PRESENTED TO

Marine Biological Library

w

ITH THE COMPLIMENTS OF

The Williams & Wilkms Company

BALTIMORE 2, MARYLAND

Dick M. Hoover

THIRD EDITION

SEX AND

INTERNAL
SECRETIONS

VOLUME I

VOLUME I

CONTRIBUTORS

A. Albert
David W. Bishop
Kichard J. Blandau
K. K. Burns
A. T. Cowie
John W. Everett
S. J. 1 olley
Thomas II. Forbes
J. W. (iowen

Koy O. Greep
A. M. Guhl
Joan G. Hampson
John L. JIampson
Frederick L. llisaw
Frederick L. llisaw, Jr.
James II. Leathem
Daniel S. Lehrman
Margaret Mead
John W. Money

Helen Padykula

Dorothy Price

Herbert D. Purves

Ari van J'ienhoven

Claude A. Villee

H. (iuN Williams- Ashman

(ieorge B. Wislocki

William C. Young

M. X. Zarrow

Baltimore • 1961

THIRD EDITION

SEX AND

INTERNAL
SECRETIONS

W^i^-

Edited by William C. Young, Ph.D. :^

Professor of Anatomy, University of Kansas, Lawrence

Foreword by George W. Corner, M.D., D.Sc.

Director Emeritus, Department of Embryology,
Carnegie Institution of Washington

The Williams & Wllkins Co

Publication was supported in part by
Public Health Service Research Grant M-464S from
the National Institute of Mental Health,
Public Health Service.

Copyright ©, 1961

The Williams & Wilkins Company

Made in the United States of America

Library of Congress
Catalog Card Number
60-12279

COMPOHKI) AND PRINTKD BY THE

WAVrCRLV PRKSS, INC.

BAL'I'IMOUFC 2, MARYLAND, U.S.A.

To the Memory of

Robert Mearns Yerkes

pp* ■'»<^

"^ CONTENTS

J oliime I

Foreword. George W. Corner i^

Edgar Allen. William C. Youny xiii

Preface to Third Edition x\i

Preface to First Edition xxiii

section a
Biologic Basis of Sex

1. Cytologic and Genetic Basis of Sex. /. W. Gowen 3

2. Role of Hormones in the Differentiation of Sex. R. K. Burns 76

section b

The Hypophysis and the Gonadotrophic Hormones in Relation

TO Reproduction

o. Morphology of the Hypophysis Related to Its Function. Herbert D. Purves . ... 161

4. Physiology of the Anterior Hypophysis in Relation to Reproduction. Roy

b. Greep ' .240

section c
Physiology of the Gonads and Accessory Organs

5. The Mammalian Testis. .4 . Albert 305

(). The Accessory Reproductive Glands of Mammals. Dorothy Price and H. Guy

Williams- Ashman 366

7. The Mammalian Ovary. William C. Young , 449

8. The Mammalian Female Reproductive Cycle and Its Controlling Mechanisms.

John W. Everett .^ 497

9. Action of Estrogen and Progesterone on the Reproductive Tract of Lower Primates.

Frederick L. Hisaw and Frederick L. Hisaw, Jr 556

10. The Mammary Gland and Lactation. A. T. Cowie and S. J. Folley 590

1 1 . Some Problems of the IMetabolism and Mechanism of Action of Steroid Sex Hor-

mones. Claude A . Villee 643

12. Nutritional Effects on Endocrine Secretions. James H. Leathern 666

\ olume II

section d
Biology of Sperm and Ova, Fertilization, Implantation, the Placenta,

AND Pregnancy

13. Biology of Spermatozoa. David W. Bishop 707

14. Biology of Eggs and Implantation. Richard J. Blandau. 797

15. Histocliemistry and Electron Microscopy of the Placenta. George B. Wislocki

and Helen Padykida 883

16. Gestation. M. X. Zarroiv ''^^^^

section e
Physiology of Reproduction in Submammalian Vertebrates

17. Endocrinology of Reproduction in Cold-blooded Vertebrates. Thomas R. Forbes. . 1035

18. Endocrinology of Reproduction in Birds. Ari van Tienhoven 1088

vii

79220

viii CONTENTS

section f
Hormonal Regulation of Reproductive Behavior

19. The Hormones and Mating Behavior. William C. Yoimg 1173

20. Gonadal Hormones and Social Behavior in Infrahuman Vertebrates. .4. AI. Guhl . . 1240

21. Gonadal Hormones and Parental Behavior in Birds and Infrahuman Mammals.

Daniel S. Lchrman 1268

22. Sex Hormones and Other Variables in Human Eroticism. John W. Money .... 1383

23. The Ontogenesis of Sexual Behavior in J\Ian. John L. Hampson and Joan G.

Hampson 1401

24. Cultural Determinants of Sexual Behavior. Margaret Mead 1433

Index 1481

FOREWORD

George W. Corner, M.D., D.Sc.

DIRECTOR EMERITUS, DEPARTMENT OF EMBRYOLOGY,
CARNEGIE INSTITUTION OF WASHINGTON

Publication of the third edition of Sex
and Internal Secretions signalizes the ac-
complishment of about a half century's
intensive work by investigators of many
countries, among whom those of the United
States have been notably active. Any such
burst of discovery as this rests, of course,
upon a long preceding period of more
gradual progress. Taking as landmarks Reg-
ner de Graaf's recognition that the "female
testis" of mammals is an egg-producing
organ comparable to the ovaries of birds
(1672) and Leeuwenhoek's description of
the spermatozoa (1674), we can trace the
continuous development of knowledge about
the reproductive system down to our own
times. Discovery of the actual mammalian
ovum by Karl Ernst von Baer in 1827
accelerated the progress of research on the
origin of the germ cells, the de^'elopment and
discharge of the Graafian follicle, transport
and fertilization of the ovum, and implanta-
tion and development of the embryo. Such
studies inevitably drew attention to the
cyclic aspects of reproductive function, par-
ticularly from students of animal breeding
and from faunal naturalists, who acquired a
great deal of information about the estrous
cycles of wild as well as domestic animals
and those of the laboratory. The work
of the English leaders in this kind of in-
vestigation, Walter Heape and F. H. A.
Marshall, reached fruition in the latter's
well known "Physiology of Reproduction,"
published in 1910. At this same period
(1890-1910) gynecologists, especially in Ger-
many and Austria, were putting their spe-
cialty on a scientific basis. Becoming aware
of the current advances in knowledge of
embryology and the biology of reproduction
of mammals in general, they were seeking
similar clues to the explanation of the
human menstrual cycle, ostensibly so dif-
ferent from the estrous cycle of domestic
animals. European workers, notably Hitsch-

mann and Adler, Robert Schroedcr, and
Robert Meyer, from about 1900 to the
beginning of the first World War, put to-
gether from operating-room material a histo-
logic description of the human cycle that
became more and more clear as the em-
bryologists related it to their understanding
of the general mammalian cycle. The young
sciences of psychology, psychiatry and an-
thropology also joined the concerted attack
upon the problems of sex and reproduction.
Since about 1870 European psychiatrists, led
by such men as von Krafft-Ebing and Forel,
had been studying sex psychology, with
the aim of understanding behavior of a
kind that was considered abnormal or con-
ducive to social difficulties such as those
created by prostitution and homosexuality.
The way was thus opened for psychology
to investigate the biologic basis of normal
sex behavior. European and American an-
thropologists had begun to document and
analyze the sex attitudes of primitive races
and distant nations, and even of their own
peoples. Nor must we forget the influence
of the Women's Rights movement, with its
fight against all forms of bondage of women
and its emphasis on standards of sex be-
havior equally applicable to both sexes.
All these new sciences and new social move-
ments called for better understanding of
basic sex physiology, which only biologists
could provide.

Thus at the beginning of the 20th century
and during the next decades investigation
in this field became more intense. Na-
turalists, animal breeders, histologists, em-
bryologists and gynecologists gradually came
to understand each other's problems, and
began a period of rapid advance not yet
ended nor even slowed down, in which
scarcely a year has passed without major
contributions.

American zoologists were already prepared
by their embryologic studies to take part

FOREWORD

in this exploration, and the rapid devc^lop-
ment of medical research in the United
States in the early part of the 20th century
provided speciahsts in human embryology,
pathologists, physiologists, and biochemists
who were ready to join the biologists in
such investigations. When in 1922 the Na-
tional Research Council was called upon by
influential groups centered in the American
Social Hygiene Association, to bring to-
gether existing knowledge and to promote
research upon human sex behavior and
reproduction, our nation already possessed
a corps of competent investigators who
rallied to the call of Robert M. Yerkes and
Frank R. Lillie, forming the Committee for
Research in Problems of Sex. This Com-
mittee, with financial support from the
Rockefeller Foundation, successfully under-
took to encourage research on a wide range
of problems of sex physiology and behavior.

The younger readers of this book will
hardly be a])le to appreciate the full sig-
nificance of such an alliance between bi-
ologists, psychologists, and physicians on
one hand, and social philanthropists on the
other. It represented a major break from
the so-called Victorian attitude which in the
English-speaking countries had long im-
peded scientific and sociologic investigation
of sexual matters and had placed taboos on
open consideration of human mating and
childbearing as if these essential activities
were intrinsically indecent. To investigate
such matters, even in the laboratory with
rats and rabbits, required of American
scientists, including some of the contribu-
tors to the first edition of Sex and Internal
Secretions, a certain degree of moral stamina.
A member of the Yerkes Committee once
heard himself introduced by a fellow scientist
to a new ac(}uaintance as one of the men
who had "made sex respectabl(\" Ccrlaiiily
the prestige of lh<^ Committee and the
successes of American in\'(>stigators working
with and without its assistance helped to
bring about a more I'ealistic attitude toward
sex research, although reactions from some
ciuarters to such important recent work as
that of the late Alfred C. Kins(\v show that
the battle is even yet not fully won.

Ten years after its formation tiie Com-
mittee for Reseai'ch in Problems of Sex,

proud of the achievements it had helped to
foster, sponsored the first edition of Sex
and Internal Secretions. The information
thus brought together in 1932 came largely
from research in genetics, cytology, em-
bryology, and endocrinology, almost ex-
clusively utilizing morphologic methods of
study. In contrast to the present situation
as reflected in the third edition, biochemistrv
of the sex glands was still in an elementary
stage, having barely achieved the pre-
liminary chemical identification of the ovar-
ian, placental, and testicular hormones;
and the psychology of sex behavior was
only beginning to develop its experimental
methods. The sum total was, however, a
deeply impressive record of progress that
drew many new workers into this fi(^ld of
research.

It would be difficult to ascribe priority
in this achievement to any one of the
biologic disciplines. Genetics and cytology
had provided one of the major clues by C. E.
jMcClung's discovery in 1902 of the sig-
nificance of the accessory chromosome, and
his brilliant conjecture that this minute
fragment of protoplasm is related to the
determination of sex. Coming just before
an outburst of discoveries concerning the
chromosomal mechanism of heredity, based
largely on the fruit fly Drosophila, the
concept of genetic determination of sex
gave rise to an immense amount of in-
vestigation and theorizing about the way
in which a developing individual is caused to
become either male or female. Three decades
after the publication of McClung's hypoth-
esis, enough information on this (luestion
was in hand to fill two ci'owded chapters
in the first edition of Si.r and I idcriud
Secretions.

The contribution of cmbryolog}' to our
sul)ject goes back to the ISth and 19th
centuries and in particulai' to the (h^scrip-
tion of the early stages of (lc\('loi)nient of
the internal reproductive system, with which
the names of Kaspar Fried rich Wolff' and
Johannes Aliiller are ind(>libly associated.
In this field, too, a period of acliv(> in-
vestigation began early in the 20tli century,
with th(^ aid of improxcd methods of microt-
omy and the apj^lication of pr(H'is(> histo-
loiiic staininti to eiubi'vonic tissues. The

FOREWORD

development of the gonads and the meta-
morphosis of the Wolffian and Miillerian
ducts into the secondary internal reproduc-
tive organs was rapidly worked out in
animals of every vertebrate order, by an
army of in\'estigators too numerous to
mention in a short resume. The story of the
first appearance of primordial germ cells
and their migration through the tissues of
the embryo to the newly forming gonads,
adumbrated in the 1880's by the work of
Semon, was confirmed and extended to
several species of mammals. If in these
latter creatures and in the human species,
th(> complete line of descent from the
fertilized ovum to the first appearance of
the germ cells is till not as clear as in many
lower forms, enough at least was discovered
within a few decades to indicate an essential
similarity in the history of the germ cells
in all vertebrates.

To the embryologists of Europe and
America w^e owe in large part also the
successful analysis of the mammalian repro-
ductive cycle that has been achieved during
this half century. In order to procure
mammalian embryos of known age the time
of ovulation and fertilization had to be
related to the outward manifestations of the
estrous cycle. Comparati^'e description of
the cycles of the various mammals for this
purpose was climaxed by the discovery, or
rather rediscovery and practical application
of cyclic changes in the vaginal epithelium
by C. R. Stockard and G. X. Papanicolaou.
The vaginal smear method, thus introduced
to the experimental laboratories, made pos-
sible a wide range of investigations on the
physiology and biochemistry of the cycle
and the ovarian hormones. As apphed to the
white rat by Herbert M. Evans and J. A.
Long it became a basic tool in such studies.

Another influence which also greatly for-
warded investigation of the ovarian cycle
has already been mentioned. This was the
efTort of the gynecologists and especially
gynecologic pathologists to interpret cychc
events in the human ovary and uterus. One
of the most notable American discoveries,
that of the influence of the corpus luteum
in decidua formation by Leo Loeb, stemmed
from his familiarity with the German studies
on the human cvcle. Much of our knowledge

of th(^ corpus luteum and its hormone,
progesterone, was in fact won by investiga-
tors who approached the problem through
gynecology.

Had it not b(>en for the first World War,
moreover, European gynecologic experi-
menters might have attained clear knowl-
edge of the estrogenic hormones, for even
before 1900 Emil Knauer and Josef Halban
of Vienna had demonstrated in a preliminary
w^ay the endocrine dominance of the ovaries
over the uterus, and by 1913 various in-
vestigators, notably Henri Iscovesco of Paris
and Otfried Fellner of Vienna, had pre-
pared crude extracts which we now know
contained estrogens. It remained, however,
for the American zoologist-anatomist Edgar
Allen and his biochemical colleague E. A.
Doisy, e(iuipped with the vaginal smear
method of testing ovarian hormone action,
to isolate an estrogen from the fluid of the
Graafian follicles, thus starting an era of
ovarian endocrinology which has ultimately
resulted in clear definition and discrimina-
tion of estrogens and progestins and their
respective effects upon the uterus and other
organs of the reproductive system. Applying
this new knowledge to the complexities of the
human reproductive cycle, the zoologists,
embryologists, gynecologists and endocri-
nologists, among them several distinguished
contributors to the first edition of this
work, have combined forces to work out a
clear account of the endocrine basis of
menstruation and the implantation of the
primate embryo.

The parallel story of the hormones of the
testis can be read in the successive editions
of Sex and Internal Secretions. Berthold's
proof, published in 1849, that in fowls the
testis presides over the development of the
cock's comb, wattles, and spurs ultimately
led to the isolation of the first known
androgen in F. C. Koch's laboratory at
Chicago, and thence to the development of
a great body of knowledge about androgenic
steroids.

The more complex history of the hor-
mones of the hypophysis, becoming some-
what clearer in each successive edition of
this work, well illustrates a main theme of
this introductory essay, namely the de-
pendence of scientific advance upon the

FOREWORD

intermingling of ideas from various fields.
It is not merely by chance that among the
American workers on the hypophysis two
names stand out, those of a surgeon, Harvey
Gushing, and a zoologist-anatomist, Philip
E. Smith.

The central achievement of this half
century of intensive work can be summarized
in one sentence. It was, first, recognition,
description and explication of the reproduc-
tive cycle of mammals and man; and,
second, identification of the chemical sub-
stances that serve to integrate the cycle and
preside over gestation. Those who took part
in these investigations recognized, of course,
that in due time their work must be ex-
tended in two directions, downward to the
domain of molecular chemistry and ul-
timately of ionic physics in order to un-
derstand the basic nature of hormone action,
and upward to the field of animal and
human behavior, where sex gland hormones
join with other forms of bodily and mental
integration in directing the life and be-
havior of the organism.

Once the histologists and embryologists
had identified the sex gland hormones it
was inevitable that further investigation of
these remarkable substances should be taken
over by biochemists, as can be seen from the
successive editions of this book. The chief
unsolved problems now demanding atten-
tion relate largely to the sites of action of
the hormones and the precise molecular
effects which they exert upon their target
organs. Recent indications that estrogens
take part in hydrogen transfer in the citric
phase of carbohydrate metabolism, and that
progesterone affects uterine muscle cells
by altering their permeability to potassium
ions, show clearly that there is hope of
understanding th(^ action at molecular level
of these remarkably specific and powerful
substances, which were bai'ely beginning
to be known when the first edition of this
book appeared in 1932.

As the investigators of the future leai'n
exactly where the sex gland hormones exert
their chemical action, and just what they
do to the fundamental elements of th(>
cells of their target organs, we may con-
fidently expect ever-increasing knowledge
even of the most complex sexual and repro-
ductive activities. Running over the chapter

headings of this third edition of Sex and
Internal Secretions, we see indeed that al-
most every author concerns himself with one
or another aspect of sex and reproduction in
the light of what is already known about
endocrine regulation. Chapter after chapter
deals with cyclic events determined by sex
gland hormones. Phenomena which in the
first edition could be explained only sup-
positionally on a hormonal basis, for ex-
ample the "free martin" state in domestic
cattle, and menstruation in primates, are
now much better understood. Others which
seemed hardly within the scope of endo-
crinology are now seen to be in some degree
influenced by hormone action, and thus
to call for discrimination between endocrine
effects and other types of regulation such
as gene action and control through the
nervous system. The experimental embry-
ologists, for example, seek to understand
the respective effects of genes and hormones
in determining the development of the
internal accessory sex organs; students of
animal psychology likewise are beginning
to discriminate between the action of hor-
mones and neuro-psychologic factors in de-
termining the patterns of sex behavior.

Among the subjects discussed in this
book, only psychiatry and anthropology are
as yet not greatly influenced by our recently
acquired stores of endocrinologic informa-
tion. The complex, high-level patterns of
human thought and behavior with which
these sciences deal are presumably far less
subject to chemical regulation than to the
integrative control of the nervous system
as it affects learning, memory, and racial
tradition. Yet when we consider the extent
to which daily life and ethnic customs are
bound up with the sexual and reproductive
activities of mankind, we are prepared to
find this edition of Sex and Internal Secre-
tions not only advancing greatly beyond its
predec(>ss()rs in the study of animal be-
havioi-, but also looking forward, through
exploi'niory chaptcn-s on psychiatric and
anthi'opologic asix'cts of liuninn s(\\ Ix^-
liavior, to a time when we shall more fully
understand the interrelations of all the con-
trolling factors even of these most complex
activities, upon which the continuation and
renewal of life dcp(>nd.

