fissiparous Strain of Dugesia

Sexuality and Neurosecretion in Freshwater Planarians •.•••.•.••• 20

Sex Differentiation in Regeneratingd' I ~ Nemertine Chimeras ••.••• 30

Sex Reversal in the SyUinae (Polychaeta: Annelida) 41

Mutual Influence on the Sexual Differentiation in the Protandric

Neurosecretory Phenomena during Reproduction in Oligochaeta .••.. 57

Study of Sex Control of Gametogenesis by Organ Culture in the

Factors of Primary Sexual Differentiation in the Simultaneous

Sex-Determination in BoneUia ••••••••••••••••••.••.••••••••••.••• 84

Hermaphroditism and Gynandromorphism in Malacostracan Crustacea. 91

Temperature-Sensitive Intersexuality and Its Determinism in

Parasite-Induced Castration and Intersexuality in Insects ••.••.. 121

Sex Specific Cell Differentiation in Different Types of

Structural and Endocrinological Aspects of Hermaphroditism in

Protandric Hermaphroditism in Prosobranch Gastropods ........... 170

The Sexuality of Pelecypod Molluscs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

Hermaphroditism in Echinoderms. Studies on Asteroids ............ 188

The Gonadal and Adenohypophysial Functions on Natural Sex

Ovarian and Testicular Intersexuality in Two Protogynous

Hermaphroditism in Cichlid Fishes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

An Unusual Approach to Experimental Sex Inversion in the Teleost

The Possible Significance of Sex-Chromatin for the Determination

Sex Determination and Differentiation among Uniparental

Protogynous Hermaphroditism in Fishes of the Family Scaridae .... 263

Ecology and Physiology of Sex Reversal in Anthias squamipinnis

The Evolution of Hermaphroditism in Fishes . . . . . . . . . . . . . . . . . . . . . . 295

Unisexual Female Offsprings in the Salamander, PZeurodeZes waUUi

Autoradiographic Study on the Mechanisms of Testosterone-

Temperature and Sex Differentiation in Embryos of Two Chelonians,

Evolution of Parthenogenetic Species of Reptiles . . . . . . . . . . . . . . . . 340

Intersexuality in Birds. Study of the Effects of Hybridization

Some Effects of Embryonic Gonad and Non-Gonadal Grafts on the

Intersexuality of the Genital System and "Free-Martinism" in

The Prenatal Development of Bovine Freemartins . . . . . . . . . . . . . . . . . . 392

The Role of Androgens in Sexual Differentiation of Mammals ...... 407

Recent Studies on the Intersexual Programming of the Genetic Rat

Chromosomes and Sex Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447

BIERNE, J., Prof., Faculte des Sciences de l'Universite de Reims,

BRIANCON, C., Dr., Faculte des Sciences, Laboratoire de Biologie

BRUSL~, Jacques, Prof., Centre Universitaire de Perpignan, Faculte des