EDGAR ALLEN

1892-1943

Soon after his untimely death, February
3, 1943, many of the important details of
Edgar Allen's life were recorded by col-
leagues who were close to him. Separated
from these memorials, however, was Sex
and Internal Secretions, understandably the
most permanent and tangible memorial.
It is appropriate, therefore, that in this
long-delayed third edition, much of the
material in those records of his life should
be combined with the review of the field
in which his substantive contributions and
directive thought were so important. With
the permission of Doctors George W. Corner
and William U. Gardner, portions of their
biographical sketches have been used here.
A few minor errors have been corrected
and supplementary information has been
added when it was felt that the picture of
Edgar Allen would thereby be enlarged and
sharpened. For much of the latter, in-
debtedness is expressed to Doctor Charles
H. Danforth, a long-time friend and senior
colleague at Washington University, and
to Doctor J. Walter Wilson, with Allen as
a graduate student at Brown University.

Doctor Allen ("Ed" to his many friends,
and "The Skipper" in his department at
Yale) was born at Canyon City, Colorado,
May 2, 1892. He was the son of a physician
about whom little seems to be known. The
Allen family moved to Providence when
he was very young. His early training was
obtained in the public schools of Providence
and at Brown University. Immediately after
his graduation in 1915 he started graduate
study in biology. This was interrupted
two years later by World War I, but was
sufficient to fulfill the not too rigorous
requirements for a master of arts degree.
His record during this period does not seem
to have been impressive and nothing that
has been learned about it foreshadowed his
later distinguished accomplishments. How-
ever, one hitherto unrecorded experience.

mentioned on one occasion to the present
biographer, may have had unusual sig-
nificance in the years that followed. Doctor
Albert Davis Mead, to whom editions 1
and 2 were inscribed, was Professor of
Biology and instructor in the course in
vertebrate embryology. At that time and
for many years later the uteri, tubes, and
ovaries of pregnant sows were collected
from the local slaughterhouse and dissected
by the class. Allen, probably as an assistant
in the course, visited the slaughterhouse
where his attention was attracted by the
numerous large follicles in most of the
ovaries. The curiosity thus engendered seems
to have been the extent of his interest in
reproductive physiology while he was at
Brown, but it could have been instrumental
in directing his attention to the ovary a
few years later in St. Louis, and it could
have prepared him to seek the sow's ovaries
as a source of follicular fluid when he was
desirous of obtaining large quantities of it
for his first tests on spayed mice.

In May, 1917, he volunteered for service
in the Brown University Ambulance Unit.
Later he transferred to a mobile unit of the
Sanitary Corps with which he served in
France. When he was discharged in Feb-
ruary, 1919, he held a commission as second
lieutenant.

Before leaving for France in 1918 he
married Marian Pfeiffer, a fellow student
enrolled in Pembroke College, the Women's
College in Brown University. Throughout
the balance of his life she was his devoted
companion. She too died as a relatively
young woman and did not long survive him. •
There are two daughters. For a man in his
position, he lived modestly. It is easy to
imagine that he valued the warmth and
affection of his family and friends and his
boat abo\T other luxuries he might have
had.

When he returned to civilian life he had

EDGAR ALLEN

no permanent po.sition in sight, but he must
have sought the help of Mead, for during
the summer of 1919 he was an investigator
in the laboratory of the U. S. Fish Com-
mission at Woods Hole. Doctor H. C.
Bumpus, an older colleague of Mead's, had
been Director of the Biological Laboratory
in the Fish Commission at Woods Hole
and summer appointments, paying two or
three himdred dollars, were a source of
help to the graduate students at Brown in
that period. The summer seems to have been
important, not because of any research
Allen did, but because it was then that
Charles Danforth, who was at Cold Spring
Harbor for the summer and had heard
Doctor FL E. Walter speak highly of Allen,
wrote him and called his attention to the
instructorship then open in Washington
University School of Medicine. Danforth
suggested that Allen communicate with
Doctor Robert J. Terry, head of anatomy
in that institution. He must have done so
promptly and been accepted, with the un-
derstanding that graduate study would hv
continued. Danforth's recollection of his
first sight of Allen is repeated:

"When I returned to St. Louis in
the fall and went up to the anatomy
department I saw a man in the hall
whose white liair and impressive bear-
ing led me to suspect that he was prob-
ably a distinguished alumnus returning
for a visit. I soon learned, however,
that this was Mr. Allen, who had been
appointed instructor in anatomy and
was already installed in an office on the
third floor."

Little time seems to have been lost in
starting his work for the Ph.D., although
the circumstances under which the choice
of a problem was made ar(^ somewhat
ohscuiv. Doctor H. H. Willier who met
Allen for- the first time that sunnner (prob-
ably on "stony beach") does not remembei-
that he mentioned any special interest in
the physiology of sex and reproduction.
Danforth, who saw nuich of him from this
time on, l;elie\-(>s that two cii'cunistaiices
may lia\-e licen important. His oflice was
on the floor with that of Doctoi' I.co Locb,
always a stimulating p(M-soii, and in the

animal quarters above was a colony of
mice which had been developed for use in
what was perhaps the first course in em-
bryology to be based exclusively on mam-
malian material — gametogenesis, follicular
growth, ovulation, fertilization, cleavage,
etc. He probably discussed problems with
Loeb and he must have read the recent
paper of Stockard and Papanicolaou in
which changes in the vaginal epithelium
in the guinea pig were correlated with the
o^'arian cycle. Whether he was sensitive to
the generally increasing interest in reproduc-
tive phenomena or was influenced more by
the fact that the mouse had been in-
adequately studied and was right at hand
and ready is not known. The latter pos-
sibility would have been consistent with
his temperament and the way he worked.
On the other hand, the fact that the first
of the three "purposes" stated in his thesis
was "to make possible a more efficient
mating for the collection of embryological
material" may indicate that the larger im-
portance of what he was about to start was
not yet apparent to him.

With the double responsibility of teach-
ing and doing the research for a thesis,
he must have worked incredibly long hours.
But the rewards were great. The observa-
tions recorded in his thesis, "The oestrous
cycle in the mouse," ignited the fire that
was to burn and to be spread during the
remaining years of his life, and to eradicate
fore\'er the diffidence which characterized
him during his earlier graduate years at
Brown. Briefly, he observed that large
follicles were present in the proestrous and
cstrous stages of the cycle, but that ovula-
tion had occurred by the time of the
metestrum. Regressive changes were noted
in the uterus and these were analogized
with menstruation in lower primates and
the human female. The reference on page
111 to Hobin.son's belief that a secretion
tVoni the follicle causes estrous changes
rcxcals how close he was to the hypothesis
that was to rccciNc gencn-al acc(>ptance only
a few years latci'. It is clear, how(>\-ei', that
he was not rcad^' for this simple and direct
conclusion. Instead, he started by rejecting
the suggestion that 1 h(^ growth stimulus

EDCAR ALLEN

to the genital tract comes from the corpus
hiteum and then, after noting that "the
follicles are the only remaining ovarian
possibility," continued, "the presence of
maturing ova in large follicles is the cause
of the prooestrum and oestrus" and "the
renewal of the ova at ovulation (or their
atresia if this fails to occur) is the primary
cause of the degenerative changes of the
metoestrum."

Except for the addition of the active role
of the estrogenic substances contained in
the follicular fluid, demonstrated by him-
self and Edward A. Doisy less than two
years later, the pattern of Allen's thought
for the next 20 years was contained in his
thesis — the cyclic origin of ova from the
germinal epithelium, the primacy of the
ovum, the growth effects of estrogens, the
consequences of their withdrawal, a dis-
counting (the word used in his thesis) of
the importance of the hormone of the
corpus luteum in the regulation of reproduc-
tive phenomena.

Rarely has so much of the important
conceptualization of a pre-eminent scientist
been "roughed in" in his thesis. Also of
interest is the place of the thesis in the
history of American anatomy. It was written
at a time when the emphasis was shifting
from structural anatomy to functional anat-
omy. Rightly or wrongly, there were then,
as there are now, those who feel that this
trend could go too far. Allen seems to have
been caught in this controversy. At all
events, he must have felt compelled to
enlist the help of friends at Brown, for it
was the men there with their orientation
toward biology who seem to ha^'e been
less concerned with the amenities of the
time and to have recommended that he be
awarded the Ph.D. The omission of any
acknowledgment to indi\4duals or to in-
stitutions could have been an understandable
oversight in his haste to test the action of
follicular fluid on the vaginal epithelium
of the mouse, or it could have been a device
for avoiding any embarrassment. It is a
coincidence that 15 years later in an office
at Brown when a younger colleague was
threatening to look into the problem of the
hormonal control of mating behavior, Allen

asked in his characteristically friendly way
and also somewhat paternally, ". . . why
don't you return to your woi-k on the
epididymis?"

With hurdles of the thesis and its publica-
tion iK^iind him, and the conviction that
the follicular fluid contains the substance
he was seeking, he must have thought of
injecting follicular fluid from the large
follicles of the sow (he had seen them at
Brown) into ovariectomized mice and ex-
amining the vaginas for the sequence of
changes he had described in intact mice.
It is clear that the idea was not suggested
to him by anything he read. As he often
said jocularly, we did the work first and
looked up the literature later. The published
statement to this effect (J. Biol. Chem.,
61: 711-727, 1924) was somewhat qualified
but almost as direct, ". . .we paid but little
attention to the papers of the various
workers who have claimed to have demon-
strated active preparations until after our
own first definitely positive results were
obtained." An unconventional approach,
but excuseable perhaps when the hunch is
as "logical" as it was to Allen.

By the early spring of 1923 he had made
promising preliminary tests and a few days
before Charles Danforth left Palo Alto for
the Anatomists' meetings in Chicago, he
received an exultant telegram saying that
he (Allen) had succeeded in inducing estrus
in a spayed mouse by injecting follicular
fluid. Others had come close to the dis-
covery, but the reason they failed was that
no one had had a clear-cut practical test.
Danforth is the authority for saying that
the ideas back of all this were Allen's,
but in order to obtain a purified product of
the active hormone in the liquor folliculi
he was using, he enlisted the cooperation of
Edward A. Doisy whose laboratory was in
an adjoining building. Time has made it
clear that Doisy was undoubtedly the most
capable collaborator Allen could have found
anywhere for this kind of research.

A side of Allen that ingratiated him to all
was seen at the time of the first announce-
ment of the discovery of the action of
follicular fluid on the vagina and is quoted

EDGAR ALLEN

from Danforth's letter to the present bio-
grapher :

"On my way to the 1923 meeting of
Anatomists I went around by St. Louis
and had dinner that night at the Allen
home. In the evening Doisy, Allen, and
I had a long and animated discussion
over how their findings should be re-
ported. Allen, sanguine and ebullient,
was all for announcing them imme-
diately at the Anatomists' meeting,
even though there w^as no place for
such an announcement on the program.
Doisy, no less convinced of the im-
portance of the findings, so far as they
went, but temperamentally careful and
thorough, thought any announcement
should be withheld till more extensive
data could be presented. Allen did not
weaken, and when he was called upon
at the meetings to present his paper,
"Ovogenesis during sexual maturity,"
whic-h had been duly submitted and
published in the abstracts, he by-passed
that paper with only a few remarks
and used the available time to tell about
the effects of follicular hormone on
spayed mice. I think this oral report was
the first public announcement, the first
published one being the AUen-Doisy
paper in the Journal of the American
Medical Association, September, 1923."
Danforth's letter continues:

"When the report was presented, I
think it was received on the whole with
reservation if not outright skepticism,
particularly since no abstract had been
published and the topic was unan-
nounced and unexpected. George Corner
said later that was true in his case.
Someone else who was there told me of
his own skepticism and said that he
thought Allen, relatively unknown and
with an unconventional idea, was re-
garded as something of an 'upstart.'
Stockard is said to have made some very
caustic comments, which I don't recall.
Herbert M. Evans seems to have been
among the few who immediately sensed
the significance of the paper. He and 1
returned to California on the same train
and over and over he kept saying, '1

think Edgar (he sometimes called him

Ezra) Allen has something,' or words

to that effect."

Cooperation with Doisy continued in a
skillful chemical analysis which soon made
possible the isolation and chemical identifica-
tion of the estrogenic hormones. It is un-
likely that this stage of the investigation
would have been greatly prolonged had
Allen adhered to a 9 to 5 o'clock schedule
or a 40-hour wTek, but it is apparent from a
letter he wTote to Doctor Carl G. Hartman
that the discovery of the Allen-Doisy test
for estrogens might ha\'e been delayed had
he not made a midnight trip to the lab-
oratory. According to Hartman, "It was
Saturday night and Ed and his wife had
been at the theatre. Would he or would he
not make a midnight visit to the animal
colony at the University in St. Louis and
examine the castrated mice which had been
receiving Doisy's extracts? He did and
much to his delight found cornificd cells in
the vaginal smears, which might not have
been there if he had waited till Monday
morning I"

Beyond this point, most of what happened
has been recorded by the earlier biographers.
In 1923, and almost certainly before the
importance of his work was fully appreciated,
Allen was appointed Chairman of the De-
partment of Anatomy in the then 2-year
University of Missouri School of Medicine.
He was made Associate Dean of the School
of Medicine in 1929 and Dean the follow-
ing year. He was happy at Columbia,
"things have worked out so nicely here. . . ,"
he wrote Danforth, and later he began to
think he "was planted definitely in Mis-
souri." But April 1, 1933, he wrote, "Dean
Winternitz visited Columbia day before yes-
terday and asked me to go on to Yale ..."
as Professor of Anatomy and Chairman of
the Department of Anatomy in \hc ^'ale
University School of McHlicine.

The unusual extent to which his reseai'ch
and lh(> i-esearch he super\is('(l in his de-
partment were suggested by his thesis has
been indicated. We find, theictoic, investiga-
tions pertaining to the problem of oxogenesis.
At a time when it was generally assumed
that the female mammal possesses a full

EDGAR ALLEN

quota of ova when she is born, he demon-
strated that new ova arise after birth and
even after sexual maturity.

As Doctor WilHam U. Gardner, a former
student and his successor at Yale, has
written, his early conviction that the ovum
is "the dynamic center of the folHcle"
persisted throughout his life; he left two
partially completed manuscripts dealing in
part with the subject. Less than a year
before he died he wrote to Danforth, "I
still think of the ovum as a dynamic center
for mitosis and there is no reason why the
fertilized ovum shouldn't be." Gardner is
of the opinion that Allen's interest in ova
prompted the collaboration with Doctors
J. P. Pratt, Q. J. Newell and L. J. Bland
that resulted in securing in 1930 the first
human ova from the oviduct, and in pro-
viding evidence bearing on the time of
ovulation in the human female.

Overshadowing these studies of the ova
were, of course, Allen's many investigations
on the relationship of the ovarian estrogenic
hormones to the growth of tissues. His
demonstration that removal of the ovaries
of the rhesus monkey under appropriate
circumstances is followed by a uterine bleed-
ing, indistinguishable from that of normal
menstruation, led to the formulation of his
estrogen-deprivation theory of menstrua-
tion. Although challenged and shown to be
in need of modification, his statement of
this theory stimulated many studies of
that phenomenon which is still not under-
stood.

After moving to Yale, he became in-
creasingly involved in investigations of the
influence of steroid hormones on carcino-
genesis, especially the relationship of es-
trogens to malignant transformation of the
uterine cervix. His interest in the growth-
stimulating capacity of the ovarian hor-
mones was fiu'ther indicated by his use of
the mitosis-stimulating and mitosis-arrest-
ing drug, colchicine, in studies on the
genital tissues.

During the relatively short span of his
professional life, he and his collaborators
published more than 140 original investiga-
tions. Of these, most were under joint
authorship. This is explained by the fact

that he rarely worked alonc^ on a scientific
problem. This may have been just as well,
for his excitement and enthusiasm reached
their peak in team work. In all such re-
lationships, but especially when younger
colleagues were involved, he gave full sup-
port and generous credit. He would have
been proud of the large number of th(^ latter
who have since worked their way to im-
portant posts in anatomy. Were he alive,
he probably would still be prodding them
to look into this or that problem.

Although the number of articles and
reviews of which Allen was author or co-
author was relatively large, it was small
compared with the many which can be
attributed to the encouragement and en-
thusiasm he inspired among his students
and associates, and to the many more which
his work inspired in other laboratories
throughout the world. In this sense, rather
than in the strictest sense of the word, he
was a foremost anatomist.

At the height of his career, he undertook
the editing of the first edition of Sex and
Internal Secretions. The editorship of the
second edition was shared with Doctors
Danforth and Doisy. The former was un-
doubtedly over-modest in recalling the parts
of the co-editors in the undertaking, but his
statement is ciuoted for the information
it contains.

"I had helped a little with the first
edition, especially with Bridges' chapter
(Bridges being in Russia at the time),
and in the second edition took I'e-
sponsibility for Section A, as Doisy did
for his group of chapters. I read the
entire book in manuscript or in proof
(mostly both) as I think Doisy did also,
but neither of us, I feel sure, thought
of ourselves as co-editors. On Jan. 4,
1939 Allen wrote, T have asked the
publishers to make the book, Allen,
Danforth and Doisy; as it has been
team work all through.' My own reac-
tion, and probably Doisy's is expressed
in my reply of January 9, 'Instead of
writing a long personal letter in reply
to yours of January 4 I am going to send
a short note by air mail protesting
against your suggestion in the last

ED(;AR ALLEN

paragraph. It would be quite unjust
and misleading for you to put Doisy's
and my names in any sense coordinate
with yours. . . . And to put our names
in any Vnit the most subordinate posi-
tion would he to give credit that is not
due.' "

Allen's services were not unrecognized.
Three of the universities with which he was
associated, Vale, Brown, and Washington
University, awarded him honorary degrees.
Had he lived a few months longer, he
would have received an honorary doctor of
law from the University of IVIissouri. In
1987 he was awarded the Legion of Honor
in l-*aris where he was guest of the Fondation
Singer-Polignac at a colloquium on the
sexual hormones. In 1941 he was honored
by the Royal College of Physicians of
London when they conferred upon him the
Baly Medal for researches on the female
sex hormones. In the last year of his life
he was president of the two national societies
most closely representative of his field of
work, the Association for the Study of
Internal Secretions (now the Endocrine So-
ciety) and the American Association of
Anatomists.

As Corner, Danforth and Stone wrote
in their memorial for the American Associa-
tion of Anatomists, "Allen was a striking
figure personally, for his broad shoulders,
ruddy face and snow-white hair made him
conspicuous in any group. He made friends
(luickly and was always the first to en-
coiu'age and to admire good work by others.
His boyish frankness and his good will
contributed more perhaps than any other
factor to the effective cordiality which has
prevailed among the American workers in
the physiology of reproduction." The writer
of this skc^tch recalls that in any meeting
Allen was always one of the first to rise to
the floor with a (juestion or lo mak(> a
constructive' coniineul .

The Viest of the spirit of I he man that we
can preserve is containe'd in \\\o form of his
own words, written inforinall>- to his fri(Mid,
( harles Danfoith, with whom he coi're-
sponded so fre(iuentl>' and xoluminously.
A numbei- have been selected for the side
of him thev will recall, the enthusiasm and

imagination that enabled him to do the
amount and quality of w^ork he did.

November 6, 1935: "We just had a
keen thing happen here. Dr. George AI.
Smith returned from Europe with in-
formation about a drug which prevents
mitotic figures from completing divi-
sion. . . . Combining this with theelin
stimulation. . . ." Alay 23, year not
stated: "Have been working since early
morning on the slides from the monkey
experiments. They are a corking good
lot of evidence and I am in one of those
elated, trembly, inspired sort of heavens
to which new ideas always bring a
fellow. If you were in reach, you would
have been dragged over precipitately
to my scope, or interrupted to talk
things over a dozen times. ... As it is, I
must run my ideas through a slow movie
instead of talking you to death. . . .
Growth in the vaginal wall is almost
unbelievable. Growth in the cervical
glands is more so. They grow so darned
much that the cervix looks like a tumor
and the cervical canal becomes a tor-
tuous thing like the lower Mississippi."
A lot of histological detail, then: "the
tubes are a knockout . . . this would
seem to make ciliation a growth phase of
the nonciliated cells and full ciliation
dependent on presence of the hormone.
The mammary gland whole moimts are
wonders. I am so elated as to be almost
damned crazy. Am sure I'd crack one
of your ribs if I could get at you."
Then, "Yesterday a letter from the
National Research Council giving me
$800 more for monkeys. . . . Ain't it
just too lucky to l)eli(n-e. . . ." And
(inallw in closing one of his letters:
"Well you darned tool, Allen, it will
take him 10 miiuitcs to read your
letter now, aren't you e\('i' going to
stop writing?"

About 10 years belore Allen died he
sulTered a >r\r]v coi'onaiy attack, whil(>
waiting foi' a train in the ,Jackson\ille,
I'lorida, Lnion Station, lie lo\-ed swinuning
as well as sailing and, after a sti'enuous and
fatiguingxisit with Doctoi' IJobei't M. Verkes

EJ)C;AR ALLEN

in the heat at Orange Park, had taken time
for a long swim in the breakers at Jackson-
ville Beach. Miraculously he reached New
Haven and recovered sufficiently to return
to his work. Nothing would have kept him
from it. When war was declared in 1941
he was tired, but he insisted on joining the
Coast Guard Auxilliary for a weekly tour
of duty on Long Island Sound, as opera-

tions officer of a flotilla. It was during one
of their patrols that he died of coronary
occlusion. There was much that was fitting
to such an end of his active life. No tribute
could have been more appropriate than
that phrased by Gardner in th(^ Edgar
Allen Memorial Number of the ^'ale Journal
of Biology and Medicine, "The 'Skipper'
left with 'all sails filled.' "

PREFACE TO THE THIRD EDITION

The impact of the first two editions of
Sex and Internal Secretions can never be
measured, but it must be near the front for
books of its kind. Few books seem to have
served their purpose better and few, 20
to 25 years after their appearance, seem
to be valued as greatly by those who are
fortunate to possess copies. It was to be
expected, therefore, that pressure would be
brought to bear for the preparation of a
third edition. Whether the Publisher's at-
tempts to find an editor miscarried because
of the character of the new order ushered
in by World War II, or because discretion
was considered the better part of valor
may never be known. The odds favored the
latter explanation because there was no
direction in which a successor to Edgar
Allen, Edward A. Doisy and Charles H.
Danforth could go except down. Neverthe-
less, there were reasons for accepting the
challenge and attempting to do for the
present generation of reproductive phys-
iologists what Allen, Danforth and Doisy
and their many colleagues did for theirs.
Most of the problems to which they ad-
dressed themselves had not been solved,
although the need for answers was as urgent
as ever, and perhaps more so. The definition
of these problems had become obscured,
partly by the addition of many contradictory
and confusing data to the literature relating
to them, and partly by the rising tide of
interest in other glands of internal secretion,
notably the thyroid and adrenal. Finally,
new technologies had pervaded the field and
there were many new data and concepts to
be evaluated and Avoven into the fabric
Allen and his contemporaries had created.

In the preparation of the third edition
little of the first two editions was to be
retained except the title. Sex and Internal
Secretions, and the ideals by which the
authors of these editions must have been
guided. Within a framework of careful
scholarship, these were seen to be a resume
of the solid facts that had been learned from

test and retest, broadly critical discussions,
enumeration of the important unsolved prob-
lems, and the preparation of lists of ref-
erences complete enough for the guidance
of any seeker of information, whether his
interest was in the extension of basic studies
or the application to clinical and agricul-
tural problems. Adherence to these ideals
has not been easy. Even if ample allowance
is made for editorial ineptitude, the period
1958 to 1961 is different from 1932 and
1939. Reviews and symposia are more nu-
merous and many are in a style that is
alien to the traditions of Sex and Internal
Secretions. There are demands on the time
of many of us which, 20 years ago, were
reserved for only a few, and the volume of
published reports has long since outstripped
oiu- capacity properly to encompass them.

Despite the difficulties and misgivings,
an effort has been made to step into the
void created by the lapse of the old Sex
and Internal Secretions. Both similarities and
differences will be noted. Relatively more
space has been given over to the role of the
gonadal hormones in the control of repro-
ductive behavior, and relatively less to the
biochemical problems of hormone synthesis,
utifization, and metabolism. This "slighting"
of the biochemical side, if it is to be so
considered, does not reflect any lack of
appreciation of the key position occupied by
this discipline. It is explained, rather, by
the opinion of the biochemists who were
consulted that another review, just at this
time, would be anticlimactic to a number
of the excellent reviews which have recently
been published. The chapter by Dr. Villee,
therefore, is, in his words, a presentation
of the general picture without being an
exhaustive citation of the tremendous body
of relevant literature.

The suggestion made above that some of
the shortcomings in this third edition are a
reflection of changes in our habits of work-
ing is not to be taken as an attempt to

PREFACE, THIRD E])ITION

excuse any errors that properly belong on
the editor's doorstep. He has learned as he
has gone along, but finds himself reacting
very much as he usually does at the end of a
lecture — if it were to be given again, parts
of it would be done differently. To the
contributors, there is a feeling of the deepest
gratitude for the time and thought they
have given to the preparation of their
chapters. The editor is indebted, too, to the
National Institute of Mental Health, Na-
tional Institutes of Health, for a research
grant, M-4648, which has taken care of a
number of the costs of publication. Justifica-
tion for this action is believed to have been

given by the expansion of the section on the
gonadal regulation of reproductive behavior.
This development, in turn, would have
pleased Robert M. Yerkes. He was sensitive
to the need for truly scientific studies of
sexual behavior, the mechanisms which par-
ticipate in its expression, in the determina-
tion of its character, in its regulation, and
in its development. But he was equally aware
of the importance of investigations that
are purely physiologic, biochemical, and
morphologic, and in a quiet but effective
way did much to encourage many that are
recorded in the first two as well as in the
present edition of this book.

PREFACE TO FIRST EDITION

It is the purpose of this book to survey
the most important recent researches in
problems of sex, especially those concerned
with internal secretions, in order that con-
cepts already established by experimental
evidence may be clearly stated and made
readily available. While general principles
can in many cases be stated concisely, the
recent data have accumulated so rapidly
that there has not yet been time for retesting
and evaluating much of the evidence. This
may account in some chapters for emphasis
upon certain work with which contributors
may have had personal contacts. Further-
more, sex and reproduction show such wide
ranges of variation in both the structures
and functions involved that major dif-
ferences, even among species of higher mam-
mals, make generalizations both difficult
and dangerous. This whole field has recently
undergone such rapid growth that many
new ciuestions have arisen to challenge the
investigator's curiosity. An attempt will be
made to indicate productive approaches to
some of these unsolved or only partially
solved problems.

This book is intended for the reader with
a moderate biological background, to whom
the less involved technical terminology may
not prove a serious handicap. It is not our
intent that it should be a "popular book on
sex." Instead, it is designed for those in-
terested in the progress of research in
problems of sex, and those who may be
already engaged in investigations thereon
or casting about for promising problems for
investigation. Physicians who are interested
in fundamentals will find much valuable
recent material. In suppljnng a biological
foundation for education in matters of sex,
it should also attract the interest of serious
students of sex function in man.

Specialization in research has reached the
point where any detailed authoritative sur-
vey requires a group effort. Conseciuently,
the editoi attempted to gather together a
group of investigators whose work has es-
tablished them in their respective fields.

Each contributor has develop(>d his chapter
in his own way and assumes full respon-
sibility for the content of his section, in-
cluding his discussion of the work of othei'
investigators.

Since it was inevitable that considerable
correspondence would be involved, intro-
ducing the time-transport factor, choice
of the group was restricted to American
investigators. Several prominent workers
in this field, whom we would have desired
as co-authors, have necessarily been omitted
because of absence abroad or press of
other work.

As the Foreword indicates, this project
saw its inception in a proposal by Dr. E. V.
Cowdry, then Chairman of the Medical
Division of the National Research Council,
to the Committee for Research in the Prob-
lems of Sex. In the publication of many of
the contributors to this book, acknowledg-
ment will be found to this Committee for
support of investigations. In a sense the
book will serve as a summary of some of the
work accomplished under grants from this
Committee. Obviously, however, it was not
desirable to limit choice of contributors to
investigators who had received grants from
the Committee, and that consideration has
not determined their choice. Instead, it is
probable that the Committee in the first
place chose to make grants to these men
because of the promise of their work shown
by their previous investigations.

The editor wishes to acknowledge the
support of the Committee for Research in
Problems of Sex, which has made provision
for the editorial expenses involved in the
preparation of the manuscript. He also
wishes to commend the cooperation between
investigators who, even though they may be
competitors in the same field, have col-
laborated so well. The cooperation has been
completely free from the secretive reserve
sometimes encountered among investigators
who may be leaders in their particular
fields. The editor wishes further to acknowl-

xxiv

PREFACE, FIRST EDITION

edge the assistance and friendly interest of
Dr E. V. Cowdry, not only during the
initial phases of this project, but through-
out its progress and consummation. Dr. F. R.
Lillie has counselled wisely in regard to
titles and content of some of the sections.
Valuable counsel and encouragement has
also been received from several of the con-
tributors. The editor also wishes to commend

the contributors for their team-work ni
eliminating or reducing overlaps between
sections which is so necessary in the unifica-
tion of interlocking material from closely
related subjects.

Edgar Allen
University of Missouri School of Medicine
Columbia, Missouri
May, 1932.

SECTION A

Biologic Basis of Sex

GENETIC AND CYTOLOGIC FOUNDATIONS

FOR SEX

John W. Gowen, Ph.D.

Department of Genetics, Iowa State University, Ames, Iowa

I. Basic Literature

II. Mechanistic Interpretations op Sex

A. Concept of Sex Determination

B. Sex as Associated with Visible Chro-

mosomal Differences

C. Changing Methods of Cytogenetics

D. Chromosomal Association with Sex

E. Balance of Male- and Female-Deter-

mining Elements in Sex Deter-
mination

III. Sex Genes in Drosophila

A. Mutant Tjqjes

B. Major Sex Genes

C. Other Chromosome Group Associa-

tions: Drosophila americana

D. Location of Sex-determining Genes
IV. Sex under Special Conditions

A. Species Hybridity

B. Mosaics for Sex

C. Parthenogenesis in Drosophila

D. Sex Influence of the Y Chromosome

E. Maternal Influences on Sex Ratio

F. Male-influenced Type of Female Sex

Ratio "

G. High Male Sex Ratio of Cienetic

Origin

H. Female-Male Sex Ratio Interactions
V. Sex Determination in Other In-
sects

A. Sciara

B. Apis and Habrobracon

C. Bombyx

VI. Sex Determination in Dioecious

Plants

A. Melandrium

B. Rumex

C. Spinacia

D. Asparagus

E. Humulus

VII. Mating Types

VIII. Environmental Modifications of
Sex

A. Amphibia

B. Fish

IX. Sex and Parthenogenesis in Birds

X. Sex Determination in Mammals...
A. Goat Hermaphrodites

B. Sex in the Mouse 50

C. Sex and Sterility in the Cat 50

D. Deviate Sex Types in Cattle and

Swine 51

E. Sex in Man: Chromosomal Basis 52

1. Nuclear chromatin, sex chro-

matin 55

2. Chromosome complement and

phenotype in man 5(1

3. Testicular feminization 50

4. Superfemale 57

5. Klinefelter syndrome 58

0. Turner syndrome 59

7. Hermaphrodites 59

8. XXXY + 44 autosome type 01

9. XXV + 66 autosome type 03

10. Summary of types 03

11. Types unrelated to sex 03

F. Sex Ratio in Man 65

XL References 00

I. Basic Literature

In the first edition of Sex and Internal
Secretions published in 1932, Bridges dis-
cussed the closely prescribed problem of the
genetics of sex, particularly as it was related
to one species, Drosophila melanogaster.
The treatment was sharply focused on the
advances made, chiefly through his own re-
searches, in understanding the functions of
the genetic and cytologic factors operating
during embryologic development which ul-
timately establish the sex types. The em-
phasis was on gene action through inter-
chromosomal balances as they may aff"cct
sex expression. The second edition of 1939
brought this material up-to-date and, at
the same time, offered a much broader treat-
ment by the inclusion of accumulated evi-
dence on how differentiation for sex comes
about in other forms. There is no substitute
for a careful reading of these two presenta-

BIOLOGIC BASIS OF SEX

tions as background material for a present
day understanding of sex determination.
The current presentation makes no attempt,
except in the barest outline, to repeat this
early material other than in those aspects
which bear on the advances that have been
made since the printing of the second edi-
tion.

Immediately following the publication of
the first edition of Sex and Internal Secre-
tions there was a resurgence of interest in
the problems considered in that book. The
resulting research led to notable advances
in available knowledge. Seven years later
the second edition was published. A second
wave of accomplished research appeared. In
the interim of the past 21 years extensive
advances have been recorded, particularly
in understanding the mechanisms of sex dif-
ferentiation in plants as well as in many
animals. Among others, the data on Melan-
drium, Asparagus, Rumex, and Spinacia
have been of first importance. Further in-
formation on Drosophila, Habrobracon, and
Xiphophorus has notably broadened our
viewpoints. Beginnings have been made to
a better understanding of the conditions for
sex separation in bacteria, protozoa, bees,
birds, goats, mice, and man. To these spe-
cific contributions may be added the basic
advances in understanding the methods by
which genes are transmitted from one gen-
eration to the next, accomplish their actions
in development, and by which chromosomes
reorganize and reconstruct gene groups.

The period has also been one of excellent
monographic treatments on different phases
of the subject. Wilson's The Cell in Devel-
opment and Inheritance (1928) and earlier,
Morgan's Heredity in Sex (1914), Schra-
der's The Sex Chromosomes (1928) and
Goldschmidt's Lymantria (1934) retain
their pre-eminence. To this list have now
been added the publication of extensive
tabulations of chromosomes of cultivated
plants by Darlington and Janaki Ammal
(1945), of animal chromosomes by Makino
(1951 ), and the yearly index to plant chro-
mosomes beginning with 1956, compiled by
a world-wide editorial group and published
by the University of North Carolina Press,
Chapel Hill, North Carolina. Those volumes
give access to the basic chromosomal consti-

tutions and their sex relations for far more
species than were heretofore available. In-
heritance information has been made more
accessible and at the same time in more de-
tail. The tabulations of mutants observed in
particular species as in D. melanogaster
(Morgan, Bridges, and Sturtevant, 1925;
Bridges and Brehme, 1944) have been fol-
lowered by those of corn, mouse, domestic
fowl, rat, rabbit, guinea pig, Habrobracon,
and many other animal forms, together
with similar tabulations for cereals and
a number of other plant species. Books
of particular interest include those of Hart-
mann. Die Sexualitdt (1956), White, Ani-
mal Cytology and Evolution (1954), Gold-
schmidt, Theoretical Genetics (1955,1,
Hartmann and Bauer, Allgemeine Biologic
{ 1953) , and Tanaka, Genetics of Silkworms
(1952), and special reviews in various vol-
umes of Fortschritte der Zoologie (1 to 12)
(Wiese, 1960). Basic normal development
of Drosophila has been presented in con-
venient book form in papers by Cooper,
Sonnenblick, Poulson, Bodenstein, Ferris,
Miller and Spencer, Biology of Drosophila
(Demerec Ed., 1950) . Special papers hav-
ing a direct bearing on the subject matter
include particularly those of Pipkin (1940-
1960) in analyzing the various chromo-
somes of Drosophila for sex loci, of Tanaka
(1953) and Yokoyama (1959) for their
studies and reviews of the genetics of silk-
worms, and of Westergaard (1958) on sex
determination in dioecious flowering plants.

II. Mechanistic Interpretations of Sex

The records of search for basic mecha-
nisms involved in the determination of sex
have been foremost in the writings of man
from the beginning of historical record.
These ideas have included all forms of
mechanisms both intrinsic and extrinsic to
the organism. The most constructive ad-
vance, although not realized at the time,
came through the recognition of cell struc-
ture, chromosome maturation and the dis-
crete behavior of the inheritance. Differ-
ences in clu-omosome behavior were noted
but not alwaj^s related to facts of broader
significance. The period was one of expand-
ing observations and development of ideas
as they applied to species in general and

FOUNDATIONS FOR SEX

also to the further complications which
arose with more extended study. These
complications included differences not only
in one chromosome pair but in several. The
significance of a balance between these
chromosome types as well as with the en-
vironment was grasped by Goldschmidt
and particularly by Bridges where his more
favorable material brought out sharper
contrasts in types and in the chromosome
behavior. Ideas related to chromosome bal-
ance as they may affect developmental
processes were developed. Goldschmidt em-
phasized that this balance could include
factors in the cytoplasm as well as in the
chromatin material. Bridges' observations
on the other hand, pointed most strongly to
the chromatin elements where changes in
chromosome numbers were often accom-
panied by sharp differentiation of new
sexual types.

Concepts of sex determination broadened.
They came to include all the chromosomes
in the haploid set, a genome, as a whole.
The important concept of single chromo-
some difference being all important has
been replaced with that of a balance be-
tween the chromosomes. It has sometimes
been emphasized that it is this balance
which is the most significant element in sex
determination. Present day research seems
to be pointing rather to even finer struc-
tures than the chromosomes in that the
main search is on for the particular genes
within the inheritance complex which con-
tribute the characteristics of sex. Bridges'
work has sometimes been asserted as com-
mitted to a particular type of balance in
the chromosome. However, his writings
1 1932) make it clear that this balance is, in
his judgment, basically due to the genes
actually contained within the chromosomes
rather than to the chromosomes themselves.
This view is further emphasized in the
second edition of Sex and Internal Secre-
tions. "Sex determination in numerous
forms with visible distinctions between the
chromosome groups of the two sexes was
at first interpreted on a 'quantitative' basis,
as due to graded amounts of 'sex-chroma-
tin.' But because even more species were
encountered in which no visible chromo-
some difference was detected, the formula-

tion was changed to include also 'qualita-
tive differences in sex-chromatin.' Now, sex
is being reinterpreted in terms of genes,
with investigation by breeding tests pene-
trating to detail far beyond the reach of
cytological investigation. The chromosomal
differences are now treated as rough guides,
and chromosomal determination is l)cing

resolved into genie determination

Hence the difference in sex must be put on
the same basis as that of any other organ
differentiation in plant or animal, for ex-
ample, the difference between legs and
wings in birds— both modifications of an-
cestral limbs."

In contrast to what some have inter-
preted. Bridges was not committed to a
specific chromosome balance, as for in-
stance the X chromosomes to the A chro-
mosomes in Drosophila, for all species.
Rather he looked on that particular balance
as but one mode of gene association in
chromosomes that could bring about dif-
ferences in gene action on development
which would lead to sexual differentiation
in its various forms. In that sense his ideas
prepared him for, and were in conformity
with, the types of sex differentiation which
later became established in such forms as
Melandrium and in the silkworm. In his
concepts of sex determination and in his
active interest in making genetics more spe-
cific, there can be little doubt that his
theory would include those of the present
and would also be welcome as refining and
more sharply defining the significance of
cytologic and genetic elements important to
sex differentiation.

A. CONCEPT OF SEX DETERMINATION

As frequently happens, observations are
made often before their significance is more
than dimly understood. Background infor-
mation may be insufficient for the facts to
become clear. This was true when a lone
chromosome was discovered as a part of
the maturation complex in sperm formation
of Pyrrhocoris. The association of the pres-
ence of this element with the idea of its be-
ing the arbitrator between male and female
development in certain species was not
made until 10 years later. Henking (1891)
observed in Pyrrhocoris 11 paired chromo-

6

BIOLOGIC BASIS OF SEX

some elements and one unpaired element
which after spermatocyte divisions led to
sperm of two numerically different classes,
one carrying 11 chromosomes plus the ac-
cessory chromosome and the other carrying
only the 11 chromosomes. His observations
were confirmed by several other investiga-
tors and in principle for other insect species
over a decade following. The first sugges-
tion that this chromosome behavior was re-
lated to sex diiTerentiation came from
McClung's (1902) observations on the "ac-
cessory chromosome" of Xiphidiiim fascia-
tum. It is of some interest to examine the
steps l)y which these conclusions were
reached.

"The function exercised by the accessory
chromosome is that it is the bearer of those
qualities which pertain to the male or-
ganisms, primary among which is the fac-
ulty of producing sex cells that have the
form of spermatozoa. I have been led to
this belief by the favorable response which
the element makes to the theoretical require-
ments conceivably inherent in any struc-
ture which might function as a sex determi-
nant.

"These requirements, I should consider,
are that: (a) The element should be chro-
mosomic in character and subject to the
laws governing the action of such struc-
tures, (b) Since it is to determine whether
the germ cells are to grow into the passive,
yolk-laden ova or into the minute motile
spermatozoa, it should be present in all the
forming cells until they are definitely es-
tablished in the cycle of their development,
(c) As the sexes exist normally in about
(■(lual projoortions, it should be present in
half the mature germ cells of the sex that
bears it. (d» Such disposition of the element
in the two forms of germ cells, paternal and
maternal, should be made as to admit of
the readiest response to the demands of en-
A'ironment i-egarding tlie ])i'oportion of the
sexes, (el It should show variations in
structure in accordance with the vai-iations
of sex potentiality observable in different
species, (f) In parthenogenesis its function
would be assumed by the elements of a cer-
tain polar body. It is conceivable, in this
regard, that another form of polar body
miglit function as the non-determinant

bearing germ cell." The important fact es-
tablished by this reasoning was that a chro-
mosome could be the visual differentiator
between the fertilized eggs developing spe-
cific adult sexual differences. It w^as of little
consequence perhaps that for sex itself, in
this species, the chromosome arrangement
was misinterpreted. The validity of the
rules was probed through examinations of
the cells of many species by many different
observers.

B. SEX AS ASSOCIATED WITH VISIBLE
CHROMOSOMAL DIFFERENCES

Variations in the chromosome complexes
contributing to sexual differentiation were
soon found. Probably the most frequent
type observed in animals and plants was
that in which a single accessory chromo-
some of the male was accompanied by, and
paired with, another chromosome either of
the same or of a different morphologic type.
This other chromosome, as distinguished
from the accessory chromosome now gen-
erally called the X, was designated the Y
chromosome. For different species the Y
ranged in size from complete absence, to
much smaller, to equal to, to larger than
the X. Besides the X and Y chromosomes,
the autosomal or A chromosomes, the ho-
momorpluc pairs, complete the species
chromosome complement. In this terminol-
ogy

Sperm (Y + A) + egg (X + A)
= XY + 2A cells

of

lie determinino; tvpe

Sperm (X + A) + egg (X + A )
= 2X + 2A cells

of female detei'mining tyi)e

Variations in numbers and sizes of chro-
mosomal pairs making up the autosomal
sets of different species are familiar cyto-
genetic facts. These variations may extend
from one pair to many chromosome pairs
depending on the species. Similar variations
may occur in chromosomes of eitluM- the X
or Y types. The X is commonly a single
chromosome but may be a compound of as
many as 8 chromosomes in Aiicaris inciirva
(doodricli. 1916). The Y mav be lacking

FOUNDATIONS FOR SEX

entirely in some species or may be redupli-
cated in others. Males which form two
types of sperm are often called heterozygous
for sex, although the term digametic may
be better, whereas the females are homoga-
metic.

In other species the females show the
chromosomal differences whereas the males
are uniformly homogametic. These types
are principally known in the Lepidoptera,
more primitive Trichoptera, Amphibia,
some species of Pisces, and Aves. The single
accessory chromosome is present in the fe-
males whereas the males have two. It may
l)e unpaired or be with another chromosome
of different morphology and size. The sex
types may be symbolized by designating
the accessory chromosome by Z and its
mate W as a means of separating possible
differences between them and the X and Y
chromosomes. The significance of these dif-
ferences is not clearly established with the
consequence that some investigators prefer
to substitute the XY designation for the fe-
males and the XX condition for the males
of these species. The zygotic formulae are:

Sperm ZA + egg ZA = 2Z + 2A male
Sperm ZA + egg WA = ZW + 2A female

A third major chromosomal arrangement
accompanying sex differentiation came to
light as a result of Dzierzon's 1845 discov-
ery of parthenogenesis in bees. Chromoso-
mal and genetic studies have shown both
Apis and Habrobracon of the Hymenoptera
to be haploid, N, for each germinal cell in
the male complex and diploid, 2N, in the fe-
male. N may stand for any number of chro-
mosomes, as 16 for Apis or 10 for Habro-
bracon. The same chromosomal pattern
characterizes, so far as known, the other
genera of Hymenoptera. Haploidy vs. dip-
loidy is viewed as the most obvious feature
predicting differentiation toward the given
sex type even though, as in Habrobracon,
there is evidence for a particular chromo-
some of the N set carrying a locus for sex
differentiating genes. The zygotic sex for-
mulations are:

No sperm + egg N = N male
Sperm N + egg N = 2X female

In development, as in some other species

oi widely diverse origins, chromosome pol-
yploidy may take place causing the soma
cells to differ from the germ cells in their
chromosome coniponents.

The common phenotype for plants and
lower animals has differentiated sex organs
which are combined in the same individual.
Plant species seldom depend on any but
hermaphroditic types for their reproduc-
tion. Lewis' (1942) tabulation for British
flora had but 8 per cent of all species de-
pend on other than this form of reproduc-
tion. Those that had perfected dioecious
systems were not all alike in the system
adopted, although species with XY + 2A
males and XX + 2A females were in high
frequency. Dioecious reproduction was
rarely the system common to a whole
genus. Recent and multiple origins of dioe-
cious types are indicated by the irregular
distribution of species with bisexual repro-
duction within the different genera and
families. Methods for preventing inbreed-
ing have taken other channels as self-
sterility genes reminiscent of fertility or
sex alleles found in the older bacteria and
protozoa. Animals of the lower phyla are
those which are most frequently hermaph-
roditic. Within hermaphroditic species
chromosomal distinctions are ordinarily ab-
sent. The higher forms with sex and chro-
mosome differences, on the other hand, may
show reversion to the hermaphroditic con-
dition from the dioecious or bisexual states.

Other less common cytogenetic controls
of sex development have become recognized
and better understood. Discussion of their
gene and chromosomal arrangements will
be considered when these cases arise. The
al)ove types will be sufficient to furnish a
basis for interpreting the newer data.

C. CHANGING METHODS OF CYTOGENETICS

The earlier studies of sex determination
depended on the natural arrangements of
chromosomes found in different species and
on occasional chromosomal rearrangements
occurring as relatively rare aberrant types.
Further development of genetics has in-
creased the tools for these studies. Mutant
genes have been shown to control chromo-
some pairing in segregation (Gowen and
Gowen, 1922; Gowen, 1928; Beadle, 1930).

BIOLOGIC BASIS OF SEX

Different drugs such as chloralhydrate and
particularly colchicine and various forms
of radiant ^energy have made it possible to
create new types by doubling the chromo-
somes, by chromosome rearrangement and
by changing the chromosome number. Bet-
ter genetic tester stocks of known composi-
tion have been organized, which together
with a more exact understanding of the in-
heritance structure of the species have
made for more critical studies. Techniques,
which have improved chromosome differen-
tiation and structural analysis through bet-
ter methods, better dyes, tracers to mark
chromosome behavior (Taylor, 1957), and
the use of hypotonic solutions in the study
of cells (Hsu, 1952), have removed doubts
that were created by the earlier technical
difficulties. Instrumentation has improved
for the measurement, physically and chemi-
cally, of cell components.

D. CHROMOSOMAL ASSOCIATION WITH SEX

The first genetic linkage groups were as-
sociated with the sex chromosomes and
were soon shown to follow the patterns of
the different chromosome sets observed
within the different species. In man, hemo-
philia inheritance was observed to follow
that which was presumed for the X chro-
mosome. Barring in birds and wing-color
pattern in Lepidoptera on the other hand
were found to follow the chromosomal pat-
terns of their species where the male was
ZZ and the female Z or ZW for the sex chro-
mosomes. The utility of these methods was
further probed by Bridges' observations
that when chromosome behavior in Dro-
sophila resulted in sperm or eggs carrying
uncxjx'cted chroiiiosomc combinations there
followed ('(nmlly unexpected phcnotypes in
the progeny. The cliaractei'istics of these
unexpected progeny, in turn, I'ollowt'd those
expected if genes for them were carried m
the sex chromosomes. The use of these link-
age groups as tracers, both as they natu-
rally occur and as they may be reorganized
through the treatment effects of such agents
as radiant energy, open the possibility ol
assigning sex effects to, not only chromo-
somes, but also to particular i-laces within
the chromosomes.

Both polyploids and aneuploids, as oc-

curring naturally and as marked by tracer
genes, give insight into sex differentiation
and its dependence on chromosome inherit-
ance behavior. True polyploids in a species
are formed as a consequence of multiplying
the entire genome. The possible types may
have a single set of chromosomes and genes
in their nuclei (haploid ) , 2 sets (diploid) , 3
(triploid), 4 (tetraploid), and so on, rep-
resenting the genomes 1. 2. 3. 4. to whatever
level is compatible with life. Multiplying
the genomes within the nucleus often in-
creases cell size but seldom gives the or-
ganism overtly different sex characteris-
tics.i Aneuploids, on the other hand, give
quite different results, the result being de-
pendent upon the particular chromosome
that may be multiplied. Particular tri-
somies in maize, wheat, and spinach, for
example, are distinguishable by marked
differences in phcnotypic appearance that
is not attributable to cell size but rather
to abrupt deviations in particular charac-
teristics. The genes in the particular tri-
somic set are unbalanced against those of
the rest of the diploid sets within the or-
ganism. Their phenotypes express these
differences.

E. BALANCE OF MALE AND FEMALE

DETERMINING ELEMENTS IN

SEX DETERMINATION

The foundations for the basic theory that
sex determination rests on quantitative re-
lationships of two genes or sets of genes
localized in separate chromosomes rests
largely on tlie work of Morgan, Bridges and
Sturtevant, 1910, with Drosophila and the
breeding work of Goldschmidt, 1911, with
Lymantria. The work of Goldschmidt soon
gave extensive descriptions of diploid inter-
sexuality in L]imantria dispar. The details
of this work scarcely net'd review because
they have \)vvn repealed several times and
have been smnmarized recently in Gold-
schnu(h's Theoretical Genetics (1955).
Fioui (hildschmidt's viewpoint, ''the basic
point was that definite conditions between
the sexes, that is, interscxuality, could be
pro.lnced at will l)y proper genetic combina-
, tions (crosses of subspecies of L. dispar)

• The notable exception of the Hynienoptera will
I )c discussed later.

FOUNDATIONS FOR SEX

without any change in the mechanism of
the sex chromosomes, and this in a typical
quantitative series from the female through
all intergrades to the male, and from the
male through all intergrades to the female,
with sex reversal in both directions at the
end point. The consequence was: (1) the
old assumption that each sex contains the
potentiality of the other sex was proved to
be the result of the presence of both kinds
of genetic sex determiners in either sex; (2j
the existence of a quantitative relation,
later termed 'balance' (though it is actually
an imbalance), between the two types of
sex determiners decides sexuality, that is
femaleness, maleness, or any grade of inter-
sexuality; (3) one of the two types of sex
determiners (male ones in female hetero-
gamety, female ones in male heterogam-
ety) is located within the X-chromo-
somes, the other one, outside of them; (4)
as a consequence of this, the same deter-
miners of one sex are faced by either one
or two portions of those of the other sex
in the X-chromosomes; (5) the balance
system works so that two doses in the
X-chromosomes are epistatic to the deter-
miners outside the X, but one dose is hypo-
static; (6j intermediate dosage (or po-
tency) conditions in favor of one or the
other of the two sets of determiners result,
according to their amount, in females,
males, intersexes, or sex-reversal individuals
in either direction; (7) the action of these
determiners in the two sexes can be under-
stood in terms of the kinetics of the reac-
tions controlled by the sex determiners,
namely, by the attainment of a threshold
of final determination by one or the other
chain of reaction in early development;
while in intersexuality the primary deter-
mination, owing to the 1X-2X mechanism,
is overtaken sooner or later — meaning in
higher or lower intersexuality — by the op-
posite one, so that sexual determination
finishes with the other sex after this turning
point. The last point is, of course, a problem
of genie action."

Bridges developed his idea of "genie bal-
ance" as a consequence of his observations
on chromosomal nondisjunction, particu-
larly as it illustrated the loss or gain of a
fourth chromosome in modifving nonsexual

characters. The similarities and contrasts
of this view from that of Goldschmidt are
indicated by the following quotation
(Bridges, 1932) : "From the cytological re-
lations seen in the normal sexes, in the in-
tersexes, and in the supersexes, it is plain
that these forms are based upon a quanti-
tative relation between qualitatively differ-
ent agents — the chromosomes. However,
the chromosomes presumably act only by
virtue of the fact that each is a definite
collection of genes which are themselves
specifically and qualitatively different from
one another. There are two slightly differ-
ent ways of formulating this relation, one
of which, followed by Goldschmidt, places
primary emphasis upon the quantitative
aspect of individual genes. The other view,
followed by the Drosophila workers, em-
phasizes the cooperation of all genes which
are themselves qualitatively different from
one another and which act together in a
quantitative relation or ratio. Goldschmidt
developed his idea through work with the
sex relations in Lymantria and has sought
to extend it to ordinary characters. The
other formulation, known as 'genie balance,'
was developed from the ordinary genetic
relations found in characters. Both are
crystallizations of fundamental ideas with
which the earlier literature was fairly
saturated and no great claim to distinctive
originality should be ventured for either or
denied for one only. Both are physiological
as well as genetic — that is, they are formu-
lations of the action of genes, not merely
statements of the genie constitutions of in-
dividuals nor merely studies of the way
genes act. The physiological side has been
emphasized by Goldschmidt and the ge-
netic side by Drosophila workers. But
Goldschmidt's tendency to represent the
view of genie balance as without, or even
as in opposition to, such physiological for-
mulation is groundless — as groundless as
would be the reciprocal contention that
Goldschmidt's theory is only one of 'pheno-
genetics.'

"A common element in the foundation of
both formulations is that if a gene is repre-
sented more than once in a genotype the
phenotypic effect is expected to be different,
though roughly in the same direction as be-

10

BIOLOGIC BASIS OF SEX

fore and roughly proportional to the quan-
titative change in the genie constitution."

Since the sexually different types ob-
served by Bridges were accompanied by
whole chromosomal differences, he could
point to losses or gains of autosomes, with
an internal preponderance of genes tending
to develop male organs, as balanced by
genes of the X chromosomes tending in the
direction of producing female organs or of
suppressing alternative male organs. The
net effect of the X chromosome favoring fe-
maleness, and of the set of autosomes fa-
voring maleness, terminates in the devel-
opment of male or female, according to the
ratio of these determiners in the whole
genotype. The effect of the X chromosome
goes on the basis of whole numbers 1, 2, 3,
4, as does the similar variation in the sets
of autosomes.

Goldschmidt, since he was dealing with

TABLE LI

Chromosomal numbers and kinds for the different

recognized sex types of Drosophila

Type

Superfemale . . .
Triploid meta-

female

Female*

Female

Female

Female

Female

Female

Female

P'emale

Female

Female

Intersex*

Intersex

Intersex

Male

Male

Male

Male

Male*

Supermale

Chromosomes

X

Y

A

3

2

4

3

4

4

3

3

3

1

3

3

2

3

2

2

2

1

2

2

2 + vS

2

2

2 + yL

2

2

2

2

1

1

3

4

2

3

2

1

3

2

1

2

2

2

3

2

4

3

X/A Balance

1.5

L3

LO

LO

LO

LO

LO

1.0

LO

LO

LO

LO
.75
. ()7
.()7
.50
.50
.50
.50
.50
.33

* These forms are cited by Bridges from his
own observations, from L. V. Morgan (1925) and
from Sturtevant (unpublished). As yet but limited
studies of these forms, which must be rare, have
been published. Fourth chromosomes generally,
but not always, equal mimber of the other indi-
vidual chromosome gn)U])s.

males and females of the diploid type, took
a corresponding view for the Z chromo-
somes of his moths with this difference.
Since the Drosophila chromosome pattern
is XY + 2A for the males as contrasted to
2X + 2A for the females and the pattern
for Lymantria is ZZ -|- 2A for the males
and ZW + 2A for the females, it was neces-
sary to use a relation which was reciprocal
to that of Drosophila; the male determining
element or elements were assigned to the Z
chromosomes. Lymantria has a rather large
number of different races found in different
geographic locations. Within any one of
these races this formulation apparently suf-
ficed. However, from crosses between races
it was soon observed that the progeny
showed ranges in sexuality all the way from
phenotypic males to phenotypic females al-
though these females were actually genetic
males. To Goldschmidt, this variation indi-
cated different potencies of the male-deter-
mining element. Similar differences were at-
tributed to the female element which he
had first assigned to the cytoplasm but for
which he later favored a W chromosome
location.

In applying this postulate of discrete
chromosome contributions to sex according
to their number, Bridges made the further
assumption that female-producing genes
l)redominate in the X and are scattered
through it in more or less random fashion
as are the genes affecting so-called somatic
characteristics as wing shape or bristle pat-
tern. The quantitative relations for the dif-
ferent chromosomal types, together with
their descriptions, are indicated in Table
1.1.

In the formulation of Table 1.1, the X
and Y chromosomes are counted separately,
whereas a set of A chromosomes (auto-
somes), is allowed a value of but one even
though comjiosed of a 2nd, a 3rd, and a
4th chromosome for the haploid genome.
The dii)loid set of autosomes is given a
weight of two, and so on. From these data
Bridges observed that the presence or ab-
sence of a Y chromosome did not affect the
sex types. The X chromosomes and auto-
somes were, however, important. He held
that their importance stood as the ratio of
thcii' prcsuiiK'd iM'oducts to each other. The

FOUNDATIONS FOR SKX

11

ratio is 1 for the perfect female and 0.5 for
the perfect male. Between these values in-
tersexual conditions develop. Beyond the
value 1, development is overbalanced by-
excess female genes resulting in the super-
female. Values less than 0.5 create a de-
ficiency in the female elements or excess of
male elements and a supermale results.
Schrader and Sturtevant (1923) proposed
another system. Instead of the ratio of X
clu-omosomes to autosomes, they suggested
that a straight difference between the prod-
ucts of the female determining elements of
the sex chromosomes and the male effects of
the autosomes causes the sex changes.
]3ridges criticized this system on the basis
of the fact that progressive polyploidy did
not change the sex type or ratio between the
X and autosomes, whereas the numerical
difference between them would be progres-
sively increased. While keeping the X/A
ratio as descriptive of the ultimate effects
of the genes in these chromosomes, he modi-
fied their proposed weight from 1 : 1 for
X:A to 1 for the X and 0.80 for the A.

All formulations for explaining sexual
differences are beset with a lack of an un-
biased quantitative scale by which these
differences can be measured. The estimates
of the changes in sexuality are left to the
insight of the observer. It seems reasonable
to suppose that the quantitative relations
between the male and female sex determin-
ing elements should have intermediate
values when the specimens under observa-
tion show a mixture of organs of either sex.
This agrees with Bridges' considerations of
this problem. It is not so clear, however,
that the so-called supersexes^ really are
what the names may connote to many
readers. The superfemales, with their three
X chromosomes and two sets of autosomes,
are quite inviable; small in size, wings re-
duced and irregularly cut on the margins,
ovaries developed to only the early pupal
stage, and reproductive tracts much re-
duced in size.^ The supermales with one X

- Recently termed metafemales bj' Stern (1959b).

'^ Further development of the ovaries is able to
take place in normal XX + 2A hosts. Larval
XXX + 2A ovaries transplanted into fes/fes hosts
IM-oduce eggs in the recipient host which on fer-
tilization are capable of developing into adult ima-
goes. These imagoes show that there is a low per-

chromosomc and three sets of autosomes
are described as resembling males but are
sterile. The wings arc somewhat spread and
bristles less in size. They are late emerging
and poorly viable. Neither type can be re-
ferred to as superior to normal female or
normal male in anatomic develoi)ment or
physiologic functioning. A new type, re-
cently described by Frost (1960j, empha-
sizes this difficulty. Females, called triploid
metafemales, Table 1.1 had 4X chromosomes
and 3 second, 3 third, and 2 fourth auto-
somes. Viability was greater than superfe-
males but still low. Fertility was about 10
per cent. Progeny per female about 10. The
flies were like triploids in bristles, eye and
wing cells large, sex combs absent. They
showed characteristics of superfemales in
rough eyes, narrow wings without inner
margins, and smaller body build. The dis-
tribution of these different Drosophila sex
types (Table 1.1) showed that optimal
development comes when the X/A values
are 1.0 and 0.5. Any deviation away from
these values tends to make the sex system
less rather than more efficient.

In normal Drosophila sex differences are
probably expressed in every cell making up
their bodies. These differences are made
visible to us only under special conditions.
In adult organ differentiation, the sexual
differences are manifest through such
things as the body size of the males being
about three-fourths that of the females,
differences in coloration of the tergites, the
appearance of sex combs, the development
of the gonads into ovaries and testes, and
the formation of a secondary reproductive
system composed of several glands and
ducts. The origin of these last elements is
of particular interest. The ovaries can be
distinguished from the testes as early as
the second instar through their size and
position within the fat body. They are lo-
cated about two-thirds the larval length
back from the mouth parts. The sex combs,
on the other hand, take their origin from
imaginal discs which are located in the
head region, possibly one-third back from
the mouth parts. The secondary reproduc-

centage of crossing over and high nondisjunction
rate in the XXX + 2A ovaries and a high mortality
rate in the offspring (Beadle and Ephrussi, 1937).

12

BIOLOGIC BASIS OF SEX

tive system for both the males and females
takes its origin from a disc at the posterior
extremity of the larvae. The testes or
ovaries are only brought together with the
secondary reproductive tracts during late
development in the pupae, whereas the sex
combs retain their separate development.
The striking action of chromosome balance
in sex determination is that all of these
organs and others are jointly affected and
simultaneously develop directly into either
the male or female sexual types. Of these
characteristics, the sex combs are particu-
larly trustworthy indices of maleness. In
the male, these combs are heavily sclero-
tized and pigmented. They have 9 to 13
rather blunt teeth on their margins. The
normal females lack these sex combs en-
tirely. The intersexes have well developed
combs, whereas the triploid females and
superfemales lack them. This all-or-none
situation resembles that observed in the
rest of the block of sexually differentiating
characteristics found in normal males or
females, save that in the intersexes develop-
ment may result in the organs of either sex
appearing in the primary or secondary re-
productive systems. The all-or-none char-
acter of the sex combs may, however, be
bridged in that the appearance of certain
mutations leads to the production of these
combs in all of the different sexual types.
The combs differ in size, in thickness and
length of the teeth, and their number. The
sex combs meet the conditions of a quanti-
tative character which may be counted or
measured in unbiased units. Thus in the
case of sex combs, it is possible to obtain
quantitative information on the effects of
clu'omosomal changes on the expression of
this form of sexuality.

TABLE L2

Mean sex conih teeth for flies having (lijj'erent X and

A chromosome

numbers and heteroziiqons for

the Hr gene

Sex Type

Chromosome

Mean Sex

Genotype

Comb Teeth

Males

X + 2A

11.4

Intersex

XX + 3A

9.1

Female

XX + 2A

().9

Superfemale ....

XXX + 2A

5.0

Triploid female.

XXX + 3A

4.8

Two gene mutations in D. melanogaster
have aided in this search. The first is the
dominant gene, Hr, which causes the male
reproductive tract to be added to that of
the female when this gene is heterozygous
(Gowen, 1942 j . The extensive effects of this
gene on the whole sex determining system
have been described by Gowen (1942, 1947)
and Fung and Gowen (1957a). The second
gene is that of Sturtevant's (1945) trans-
former, tra, which operates on the female
phenotype to convert it to that which cor-
responds to the male in having sex combs,
full male reproductive system including
testes, while still leaving the female charac-
teristic body size. These genes are located
in the third chromosome of D. melanogaster.
Crosses between them show allelomorphic
effects. Combinations of these genes in con-
junction with different chromosomal ar-
rangements make possible a series of differ-
ent sex types which are distinguishable
from ordinary males and females (Gowen
and Fung, 1957). As some of these types
bear sex combs, a quantitative character is
furnished in the variation of the sex comb
teeth which may be used as an impersonal
measure of departure of these types from
ordinary males or females. The groups hav-
ing particular interest are those carrying
the Hr gene in heterozygous condition with
the X and A chromosomes having various
numbers. The mean numbers of sex comb
teeth for these various groups are shown in
Table 1.2.

Analysis of these data for the contribu-
tions made by the sex chromosomes and
autosomes to the numbers of teeth found on
the sex combs of these different genotypes
shows the following relation.

Sex combs == 12.82 - 3.42 X + 0.89 A

From this equation it is seen that the X
chromosome has four times as much effect
on lowering the number of teeth on the sex
combs as a set of autosomes. The direction
of effect is, as would be expected, increasing
the number of X chromosomes tends toward
making the individual more female-like in
that the sex combs become smaller and less
l)ronounced. Increasing the number of au-
tosome sets on the other hand tends to push
the indiviihial toward the male type with

FOUNDATIONS FOR SEX

13

larger sex combs. Some 95 per cent of the
variation in sex comb teeth has been ac-
counted for by this equation.

The above equation results when the ef-
fect of the sex chromosomes and autosomes
is considered as operating on a simple addi-
tive basis. It is interesting to consider these
effects on the basis of the ratio of sex chro-
mosomes to autosomes as utilized by
Bridges. As is customary, the male geno-
type is given a weight of 0.50, the intersex
0.67, the female 1.00, the superfemale 1.50,
and the triploid female 1.00. With these
values the data on the sex combs are fitted
by the equation

Sex combs = 13.40 - 6.38 X/A

The fit of this equation to these data
shows control of less of the variation in the
sex comb teeth. Only 76 per cent of the
variation is accounted for by these methods
whereas 95 per cent is accounted for when
the effects of the X and A chromosomes are
considered as additive.

If it is agreed that the condition of the
sex combs is a good unbiased measure of
the degree of sexuality of the Drosophila,
it follows that it would be more probable
that the genes in the X chromosomes oper-
ate additively with those of the autosomes.

III. Sex Genes in Drosophila

A. MUT.\XT TYPES

Bridges' concept of sex determination
turned on the action of sex genes located
more or less fortuitously throughout the
inheritance complex of the species. In Dro-
sophila it happens that the major female
determining genes seem to be located in the
X chromosome and the male determining
genes in the autosomes, whereas the Y
chromosome seems essentially empty of sex
genes. In support of this concept limited
data are cited on specific genes affecting
the reproductive system or its secondarily
differentiated elements and two cases where
genes affected the primary reproductive
system as a whole. During the interim be-
tween 1938 and the present, the numbers
of these genes and the breadth of their
known effects have been notably increased.
Again the genes as a whole affect every
phase of sexuality, morphology, fertility,

and physiology. Single genes may occa-
sionally alter both sexes or may frequently
affect only male or only female phenotypes.
Single genes may appear to influence two
or more distinct characteristics observable
in the developing flies, although this multi-
ple phenotypic expression may go back to
a gene action which is controlling a single
event in development. Genes affecting the
structural development of either male or
female organs frequently are accompanied
by sterility of various degrees. A very large
category of genes is known only through
its effects on sterility of either or both
sexes. Experience has shown that when
properly analyzed anatomic changes are
probably basic to the sterility. In this sense
genes for sterility should be considered
genes for sex characters. Berg (1937) fur-
nished data on the relative frequency of
sterility mutations in the X chromosome
as against those in the autosomes. 12.3 per
cent mutations in which the males were
sterile were found in the X chromosome
against 4.5 per cent found in the second
chromosome. These results show that the
X chromosome has many gene loci occu-
pied by genes capable of mutating to ste-
rility genes which affect males. Sterility is
also common for the females but requires
more testing. The loci for these genes are
widely distributed both within and among
the chromosomes.

Genes affecting sex morphology are found
in all Drosophila chromosomes. Of 17
which have been recently studied; 6 were
in the 1st chromosome, 5 in the 2nd, 5 in
the 3rd, and 1 in the 4th. Insofar as can be
determined these genes are no different than
those affecting other morphologic traits.
They may be dominant, they may be re-
cessive, and a limited number of them may
show partial dominance. They affect a vari-
ety of sex characteristics and do not always
involve sterility of one or the other sex.
The loci occupied by these sex genes may
have several alleles, some of which may
lead to sterility, others not. Most affect
characters like size and development of the
ovaries, the characteristics of the eggs, duct
development, spermathecae, ventral recep-
tacles, parovaria, paragonia, sex combs, po-
sition of genitalia, and so on.

14

BIOLOGIC BASIS OF SEX

Altlioiigh Drosophila is the leader in fur-
nishing types for analyses of the inherit-
ance basis for sexuality, it is well known
that similar conditions exist in other forms
of life.

B. MAJOR SEX GENES

Major sex genes affecting the dichotomy
of the sexes have appeared among the ob-
served mutational types. Sturtevant (1920a,
1921 j in Drosophila simulans isolated a
gene in its second chromosome which when
homozygous could convert diploid females
into intersexuals and render XY males
sterile. Phenotypically these intersexuals
were female-like in that they lacked sex
combs and had 7 dorsal abdominal tergites,
ovipositor of abnormal form, 2 sperma-
thecae, and lacked the penis. They were
male-like in having first genital tergite al-
though abnormal in form, lateral anal
plates, claspers, black pigmented tip to the
abdomen. The gonads were rudimentary.
The gene was recessive and, as expected,
showed no effect on D. simulans X D. mel-
anogaster hybrids.

In 1934 a new intersexual type was ob-
served by Lebedeff in D. virilis. Intensive
study of this type showed that it depended
on a fairly complex inheritance. A 3rd chro-
mosome recessive gene ix™ at 101.5 con-
verted the XX females into sterile males,
showing only one noticeable female charac-
teristic, the presence of a rudimentary 5th
stcrnite. Sterility could be accounted for,
even though the internal genitalia were
completely male, by the small size of the
testes and the degeneration of the germ
cells largely at the spermatocyte stage. Ob-
servable effects of this gene appeared only
in females, XX + 2A. Genetic modifiers
were isolated which acted on the ix"yix'"
complex. Other gcnotyi)es were derived
through recombinations of these genes. The
resulting phenotypes showed a greater vari-
ation of the male-female mosaics. One of
these types was sterile but fully hermaph-
roditic having a complete set of external
and internal genitalia of both male and fe-
male. Genetically, this type was homo-
zygous for the ix'" gene but in addition luid
both a previously found dominant scmisu})-
pressor and also a second semisuppressor.
Another line having the ix'" gene homozy-

gous had a full set of male organs but there
were rudimentary ovaries attached to the
testes. This line showed the dominant semi-
suppressor as the restraining element on the
developmental pattern of the ix'" gene. An-
other type was separable in that it was still
more female-like, yet had male external
genitalia and rudimentary testes. This type
was the result of the homozygous ix'" genes
operating in conjunction with 3 different
semisuppressors. Extensive embryologic
studies were interpreted as indicating that
gonads, ducts, and genitalia had started as
in females with the XX constitution.
Shortly thereafter male organs appeared as
new outgrowths from the same imaginal
disc. The development of the two sets of
organs was then simultaneous but still de-
pended on the gene pools present in the
particular strain.

Another intersexual type appeared in D.
pseudoobscura as a mutation, presumed due
to a single dominant gene (Dobzhansky
and Spassky, 1941). In a series of cultures
having "sex ratio," two cultures gave 234
females, 7 males, and 266 intersexes. The
females' progeny transmitted only the nor-
mal condition to their Fs progeny. The
males were sterile presumably because they
lacked a Y chromosome. The intersexes
were also sterile so that further study of
the genetic condition became impossible.
The evidence, however, is interpreted as
showing that the intersexes were trans-
formed females which had inherited a domi-
nant gene governing this condition. The
intersexes were characterized by two sets
of more or less complete genital ducts and
external genitalia but only one pair of gon-
ads. One set of ducts and genitalia was al-
most always more female-like and the
other more male-like. Sex combs were pres-
ent, the distal comb had 2 to 4, and the
proximal 4 to 6 teeth, compared with the
normal male number of 4 to 7 on the distal
and 6 to 9 on the proximal comb. Body de-
velo{)ment was more like that of the female.
Cytologic examination showed two of the
intersexes had (wo ()\-aries each. One of
these two had a rudimentary testis. The
chroinosoiiu' complement was 2X + 2A.

A dominant gene which causes intersex-
iiality in diploid females of D. virilis was
estal")lished by Briles. Stone (1942) located

FOUNDATIONS FOR SEX

15

this gene in the second chromosome. Price
( 1949j placed the gene within the second
chromosome near the locus of "brick" by
inducing crossing over in males by expos-
ing them to x-rays. Newby (1942) exten-
sively studied the embryologic sequence in
development of the organs of these inter-
sexual types. The Ix^ gene did not affect
the males, XY 4- 2A, but did change the fe-
males XX + 2A into intersexes when it was
in the heterozygous condition. The intersex-
uals had 9 tergites; the first 6 were like
those in the normal female and the last 3
were small and irregularly formed. There
were 6 sternites, the first 5 being normal,
but the 6th malformed. Anal valves were
lateral as in the male but a third small
valve was also present at the ventral side
of the anus. The plates forming the claspers
were of irregular pattern and found ventral
to the anus. The vaginal plates were often
extruded into a genital knob and were be-
low the claspers. The knob occasionally be-
came heavily pigmented. The internal or-
gans ranged from nearly female through
those which were of hermaphroditic type
containing representative organs of both
sexes to individuals almost wholly male.
Newby concluded that intersexuality ex-
presses itself as a response to the develop-
mental pressures of both sexes, not as de-
velopment in the one direction followed by
a change.

Gowen in 1940 established a stock carry-
ing the dominant gene Hr which had ap-
l)eared as a mutant in one of his cultures of
D. melanogaster. This gene affected diploid
females of XX + 2A type but not the males
of XY -I- 2A constitution. In the presence
of the Hr gene the diploid phenotype of the
females changed into a sterile type with
male secondary reproductive system associ-
ated with the female counterpart. The first
6 segments were complete with 6 spiracles.
The 7th was small with spiracle. The 8th
was small but without spiracle. Sternite
forming rudimentary ovipositor was usu-
ally protruded. Ninth and 10th segments
resembled those of males with large tergal
plates. Claspers were abnormal and had a
pair of small plates flanking the anus ma-
joi-ly in vertical position. Organs formed,
although sometimes modified or missing,
included: sex combs of 6.9 long slender

teeth, gonads distinguishable from those of
the ordinary male or female in the 3rd and
possibly the 2nd instar, genital ducts male
and/or female, male accessory gland, penis
deformed, sperm pump, vas deferens, sper-
matheca, ventral receptacle often displaced,
and occasionally parovaria. The primary
gonads were often abnormal ovaries but in
rarer instances bore a crude resemblance
to testicular tissue. The yellow of the testes
was frequently present as material clinging
to the ovary.

Superfemales with one dose of the Hr
gene had sex combs and developed parts of
both the male and female external and in-
ternal reproductive systems. Sex combs had
an average of 5 long and slender teeth. Ab-
dominal segment 8 developed as in the fe-
male, and formed the vaginal plate. The
latter was abnormal in shape, ordinarily
becoming a sclerotic protuberance. Seg-
ments 9 and 10 developed more as in the
male but were incomplete and abnormal.
The genital arch did not develop but the
inner lobe of tergite 9 showed irregular and
abnormal growth as for the claspers. Seg-
ment 10 developed, as in the males, into
longitudinal plates flanking the anus. The
internal genitalia were underdeveloped but
consisted of mixtures of male and female
organs. Gonads were rudimentary but gen-
erally consisted of a pair of ovaries with
small traces of yellow pigmentation.

The triploid fly with one dose of the Hr
gene was largely female with developed
ducts, ovaries and eggs, but was sterile. The
male characteristics were small sex combs
and dark abdominal plates. In superfe-
males, sex combs were present and teeth
were intermediate between those of the dip-
loid female and triploid female. The gene
showed a dosage effect in triploids which
was less than that observed in diploids and
was in relation to the relative balance of
the gene with its normal alleles, 1:2 for
the triploid and 1 : 1 for the diploid. The
developmental effects of Hr as well as the
pigment producing potentialities of testes,
ovaries and hermaphroditic gonads have
been discussed by Fung and Gowen (1957a,
b).

Hr has been shown to be allelomorphic to
a recessive gene, tra, described by Sturte-
vant (1945) and known to be located in the

16

BIOLOGIC BASIS OF SEX

3rd chromosome. The location of this allele,
tra, is at 44 to 45 or between the genes scar-
let and clipped. When homozygous the gene
transformed diploid females into sterile
males. Heterozygotes showed no detectable
differences from normal females of XX con-
stitution. Males XY homozygous for tra
or heterozygous for it were indistinguish-
able from normal males. The homozygotes
XX, tra/tra were female in body size, but
otherwise were nearly male in appearance.
They had fully developed sex combs, male
colored abdomens, normal male abdominal
tergites, anal plates, external genitalia,
genital ducts, sperm pumps, paragonia, and
showed the usual rotation of the genital and
anal segments through 360 degrees. They
mated with females readily and normally.
The testes, however, although normal in
color, elongated, curved, and attached to
the ducts were of small size. Testis size was
never that found in normal brothers. The
addition of a single Y or two Y's did not
alter fertility.

The triploid females 3X + 3A homozy-
gous for tra, had large bodies, ommatidia
and wing cells. They resembled the diploid
homozygous tra individuals in having male
external genitalia, well developed ejacula-
tory ducts, sperm pumps, and accessory
glands, testes elongated but narrower than
those of normal males, and sex combs aver-
aging about 9.6 teeth. They mated with fe-
males but were completely sterile.

Triploids with one or two doses of tra
were like wild type triploids in having no
sex combs and being female throughout. In-
tersexes having one or two doses of tra were
similar to intersexes having only wild type
genes in the locus.

Sturtevant obtained one superfemale
which was homozygous for tra. It had male
genitalia and sex combs with only about
half the normal number of teeth. This in-
dividual argues for a greater balance to-
ward the female side of sexual development
than either the diploid or triploid females
previously discussed. The evidence is, how-
ever, contradictory to that furnished by the
Hr gene as indicated earlier.

A combination of two or more genes,
Beaded and various Minutes, having well
known phcnotypic effects, has sometimes

produced phenotypes which have been in-
terpreted as peculiar, low grade types of
intersexuality in males (Goldschmidt,
1948, 1949 and 1951). The data showed that
the Beaded cytoplasm favors the low grade
intersexual male whereas the Minute cyto-
plasm favors the reduced male with the
hetcrozygote being intermediate. Just how
far these types may be related to the other
types strongly affected by specific genes is
a matter of question, having at least other
interpretations (Sturtevant, 1949).

In 1950, Milani trapped an inseminated
female of D. subobscura w^iich segregated
intersexual progenies. Spurway and Hal-
dane (1954) studied these intersexual
types. A recessive guiding development to-
ward these intersexes was located on the 5th
chromosome of subobscura. When present
it caused the XX homozygous females,
ix/ix + 2A, to have sex combs on both the
first and second tarsal joints. The numbers
of teeth making up the sex combs w^ere re-
duced as also were the sizes of the teeth.
The illustration in Spurway and Haldane's
(1954) paper indicates that the number of
teeth was 7 on the first tarsal joint and 5
on the second joint, whereas the sex combs
of the males had 11 teeth on the first joint
and 9 on the second. A series of changes
were observed in the genital plates which
graded from those resembling true females
to those approaching the male type.

C. OTHER CHROMOSOME GROUP ASSOCIATIONS I

DROSOPHILA AMERICANA

I), aniericana has 4 chromosomes in the
female genome and 5 chromosomes in the
male genome. As compared with D. virilis
the X chromosome is fused with the 4th
chromosome and the 2nd chromosome is
fused with the 3rd, the 5th and 6th chromo-
somes are free in the female, whereas in the
male genome the Y chromosome, 4th, 5th
and 6th chromosomes are free and the 2nd
and 3rd fused. Stalker (1942) has shown
that the three female genomes are balanced
and lead to triploid females as they do in

D. melanogaster. D. aniericana triploids
differ from their diploid sisters in having
bigger ommatidia, larger wing cells and
somewhat larger bodies. When these trip-
loids are bred to diploid males they give

FOUNDATIONS FOR SEX

17

rise to 6 chromosomally different types of
offspring: diploid males, diploid females,
triploid females, intersexes, females carry-
ing a Y and a male limited 4th chromosome
hut otherwise diploid, and tetraploid fe-
males. All intersexes cytologically show a
Y and a 4th chromosome present. Intersexes
without the Y are presumed also w^ithout
male limited 4th chromosomes and would
be expected to be inviable or very weak.
No supermales or superfemales, that w^ould
correspond with those found in D. melano-
gaster, were observed so are presumed to
l)e inviable due to unbalance for the 4th
chromosomes. Among 948 progeny of trip-
loid females X diploid males there were 9
individuals that were phenotypically abnor-
mal females. They had slightly spread, ven-
trally curved wings with slightly enlarged
wing cells. In 8 of the 9 the first section of
the costal vein was shortened so that no
junction was made with the first vein at the
distal costal break. Heads were large with
rough eyes, thoraxes shortened, legs fre-
(luently malformed, and abdomens small
with unusually wide 7th sternites. Genitalia
were apparently normal with well devel-
oped ovaries. This type carries three doses
of any genes contained in the 4th chromo-
some to two doses of the genes in the other
chromosomes. Its phenotype represents a
trisomic condition.

The sex characteristics of the flies ob-
served in D. americana seem to follow the
same patterns as those of D. melanogaster
as judged by the numbers of the X chromo-
somes and autosomes. A Y chromosome in
the intersex was not observed to affect sex
expression. The intersexes could be grouped
into six classes ranging from extreme male
type to the most female type. The male
type showed largely male organs, courted
females, and had motile but nonfunctional
sperm. The most female type had nearly
normal ovipositor plates, well developed
uterus, ventral receptacle, spermathecae
and oviducts. At least one gonad showed
egg strings, although a small patch of
orange-red tissue was present at the tip.
Two types of chromosomes were observed
in the nuclei of these extreme female type
intersexes; those like the chromosomes in
any normal diploid cells and some which

were so swollen as to be almost unrecog-
nizable. Such swollen chromosomes were
not found in the other classes of intersexes
or in diploid or triploid individuals. They
are suggestive of some noted by Metz
(1959) in Sciara. Most of the intersexes
were of the male type, 45 per cent, with de-
creasing numbers for each of the other five
classes until those in the most female class
constituted only about 4 per cent of the
total.

D. LOCATION OF SEX-DETERMINING GENES

The problems of isolating and determin-
ing the modes of action of the factors nor-
mally operating in sex determination have
received extensive study since they were
reviewed by Bridges in 1939: Patterson,
Stone and Bedichek (19371, Patterson
(1938), Burdette (1940), Pipkin (1940-
1942, 1947, 1959), Poulson (1940), Stone
(1942), Dobzhansky and Holz (1943),
Crow (1946), and Goldschmidt (1955).
From his work on Lymantria dispar, Gold-
schmidt concluded that sexual differentia-
tion was controlled by a major male factor,
M in the Z chromosome and a factor F
directing development toward the female
and at first assumed to be in the cytoplasm
but later considered to be in the W chromo-
some. The heterogametic female of this
species would then be FM and the male be
MM. In considering this problem, Gold-
schmidt attempted to distinguish between
the sex determiners responsible for the F/M
balance and modifiers affecting special de-
velopmental processes (Goldschmidt, 1955).
At the other extreme Bridges' study of trip-
loids led him to consider that sex in Droso-
phila was determined by the interaction of
a number of female tendency genes found
largely in the X chromosome and of genes
having male bias located largely in the au-
tosomes. These numerous genes were con-
sidered as being distributed throughout the
whole inheritance complex. Search for the
more exact locations of these genes within
the different chromosomes of Drosophila
has largely taken the form of determining
the variation in sex types as induced by the
addition or deletion of various pieces of the
different chromosomes to either the normal
male, normal female, or triploid complexes.

18

BIOLOGIC BASIS OF SEX

The sections of the chromosomes added
were derived from previous translocations
generally to the 4th chromosome and were
of varying lengths determined through cy-
tologic and genetic study. Summaries of
these comparisons are found particularly in
the papers of Pipkin (1940, 1947, 1959). In
their search for a major female sex factor
in the X chromosome, Patterson, Stone and
Bedichek (1937), Patterson (1938), Pipkin
(1940), and Crow (1946) finally were un-
able to show that any single female sex de-
terminer, located in the sex chromosome,
was of primary importance to sex. Evidence
for multiplicity of genes with a bias toward
female determination was found by Dob-
zhansky and Schultz (1934) and by Pipkin
(1940). They were able to transform dip-
loid intersexes into weakly functioning hy-
potriploid females by the addition of long
fragments of the X chromosome to the
2X + 3A intersex complement. Short sec-
tions of the X chromosome in some cases
shifted the sex type in the female direction.
Pipkin found that additions of short X
chromosome sections, to the 2X -|- 3A chro-
mosome sets, although covering in succes-
sion the entire X chromosome, were insuffi-
cient to make the flies other than of the
intersexual type. Longer and longer frag-
ments from either the left or right end of
the X chromosome caused a qualitatively
progressive shift toward femaleness.
Weakly functional hypotriploid females re-
sulted when either right- or left-hand sec-
tions of two translocations with t-lz (17)
and Iz-v (W13) breaks were present in the
2X + 3A chromosome complement. These
duplication intersexes possessed 1 or 2 sex
comb teeth when reared at 22 °C. and up to
5 well developed teeth when reared at 18°C.
These facts give support to the multiple
sex gene theory of Bridges or at least that
quantitative differences in sex potencies
exist within the X chromosome. This con-
clusion was further strengthened by the hy-
pointersexes lacking a short portion of one
of their two X chromosomes although pos-
sessing three of each autosome, inasmuch
as these types were shifted strongly in the
male direction. These studies of the X cliro-
mosonie show that several parts of this
chi-omosomc are concerned witli female dif-

ferentiation and that the effects are irregu-
larly additive.

Similar search of the autosome II and
III for genes of male potency showed that
small shifts in the male direction were
found in hyperintersexes for several short
regions of chromosome III but for none of
chromosome II (Pipkin, 1959, 1960). Tl^ree
slightly different right-hand end regions of
chromosome III produced the largest shifts
in the male direction in hyperintersexes,
but no increase in number of sex comb
teeth. These changes were comparable with
those produced in the female direction by
the addition of very short sections of the
X-chromosome to the 2X + 3A intersex
complement. On the other hand, none of the
seven different hypointersexes lacking a
short section of the 3rd chromosome from
the 2X + 3A complement showed a shift in
the female direction. This is rather surpris-
ing as hypointersexes for two short regions
in the X chromosome were shown by Pip-
kin (1940) to shift the sex type in the male
direction as was to be expected. From these
results Pipkin (1959) derives the conclu-
sion that 3rd chromosome aneuploids as
well as those of the 2nd chromosome and
X chromosome support the deduction that
dosage changes of portions of the X chro-
mosome are more powerful than dosage
changes of portions of either of the large
autosomes in affecting sex balance. This
view receives further support through
changes of size and number of sex comb
teeth as observed in the chromosomal types
carrying the gene Hr and reviewed earlier.

Influence of the Y chromosome on sexual
differentiation has generally been ruled out
as XO nondisjunctional flies are male al-
though they are sterile (Bridges. 1922).
Similarly Dobzhansky and Schultz (1934)
ruled out the Y chromosome as an effective
influence on sex types of triploid intersexes
of D. melanogaster since the mean sex type
of Y -(- 2X 4- 3A intersexes did not differ
significantly from the sex tyyies of siblings
2X + 3A. "

The steps taken in the studies of chro-
mosome IV are of interest. Dobzhansky
and Bridges in 1928 concluded that the 4th
chromosomes i^lay no part in sex determina-
tion in D. melanogaster. The evidence was

FOUNDATIONS FOR SEX

19

of two kinds. Triploid females were out-
crossed separately to males of two diploid
stocks. The triploid daughters from these
crosses were again crossed to males like
their fathers. This repeated outcrossing to
the different stocks resulted in a shift in
the grade of the intersexes in both cases, in
one case to a very high proportion of ex-
treme male-like intersexes and in the other
to nearly as high a proportion of extreme
female-type intersexes. These results were
interpreted as showing that the grade of
development of the sexual characters was
dependent on genetic modifiers. The second
experimental test consisted of subjecting a
triploid stock to selection toward a line
which produced a high proportion of ex-
treme male type intersexes and to another
line which would have a high proportion of
extreme female type intersexes, each being
much higher than the original stock. The
l)rocedure established a line with a high
proportion of extreme male type and an-
other line which was not so extreme in its
proportions but was definitely higher in fe-
male type intersexes than the original stock.
Again these results w'ere interpreted as in-
dicating the selection of modifying genes
of unknown positions within the inheritance
complexes.

This evidence had been preceded by
Bridges' (1921) discussion in which he
wrote "the fourth-chromosome seems to
have a disproportionately large share of the
total male-producing genes; for there are
indications that triplo-fourth intersexes are
predominately of the 'male-type', while the
dijjlo-fourth intersexes are mainly 'female-
type'." In 1932, Bridges concluded for Dro-
sophila intersexes that, in spite of the fa-
vorable genetic checks, in repeated and
varied tests, it has been impossible to state
with any assurance whether the 4th chro-
mosome is or is not a large factor in the
variability encountered.

In our own work a stock of attached X
triploids has for many years consistently
produced only male type intersexes. This is
in contrast to what we frequently see within
other lines of triploids as made up utilizing
the cIIIG gene (Gowen and Gowen, 1922).
Lines established from these triploids or-
dinarily have three intersexual types: male,

intermediate, and female. These lines, how-
ever, may be subjected to selection in both
directions. In our experience, male intersex
lines are established rapidly and remain
relatively permanent. On the other hand,
female intersex lines take many more gen-
erations and are less stable. These lines
have been extensively examined for their
4th chromosome constitutions (Fung and
Gowen, 1960). The male intersex lines sel-
dom show more than two 4th chromosomes.
On the other hand, the female intersex lines
rarely show two 4th chromosomes but gen-
erally have more than three, the number
sometimes going as high as four. More tests
are needed but the evidence would seem to
indicate that the fourth chromosome does
have sex genes. These genes, contrary to the
first notion of Bridges, are more frequently
of the female determining type than of the
male determining type. This would make
the 4th chromosome like the X in that it
carries an excess of female influencing genes
and is not like the rest of the autosomes
which have an excess of male determining
genes. These observations are of particular
interest in view of Krivshenko's (1959) pa-
per. In this investigation on D. busckii, cy-
tologic and genetic evidence was presented
for the homology of a short euchromatic
element of the X and Y chromosome with
each other and also with the 4th chromo-
some or microchromosome of D. melano-
gaster. This conclusion is based on (1) ob-
served somatic pairing of the X and Y of
D. busckii by their proximal ends in gan-
glion cells and the conjugation of the short
euchromatic elements of these chromosomes
at their centromeric regions in the salivary
gland cells; (2) the presence in the short
Y chromosomal element of normal allelo-
morphs to four different mutant genes of
the short X chromosomal element; (3) the
presence in the short element of the D.
busckii X chromosome of chromosome IV
mutants: Cubitus interruptus. Cell and
shaven of D. melanogaster. These consider-
ations furnish proof for the homology of
this X chromosomal element with the 4th
chromosome of Drosophila.

These observations of Krivshenko sup-
port our findings that the 4th chromosome
of D. melanogaster has an excess of female

20

BIOLOGIC BASIS OF SEX

determining functions. It would further
show that autosomes may behave differ-
ently with regard to their sex-determining
properties according to the chance distribu-
tion of sex genes which happen to fall
within them as they do in Rumex (Yama-
moto, 1938j. The finding that the chromo-
some IV has a bias toward female tenden-
cies further strengthens Bridges' multiple
sex gene theory and weakens the theory of
an all-or-none action of the whole X chro-

IV. Sex under Special Conditions

A. SPECIES HYBRIDITY

Hybrid progeny coming from species
crosses are apt to represent but a very few
of the possible genotypes of the total num-
ber that conceivably could come from the
gene pool. The hybrid phenotypes may dis-
play three kinds of characteristics. The
common set is that derived from genes in
either or both parents through ordinary
meiotic segregations and dominance. The
second set shows intermediate develop-
ment of the characters found in the two
parent species. The third set of characters
that complete the animal is new to those
observed in either parent species. These new
characters may be the loss of a few dorso-
central and scutellar bristles, broken or
missing cross veins, or abnormal bands in
the abdomen as in D. simulans x D. melano-
gaster hybrids (Sturtevant, 1920b), extra
antigenic substances as found in dove hy-
brids (Irwin and Cole, 1936), or more nu-
merous characteristics as in the mule. Fre-
quent among these new characteristics is
sterility. The sterility may extend to either
or both sexes and affect the secondary sex
ratios. As Sturtevant (1920b) points out,
crosses between the domestic cow and male
bison give male offspring with humps de-
rived from the bison which are so large as
to prevent their being born alive. The fe-
male hybrids lack these humps and are con-
se(iuently born normally. The abnormal sex
ratio observed at time of birth is due to
causes external to the hybrid itself and at-
tributable to the structure of the mothers.

A comparable case was found during the
study of female sterility in interspecies hy-
brids of Drosophila pseudoobscura in which

Mampell (1941) showed that in the hybrids
of certain strains, the females produced no
or few offspring because of interspecies le-
thal genes connected with a maternal effect.
Comparable cases as well as those depend-
ent on other mechanisms are known for
other groups. The progeny may also be al-
tered to give new sex types, generally inter-
sexes. These intersexes often replace either
the male or the female sex group. However,
despite their apparent relation, the changes
in the sex ratio and the appearance of in-
tersexes can have different causes. D. sim-
vlans X D. melanogaster hybrids emphasize
that there may be no relation between the
peculiar hybrid sex ratios and the intersexes
since extreme differences in sex ratio occur
but no intersexual types.

Species, however, may have natural dif-
ferences in the sex potencies of their X
chromosomes and/or their autosomes. In
crosses between D. repleta and D. neo-
repleta involving a sex-linked recessive
white-eyed mutant type of D. repleta Stur-
tevant (1946) obtained about 15 per cent
fertile matings in 500 mass cultures, a total
of 532 females to 635 males. All progeny as
expected were wild type in character. The
males, however, had long narrow testes and
were totally sterile, a condition later shown
to be due to a gene in the X chromosome
located near the white locus. Females sug-
gested intersexuality in having three anal
plates instead of the usual two. Mating of
Fi hybrid females to white D. repleta males
gave 9 per cent fertility, the 179 offspring
being distributed as 70 wild type females,
9 white females, 42 wild type males, and 58
white males, although the expectation for
the classes was equality. Evidence indicates
that some of the 9 white females were in-
tersexes as were possibly some of the white
males. The wild-type males again had the
long narrow testes and sterility of the Fi
male progeny. Wild type females were mod-
erately fertile. By continued backcrossing
to 1). repleta males having white or white-
singed, a female line was picked up which
continued to have the unusual sex ratios
but had more fertility. It was presumed
that the D. neorepleta gene responsible for
the unusual ratios was originally associated
with MHotlier gene that decreased fertility

FOUNDATIONS FOR SEX

21

in females largely of D. repleta constitution
and that the foundation female for the
more fertile line came as a result of a cross-
over between an infertility gene and that
responsible for the unusual sex ratios. Con-
tinued back crosses of females of this line
to white D. repleta males have been made.
Out of 33 fertile cultures, 16 gave approxi-
mately equal ratios of wild-type and white
females, wild-type and white males; and
17 gave 472 wild-type females, 5 white fe-
males, 63 white intersexes, 482 wild-type
males, and 339 white males. The white fe-
males presumably represented crossovers
between the loci of white and the critical
gene in the X derived from D. neorepleta.
The intersexes were of extreme type with
gonads very small (rudimentary ovaries in
those cases where they were found at all).
External genitalia were missing or of ab-
normal male type. Other somatic character-
istics included weakness which prevents
emergence and accounts for the loss of
about 88 per cent of the flies expected in
that class. The intersexual condition was
suggested as being caused by an autosomal
dominant gene derived from D. neorepleta
which so conditions the eggs before meiosis
that two D. repleta X chromosomes result
in the development of intersexes rather
than females. The action of this gene occurs
before meiosis and may in fact be absent
from the intersexes themselves. This was
confirmed by crosses of white brothers of
the intersexes to pure D. repleta females
when the offspring were normal for both
sexes; but when these Fi daughters were
mated to D. repleta males only intersexes
and males resulted. This last cross further
showed that although this gene was derived
from D. neorepleta in D. neorepleta cyto-
plasm, the D. neorepleta cytoplasm was not
necessary for the intersexes to result. The
case also has an important bearing on the
location of the sex-determining factors, for
in this cross the characteristics were only
secondarily governed by the cytoplasm
through earlier determination by genes of
the mothers' nuclei.

Significant parallels are found between
the autosomal gene of D. neorepleta and
the third chromosome Ne gene of D. mela-
nogaster (Gowen and Nelson, 1942) de-
scribed in the section on high male sex

ratio. The D. neorepleta gene caused the
cytoplasms of the eggs laid by mothers
carrying it to become more male potent.
The female potencies of two X chromosome
D. repleta zygotes were unable to balance
these male elements. Many died late in de-
velopment. Those able to emerge became
intersexes. The Ne gene also sensitizes the
cytoplasms of all eggs of mothers carrying
it causing any 2X + 2A, 3X + 3A, 3X +
2A or 2X -h 3A intersexes of female type to
die in the eggs at 10 to 15 hours whereas
males XY + 2A and male-type intersexes
live.

Other mechanisms for causing sex and
sex ratio changes are known, i.e., Cole and
Hollander (1950), but few are as well
worked out as that of D. repleta x D. neo-
repleta. New mechanisms will certainly be
found for the opportunities for genetic anal-
ysis of sex in hybrids are many.

B. MOSAICS FOR SEX

Recent genetic work has emphasized the
fact that individual D. melanogaster may
be composed of cells of more than one genie
or chromosome constitution. The main type
of sex mosaic is the gynandromorph com-
posed of cells of female constitution on one
side, XX + 2A, and male, X + 2A on the
other, the loss of the X chromosome coming
at an early cleavage (Morgan and Bridges,
1919; L. V. Morgan, 1929; Bridges, 1939).
The mosaic areas are large since the cells of
each type may be in nearly equal numbers.

At the other extreme Stern (1936) has
shown that phenotypic mosaics may de-
velop as a consequence of the somatic chro-
mosome pairs crossing over at late stages
in embryologic development. Special genes,
Minutes, materially increase the frequen-
cies of these crossovers. The proportion of
the body occupied by the cross-over type
cells is small because crossing over takes
place so late in development.

Recently another agent in the form of a
ring chromosome has been discovered which
greatly increases the production of sex mo-
saics. Some ring chromosomes are relatively
stable whereas others are quite unstable,
the instability depending to some extent on
aging of the eggs and environmental factors
(Hannah, 1955). The instability is manifest

22

BIOLOGIC BASIS OF SEX

by frequent gynandromorphs, XO males,
and dominant lethals among the rod and
ring zygotes. It has been suggested that the
instability is due to heterochromatic ele-
ments. Hinton (1959) has observed the
chromosome behavior of these types in
Feulgen mounts of whole eggs that were in
cleavages 3 to 8. He found strikingly ab-
normal chromosome behavior in these
cleaving nuclei. For some cell divisions
chromosome reproduction was interpreted
as being through chromatid-type breakage
fusion bridge cycles. As a result of this be-
havior mosaics are formed which are inter-
mediate between those of the half gynan-
dromorphs and those which occur much
later because of somatic crossing over. In
terms of volume of cells included, the ab-
normal types may include only a few cells
of the total organism, a fair proportion of
the cells, or a full half of the whole body.
These unstable ring chromosome mosaics
may be a part of the secondary reproduc-
tive system or for that matter any other
region of the body. When the mosaic cells
are incorporated in the region of sex organ
differentiation male or female type organs
or parts of organs may develop as governed
by the cell nuclei being X, XX or some frac-
tion thereof.

Gynandromor|)lis appear sporadically and
rarely in many species but in some in-
stances genes which activate mechanisms
for their formation are known. In the pres-
ence of recessive homozygous claret in the
eggs of D. siniulans, gynandromorphs con-
stitute a noticeable percentage of the
emerging adults. The gene nearly always
operates on the X received from the mother
causing it to be eliminated from the cell.
The resulting gynandromorphs are similar
to those of D. melanogaster. The fact that
the claret gene should affect the X and a
particular X chromosome is suggestive of
the manner in which given chromosomes arc
eliminated in Sciara. Other types of sex
mosaics will be found in the descriptions of
other species, i)articularly in the Hyme-
noptci'a.

C. PARTHENO(iENESIS IN DROSOPHILA

Parthenogenesis is of interest as it
changes the sex ratios in families and brings
to light new sex types and novel methods

for their development (Stalker, 1954), A
survey of 28 species of Drosophila showed a
low rate of parthenogenesis in 23 species.
Adult progeny were obtained for only 3
species. For D. 'parthenogenetica the origi-
nal rate was 8 in 10,000 whereas that for D.
polymorpha was 1 in 19,000. These rates
could be increased by selection of higher
rate parents: 151 and 70 per 10,000 unfer-
tilized eggs of the first and second species
respectively.

D. parthenogenetica diploid virgins pro-
duced diploid and triploid daughters as well
as rare XO sterile diploid males. Triploid
virgins produced diploid and triploid fe-
males and large numbers, 40 per cent, of
sterile XO diploid males. Diploid virgins
heterozygous for sex-linked recessive garnet
produced homozygous and heterozygous
diploid females as well as +/+/g and
+ /'g/g triploid females. No homozygous
wild-type or homozygous garnet triploid fe-
males or garnet mosaics were found. Dip-
loid females crossed to fertile diploid males
produced few if any polyploid progeny or
jjrimary X chromosome exceptional types.
Of the unfertilized eggs from diploid virgins
which started development, 80 per cent died
in late embryonic or early larval stages.
The i)arthenogenesis in diploid females de-
pended on two normal meiotic divisions fol-
lowed by fusion of two of the derived hap-
loid nuclei to form diploid progeny, or the
fusion of three such nuclei to form triploitl
progeny. In the triploid virgins similar fu-
sions of the maturation nuclei may produce
diploid and triploid females but the large
number of dii)loid XO sterile males were
picsunicd to be the result of cleavage with-
out prior nuclear fusion. Such cleavages
without fusion in eggs of dijiloid virgins
would lead to the production of haploid
embryos. They were presumed i-esponsible
for the large early larval and embryonic
<h'atlis. These obser\ali()ns have been con-
firmed by the study of S|)rackling (1960)
in\-olving some 2200 eggs at various stages
of cleavage. Evidence from XXY diploid
virgins indicated that biiuiclcar fusion in
unfertilized eggs involved two terminal
haploid nuclei or two central nuclei. The
fact that tetraploids were not observed as
progeny of ti'iploid vii'gins was considered
indicative of relative inviability of this

FOUNDATIONS FOR SEX

23

type. Successful parthenogenesis was under
partial control of the inheritance as 80 gen-
erations of selection increased the rate
about 20-fold. Similarly outcrossing to bi-
sexually reproducing males and reselection
resulted in both pronounced increases in
and survival of the parthenogenetic types.

Carson, Wheeler and Heed (1957) and
Murdy and Carson (1959) have established
a strain of Drosophila mangabeirai with
only thelytokous reproduction. Males have
been captured in nature but are rare. Fe-
cundity of the virgin females is low but the
egg hatch is 60 per cent and 80 per cent
survive to adult stage. The progeny of the
virgins are always diploid. In meiotic spin-
dle formation D. mangabeirai differs from
other Drosophila species in that its orienta-
tion increases the probability for fusion of
two haploid nuclei into structurally het-
erozygous diploid females. The study of
Feulgen whole mounts of freshly laid eggs
indicated automictic behavior with two
meiotic divisions followed by a fusion of
two of the four haploid meiotic products.
The absence of adult structural homozy-
gotes in wild populations is probabl}^ ex-
plained by death during early development
and by possible fusion of second division
meiotic products derived from different sec-
ondary oocytes. Stalker (1956a) postulated
such selective fusion in order to account for
the heterozygous condition in females of
Lonchoptera dubia, an automictic, parthe-
nogenetic fly in which males are rare or un-
known.

A case in which jihenotypically rudimen-
tary females, supposed homozygous for r/r
give rare and unexpected type progeny, has
suggested that polar body fusion may also
take place in D. melanogaster (Gold-
schmiclt, 1957). The rudimentary mothers
producing the peculiar type are interpreted
as formed by a most unusual series of
events: a fertilization nucleus derived from
the fusion of an r containing egg fertilized
by an r containing sperm and a polar copu-
lation nucleus derived from the fusion of a
polar body containing an r genome and one
having wild type. Both cell types become
incorporated into the ovary. The progeny
which come from these supposed rudimen-
tary mothers are presumed to be derived
from the maturation into eggs of the cells

derived from the heterozygous i)olar copu-
lation nuclei. If these progeny-producing
rudimentary mothers arise in the presumed
manner they give the basis for a partheno-
genetic mode of reproduction and the sex
types which have been described in other
Drosophila species by Stalker and Carson.
There are similar, as well as other forms
of parthenogenesis which affect sex (Smith,
1955) or assist in maintaining trijiloid con-
ditions (Smith-White, 1955). A number of
these types have been reviewed by Suoma-
lainen (1950, 1954). The reader may be re-
ferred to this material for other cases and
chromosome behaviors.

D. SEX INFLUENCE OF THE Y CHROMOSOME

The first function discovered for the Y
chromosome in D. melanogaster was that it
was necessary to male fertility (Bridges,
1916). Two and possibly more Y chromo-
some-borne, genetic factors were involved
(Stern, 1929). Gamete maturation when
these factors were lacking ceased just short
of the sperm's becoming motile (Shen,
1932). The motility conferred on the sperm
by the presence of the Y chromosome fac-
tors was fixed for the testes at an early
stage of development as transplantation ex-
periments, sterile testes to fertile larvae
and fertile testes to sterile larvae, showed
motility to be a property determined by
early localized somatic influences on the
developing gametes or predetermined in the
diploid phase (Stern and Hadorn, 1938).
This Y chromosome function was sex lim-
ited, because females without a Y were the
normal fertile females and those with an
extra Y also were fertile.

Neuhaus (1939) followed by Cooper
(1952, 1959) and Brosseau (1960) further
analyzed the Y chromosome for fertility
loci. The latter showed at least two fertility
loci on the short arm and five on the long
arm of the Y chromosome. Data are com-
patible with a linear order of the genes. An
additional fertility factor common to the X
and Y was suggested. The Y chromosome
fertility factors consequently fall in line
with Bridges' concept of multiple gene loci
distributed in a more or less random man-
ner which may affect sex.

The Y chromosome has other attributes
which help to explain its significance to sec-

24

BIOLOGIC BASIS OF SEX

ondary if not primary sex characteristics.
As with many other species it has loci for
genes which are also found in the X chro-
mosome, i.e., bobbed, as well as a limited
number of bands in the salivary gland chro-
mosome. Nucleolus organizers and at least
two specialized pairing organelles similar
to those in the X chromosome are found
within the Y chromosome (Cooper, 1952,
1959). One of the most significant proper-
ties is the effect of extra Y chromosomes on
the variegation observed for various gene
phenotypes either in the normal chromo-
some pattern or in that accompanying
translocation. In variegation one Y chro-
mosome as extra to the normal complex is
sufficient to eliminate or more rarely to
much reduce the variegated expression
(Gowen and Gay, 1933). When the Y chro-
mosomes are 2 above the normal comple-
ment, the phenotypes gain two new features
(Cooper, 1956). Both males and females
become variegated in the expression of their
eye characteristics. The males become ster-
ile. These effects are unexpected for they
are counter to any previous trends in the Y
effects on these characteristics. They have
reversed the direction of the effects as es-
tablished by the two previous chromosome
types. The variegation of the XX2Y + 2A
females and X3Y + 2A males resembles
that of the XX + 2A females and XY +
2A males but is more extreme, whereas the
XXY + 2A and X2Y + 2A are largely
nonvariegated. The fertility relations are
equally aberrant: the X3Y + 2A males
have the same type sterility through loss of
sperm motility as that of the XO + 2A
males. Bundles of sperm are formed but
they do not become motile. Full-sized Y
chromosomes are not required to bring
about these effects, because females having
a whole Y plus a piece of a second, or males
with 2Y plus a piece of a third, will show
the effects.

The fractional Y chromosomes furnish
opportunities to test for the partial inde-
pendence of the variegation and sterility ef-
fects. The two Y hyperploid males differ in
their degree of fertility according to the
fraction of the Y chromosome which may
be present whereas the effects on variega-
tion may be constant among groups. This is
in accord with the sterility l)eing in part

independent of the factors causing the var-
iegations. Similarly, the variegations may
be shown to be partially free of the action
of some elements that are not themselves
members of the two sets of factors influ-
encing fertility in the normal male.

Other phenotypic irregularities appear;
eye facets may be roughened, legs short-
ened, and wing membranes become abnor-
mal. On the negative side two extra Y chro-
mosomes in females homozygous for the
transformer gene, tra, do not increase in
maleness or function.

The variegations of the so-called V-type
position effects with translocations are sup-
pressed, as in the normal type described,
by one extra Y but are nonetheless vari-
egated when two extra Y chromosomes are
present. That these effects are caused by
there being two supernumerary Y's is in-
dicated by the fact that XX2Y females are
fertile and X3Y males sometimes lose a Y
chromosome in their germinal tracts and
become fertile. A number of mechanisms
have been suggested to account for these
results but most have proven unsatisfac-
tory. A balance interpretation for the X, Y
and autosomes like that for sex, as sug-
gested by Cooper (1956) is compatible with
the somatic cell variegations for euchro-
matic loci transferred to the heterochro-
matic regions and the sterility-fertility
relations expressed by the different chromo-
somal types.

The variegation effects of the Y chromo-
some take on further significance. Baker
and Spofford (1959) have shown that 15
different fragments of the Y chromosome
when studied for their contributions to var-
iegation differ in their effects with the dif-
ferences often not related to the size of the
fragment, thus indicating that the Y chro-
mosome has linearly differentiated factors
capable of modifying the variegated pheno-
type.

Other indications of genetic activity of
the Y chromosome were given by Aronson
(1959) in her study of the segregation ob-
served in 3rd chromosome translocations. A
deficiency for the region of the 3rd chromo-
some centromere is lethal when homozy-
gous. Both males and females are fully
\'iable when this deficiency is heterozygous.
XO or haplo IV deficient males die. How-

FOUNDATIONS FOR SEX

25

ever, in the presence of a Y chromosome the
deficient haplo IV males become viable.
The lability of this last class indicates that
the Y chromosome is genetically active, can
compensate for the autosomal deficiency,
and thus alter the progeny sex ratio.

In D. virilis the situation is somewhat
different from that observed by Cooper in
1956 in D. melanogaster. Baker (1956) has
shown that, in a translocation, males having
two Y chromosomes plus a Y marked with
a 5th chromosome peach are fertile. These
results seem to indicate a species difference
in the effect of the extra Y on fertility or
the Y with the inserted peach locus is not
a complete Y and in consequence the true
composition of these males is X + 2Y plus
a fragment of the Y. The X chromosomal
associations in the multichromosomal types
are shown to be by trivalents or by tetra-
valents. The segregation data indicate that
the pattern of disjunction of trivalents is a
function of the particular Y chromosome
involved. In X2Y males with normal Y's or
with one normal and one marked Y, the Y's
disjoin almost twice as frequently as they
do from trivalents with two identical Y's.
Tetravalent segregation is almost entirely
two by two, with no preference for any of
the three types of disjunction.

An odd situation was reported by Toku-
naga (1958) in substrains of Aphiochaeta
.vanthina Speiser. When a male of the sub-
strain was crossed to individuals bearing
3rd chromosome genes of the original
strains, the mutant genes for brown, and so
on, behaved as if they were partially sex-
linked in the following generations. On the
other hand, when a female carrying the par-
tially sex-linked genes on the X and Y
chromosomes. Abrupt or Occhi chiari, was
crossed to the male of the substrain the
characters segregated as though they were
autosomal. As a working hypothesis it was
suggested that in this species the Y chro-
mosome had the major male determining
factors. The "special" male arose as a trans-
location of these factors to the third chromo-
some with the consequent change in linkage
relations. Data on the role of the X chromo-
some in sex determination in this species
have not yet been obtained but if they sup-
port the interpretation they indicate real
differences between this species and that of

Drosopliila in the location of the sex genes.
The results are reminiscent of those ob-
tained by Winge in Lebistes as well as those
in Melandrium and other forms in which
the Y or W chromosomes may contain
strong sex genes for either sex. They would
further support the thesis that sex genes
may be distributed to almost any loci within
the inheritance complex.

E. M.\TERNAL INFLUENCES ON SEX RATIO

Aside from chance and specific genetic
factors, sex ratio is subject to effects from
agents intrinsic in the cells of the mothers
(Buzzati-Traverso, 1941; Magni, 1952).
Strains of Drosophila bifasciata (Magni,
1952, 1953, 1957), D. prosaltans (Caval-
canti and Falcao, 1954), D. willistoni and
D. paulistorum Spassky, 1956 have been
isolated which were nearly all of the female
sex even though the mothers were outbred
to other strains having normal ratio bisexual
progeny. Study of these strains by the above
workers and Malogolowkin (1958) IVIalo-
golowkin and Poulson (1957), and Malo-
golowkin, Poulson and Wright (1959), as
well as by Carson (1956) have shown in-
heritance strictly through the mother re-
gardless of the genetic nature of the males
to which they were bred. The unbalanced
ratio was retained even when the original
chromosomes had been replaced by homolo-
gous genomes from lines giving normal male
and female progenies. Transmission of this
unbalanced sex ratio was through the cyto-
plasm. Eggs fertilized by Y-bearing sperm
died early in the course of development.
Malogolowkin, Poulson and Wright (19591
have shown that the high female ratio and
embryonic male deaths may be transferred
from affected females to those which do not
normally show the condition through in-
jection of ooplasm from infected females.
Ooplasm of this same type is, w^hen injected
into males, suflficient to cause death to oc-
cur within 3 days. In the females a latent
period of 10 to 14 days after ooplasm injec-
tion was apparently necessary to establish
egg sensitization. Once established the con-
dition could be transmitted through the
female line for several generations. The
ooplasm injections were not as efficient in
establishing these lines as females found
naturally infected. Unisexual broods may

26

BIOLOGIC BASIS OF SEX

fail to appear, in some instances fail to
transmit the condition or produce inter-
mediate progeny ratios, as well as in some
instances to skip a generation. The infec-
tious agent was found to vary in different
species. Magni (1953, 1954) for D. bi-
fasciata found that temperature above nor-
mal tended to remove the cytoplasmic agent
making it ineffective. This result has a par-
allel to the action of temperature on certain
viruses, as for instance that involved in one
of the peach tree diseases. On the other
hand, Malogolowkin (1958) ior D . willistoni
found no temperature effect. Malogolowkin
further found that the cytoplasmic factor
was not independent of chromosomal genes
since some wildtype mutant strains induced
reversions to normal sex ratio. Recently
Poulson has established the complete corre-
lation of the high female progeny character-
istic with the presence of a Trepomena in
the fly's lymph. As with mouse typhoid vari-
ations, lethality was genotype dependent.

These cases have interest from more than
the sex ratio viewpoint. Cytoplasmically
inherited susceptibility of some strains of
D. 7nelanogaster to poisoning by carbon di-
oxide as studied by L'Heritier (1951, 1955)
depended on the presence of some cyto-
plasmic entity which passed through the egg
cytoplasm and also, but less efficiently, by
means of the male sex cells to the progeny.
The carbon dioxide susceptibility was trans-
mitted through injections of hemolymph or
transplantation of organs of susceptible
strains. The transmissible substance had a
further property of heat susceptibility. The
COo susceptibility differed from that of "sex
ratio" in being partially male transmitted.
It agrees with ''sex ratio" D. bifasciata in
being heat susceptible but differs from D.
willistoni "sex ratio" in this respect.

D. bifasciata may behave differently than
D. willistoni in that Rasmussen (1957) and
Moriwaki and Kitagawa (1957) both con-
ducted transplantation experiments with
negative results. However, it is jiossible that
these experiments may be affected by a
different incubation period for the injected
material as contrasted with that for D.
willistoni. A range of possibilities evidently
exist for extrinsic effects in sex ratio.

Carson (1956) found a female producing
strain of D. borealis Patterson, which car-

ried on for a period of 8 generations, pro-
duced 1327 females with no males. The
strain showed no chromosomal abnormali-
ties. It had 3 inversions. The females would
not produce young unless mated to males
from other strains having biparental in-
heritance. This requirement, together with
gene evidence, showing that the females
were of biparental origin, is against the fe-
male progenies being derived by thelytokous
reproduction.

F. MALE-INFLUENCED TYPE OF
FEMALE SEX RATIO

Sturtevant in 1925 described exj^eriments
with a stock of D. affinis in which a great
deficiency of sons was obtained from certain
males, regardless of the source of females
to which such males were mated. The few
males obtained from such matings were nor-
mal in behavior but some of the sons of
females from such anomalous cultures again
gave very few sons (]Morgan, Bridges and
Sturtevant, 1925).

Gershenson (1928) in a sampling of 19
females caught in nature found two that
were heterozygous for a factor causing
strong deviations toward females (96 per
cent to 4 per cent males) whereas the nor-
mal D. pseudoobscura ratio was nearly 1 to
1. The factor was localized in the X chromo-
some and was transmitted like an ordinary
sex-linked gene. Its effect was sex limited
as it was not manifest in either hetero-
zygous or homozygous females. It had no
effect on the development of zygotes al-
ready formed but strongly influenced the
mechanism of sex determination through the
almost total removal of the spermatozoa
with the Y chromosome from the fertiliza-
tion process. Egg counts showed the di-
vergent ratio toward females was not caused
by death of the male zygotes inasmuch as
there was no greater mortality from such
cultures than from controls giving 1 to 1 sex
ratios.

Sturtevant and Dobzhansky (1936)
showed that an identical or nearly identical
phenotypc to that obscM'ved in Drosophila
obscura was present in D. pseudoobscura.
The D. pseudoobscura carrying the factor
were scattered over rather wide geographi-
cal areas. Comparable types also were found
in two other species D. athabasca and D.

FOUNDATIONS FOR SEX

27

azteca. This genie "sex ratio"' sr lies in the
right limb of the X of races A and B of
D. pseudoobscura. Like the other cases
analyzed, males carrying sr have mostly
daughters and few sons regardless of the
genotypes of their mates. Structurally the
female sex ratio came through modification
of the development of the sex cells of the
male to give a majority of X-bearing sperm.
Cytologic study showed that in "sex ratio"
males the X underwent equational division
at each meiotic division, whereas the auto-
somes behaved normally. The Y chromo-
some lagged on the first spindle, remained
much condensed, and its spindle attachment
end was not attenuated. The Y chromosome
was not included in either telophase group
of the first meiotic division but was left be-
hind. It was sometimes noted in one of the
daughter cells where it formed a small nu-
cleus. Among 64 spermatocytes examined,
all had an X chromosome and none a Y
chromosome in their main nuclei. The mi-
cronuclei containing Y played no further
part in the division, became smaller and
exceedingly contracted. The final fate of the
micronuclci was uncertain, but there was no
indication that any spermatids died or were
abnormal.

G. HIGH MALE SEX RATIO OF GENETIC ORIGIN

In 1920, Thompson described a recessive
mutant in D. melanogaster which killed all
homozygous females but changed the male
phenotypes only slightly, the wings stand-
ing erect above the back. The locus of the
mutant was 38 in the X chromosome. This
mutant was a leader for a class in which the
genes affect only one sex but are innocuous
to the other.

Bobbed-lethal, a sex-linked gene found
i)y Bridges, is a gene of this class but one in
which the mechanism of protection to the
other sex is known. It kills homozygous fe-
males but does not kill the males because of
the wild type allele which the males have in
their Y chromosomes. The presence of this
bobbed-lethal in a population consequently
leads to male ratios higher than wild type.

A recessive gene in chromosome II, dis-
covered by Redfield (1926), caused the
early death of the majority of female zy-
gotes and led to a sex ratio of about 1 fe-
male to 5.5 males. The effect was trans-

mitted througli both males and females. The
missing females may have died largely in
the egg stage although disproportionate
losses also occurred in the larvae and pupae.
The maternal effect was attributed to an
influence exerted by the chromosome con-
stitution of the mother on the eggs before
they left the mother's body. Because of this
maternal lethal, the families from these
mothers have the normal number of sons
but few or no daughters.

A culture was observed by Gowen and
Nelson (1942) which yielded only male
progeny, 136 in all. Some of the male prog-
eny were able to transmit the male-produc-
ing characteristics to half of their daughters
without regard to the characteristics of the
mates to which they were bred. The in-
heritance was without phenotypic effect on
the males. When present in the hetero-
zygous condition in females, the female's
own phenotype gave no indication of the
gene's presence. The inheritance was sex
limited in that it affected only the eggs laid
by the mothers carrying it. In these eggs it
acted as a dominant lethal for the XX zy-
gotes. The gene was found located in the 3rd
chromosome between the marker genes for
hairy and for Dicheate at approximately 31.
The X eggs carrying the gene, Ne, died in the
egg stage at between 10 and 15 hours under
25°C. temperature. The gene was as ef-
fective in triploids as in diploids. One dose
of this gene in triploids caused them to have
only male progeny and male type intersexes.
The presence of this gene caused the elimi-
nation of any embryos with chromosomal
capacities for initiating and developing the
primary or secondary female sexual sys-
tems. Since the gene itself was heterozygous
in the female the meiotic divisions of the
eggs would cause the gene to pass into the
polar bodies as often as to remain in the
fertilization nucleus yet the lethal effects
are found in all eggs. The antagonism was
between the egg cytoplasm and the XX
fusion products. Supernumerary sperm of
the Y type were not sufficient to overcome
the lethal effects. An XY fertilization nu-
cleus was necessary for survival.

This case has parallel features with that
observed by Sturtevant (1956) for a 3rd
chromosome gene that destroys individuals
bearing the first chromosome recessive gene

28

BIOLOGIC BASIS OF SEX

for prune. The gene responsible was found
to be a dominant, prune killer, K-pn, which
was located in the end of the 3rd chromo-
some at 104.5. Prune killer was without
phenotypic effect so far as could be ob-
served except that it acted as a killer for
flies containing any known allele of prune.
It was equally effective in either males or
females, hemizygous or homozygous, for
prune. The larvae died in the 2nd instar
but no gross abnormalities were detected
in the dying larvae. This case has interest
for the Ne gene in that the killer gene oc-
cupies a locus in the 3rd chromosome al-
though removed by about 70 units from Ne,
and acts on a specific genotype, the prune
genotype, whereas Ne acts on the specific
XX genotype. The known base for action
of the prune killer is genetically much nar-
rower than that for the female killer, Ne,
in that prune killer acts on an allele within
a specific locus in the X chromosome, and
Ne acts on a type coming as a product of
the action of two whole sex chromosomes. It
is, of course, conceivable that when ulti-
mately traced each action may be dependent
upon specific changes of particular chemi-
cal syntheses. The action of these genes is
also of interest from another viewpoint.

In mice there is a phenotype caused by
the homozygous condition of a recessive
gene (Hollander and Gowen, 1959) which
acts on its own specific dominant allelic
type in its progeny so as to cause an in-
creased number of deaths between birth
and two weeks of age, as well as causing the
long bones to break and the joints to show
large swellings. The lethal nature of this
interaction is not a product of the mother's
milk nor does it show humoral effects such
as those observed with erythroblastosis in
the human. The interallelic interaction is
sex limited in that it is confined to the
mother and is without effect when the male
has the same genotype.

H. FEMALE-MALE SEX RATIO INTERACTIONS

A case in D. affinis involving the interac-
tion of a "female sex ratio" factor and an
autosomal "male sex ratio" factor has been
studied by Novitski (1947). Starting from
stock which had a genetic constitution for
the "sex ratio" X chromosome which ordi-
narily causes males carrying it to produce

only daughters, he was able to establish a
recessive gene in chromosome B in whose
presence only male offspring resulted. The
genetic constitution of the male was alone
important. High sterility accompanied the
"male sex ratio" males breeding perform-
ance. The "male sex ratio" parents yielded
95 fertile cultures having an average of 25
individuals per culture. Females appeared
in 10 of these cultures with an average of 3
per culture for those producing females. The
total sex ratio was 77 males per female. The
males were morphologically normal. They
carried an X chromosome of their mothers.
Cytologic observation of spermatogonial
metaphases of 3 progeny showed that the
Fi males may or may not have carried a Y
of their fathers. This agreed with the tend-
ency for sterility in these "male sex ratio"
males. The ventral receptacles of the fe-
males from 4 such sterile cultures when ex-
amined proved devoid of sperm. The occa-
sional female offspring of the "male sex
ratio" males had one X chromosome from
each parent. Females mated to "male sex
ratio" males show large numbers of sperm
in the ventral receptacles (7 out of 8 cases) ,
although such females had usually produced
no or very few offspring. The sperm, if
capable of fertilization, must have had a
lethal effect on the zygotes. The recessive
nature of the factor in chromosome B indi-
cated that its potential lethal effect origi-
nated during spermatogenesis rather than
at the time of fertilization. This lethal pe-
riod corresponds to that when the "female
sex ratio" factor in the X chromosome is
active.

The research on sex ratio in Drosophila
reviewed shows that through the interplay
of the sex chromosome-located and auto-
somal-located factors all types of sex ratios
from only females in the family to only
males in the family may be generated.

V. Sex Determination in Other Insects

Sciara, a fungus gnat, offers sex-deter-
mining mechanisms quite different from any
yet offered in Drosophila. Sciara is unique,
yet in its uniqueness, it illustrates basic
facts that were rediscovered in other spe-
cies only through the study of abnormal

FOUNDATIONS FOR SEX

29

forms some of which were created as in-
duced developmental, chromosomal, or hor-
monal abnormalities. The complexities re-
sponsible for the remarkable facts were
analyzed by Metz and his collaborators.
The following brief review is based upon
Metz's (1938) summarization of the chro-
mosome behavior problem to which the
reader is referred for further information.
Of Sciara species studied 12 out of 14 have
their basic chromosome groups composed of
three types of chromosomes: autosomes,
sex chromosomes, and "limited" chromo-
somes. Two species, S. ocellaris Comstock
and S. reynoldsi Metz, lack the "limited"
chromosomes. Two sets of autosomes and
three sex chromosomes are found in the
zygotes of all Sciara species at the comple-
tion of fertilization. The behavior of these
chromosomal types will be considered in
sequence.

The autosomes behave normally in so-
matic mitosis and in oogenesis. There is
cytologic and genetic evidence for synapsis,
crossing over, random segregation, and reg-
ular distribution of chromosomes and genes
in the female.

In spermatogenesis the story is quite dif-
ferent. The first maturation division is uni-
polar. Genetic evidence shows a complete,
selective segregation of the maternally de-
rived autosomes and sex chromosomes from
those of paternal origin. The paternal homo-
logues move away from the pole, are
extruded, and degenerate. The second
spermatocyte division is likewise unequal
and one of the products, a bud, degenerates.
Thus a spermatogonial cell passing through
meiosis gives rise to only one sperm. At the
second spermatocyte division all of the
chromosomes (maternal homologues) except
the sex chromosome undergo an equational
division but both halves of the sex chromo-
some enter into the same nucleus. This nu-
cleus becomes the sperm nucleus. The chro-
mosomes at the opposite pole form a bud
and degenerate. Fertilization of the Sciara
egg is normally monospermic not poly-
spermic.

The sex chromosomes XXX of the fer-
tilization nucleus are derived, two from the
sperm and one from the oocyte. Their des-
tiny depends on whether the egg they are

in develops into a male or a female imago.
In male early development, those nuclei
which are to become male soma lose the two
sex chromosomes XX, contributed by the
father's sperm, to give X + 2A soma. These
chromosomes fail to complete mitosis at
the 7th or 8th cleavage division and are left
to degenerate in the general cytoplasm when
there are no true cells in the soma region
and no membranes surround the nuclei.
Those embryos which are to become female
soma at the same cleavage cycle eliminate
but one paternal X chromosome. On a chro-
mosome basis the soma cells respectively
become X, plus two sets of autosomes, give
rise to males on differentiation, or become
XX + 2A and develop female organs (Du
Bois, 1932). The germ line nuclei for each
sex, on the other hand, remain unrestricted
in their development. They retain their
XXX constitutions until the first day of
larval life, or about 6 hours before the for-
mation of the left and right gonads (Berry,
1941), when they eliminate a single pater-
nally derived X. The X which is rejected or
makes its own exit is always one of two
sister chromosomes contributed by the
father. The process of loss is strikingly dif-
ferent from that noted in the soma-building
nuclei. Both cell walls and nuclear mem-
branes are present. The path of the X chro-
mosome is through the nuclear membrane
into the cytoplasm where degeneration
eventually takes place. The loss occurs at
a time when there is no mitotic activity and
the chromosomes are separated from the
cytoplasm by an intact nuclear membrane.
Some Sciara species are characterized by
females which produce only unisexual fam-
ilies — practically all females or practically
all males. Genetically, the sex of progeny is
accounted for by sex chromosomes, X' and
X, which are so designated because of their
physiologic properties. Mothers having only
female progeny are characterized by always
having the X' chromosome in all their cells,
X'X + 2A, whereas mothers whose progeny
are all males are XX + 2A (Moses and
Metz, 1928). The all female and all male
broods are observed in about equal num-
bers. The male has no influence on sex ratio.
Normally the progeny in all female broods
will be X'X + 2A or XX + 2A in soma

30

BIOLOGIC BASIS OF SEX

and germ lines, whereas the all male prog-
eny broods will be only X + 2A in the soma
and XX + 2A in the germ line.

Studies of Grouse (1943, 1960a, b) show
nondisjunction may occur and eggs entirely
lacking sex chromosomes or having double
the normal number may be produced. When
the nondisjunctional eggs lacking X chro-
mosomes are from X'X females and are fer-
tilized by sperm contributing two sister X
chromosomes, one X (as expected of X'X
mothers, not two as expected of XX cells) is
eliminated at the 7th or 8th cleavage to give
cells which develop into "exceptional" males
with XO + 2A soma and XX + 2A germ
line, all X chromosomes being of paternal
origin. The casting out of the single X chro-
mosome is consequently controlled by the
characteristics of the egg cytoplasm derived
from the X'X mother rather than by the
kind of chromosomes present in the nuclei
at the 7th or 8th cleavage stage of the em-
l)ryo.

Similarly when nondisjunctional eggs
having both XX chromosomes of the XX
mothers are fertilized by the normal sperm-
bearing XX chromosomes, the zygotes are
XXXX. At the 7th or 8th cleavage both
paternal X chromosomes are eliminated and
the resulting soma develops into an ''excep-
tional" female, XX + 2A, often with 3 X's
in her germ line. These observations con-
firm the earlier interpretations that the X'X
or XX constitution of the mother conditions
her eggs respectively to cast out 1 or 2
parental X chromosomes at the 7th or 8th
cleavage according to the cytoplasmic con-
stitution of the egg. Somatic sex develop-
ment is visually determined only after this
stage when the chromosomes of the somatic
cells take on the familiar aspect of either
XX + 2A for the females or X + 2A for the
males.

Although at present there are few exam-
ples in other species, this cytoplasmic con-
ditioning of early embryologic development
by the mother's genotyj)e may be a common
rather than a rare occurrence of develop-
ment. The expression of the events to which
this control may lead may be quite differ-
ent in different cases. The spiral of snail
shells was an early recognized case (Sturtc-
vant, 1923) . In Drosophila, a gene acting
through the mother's genome allows onlv

male embryos to survive (Gowen and Nel-
son, 1942), suggesting that this gene acts
through her cytoplasm in a manner com-
parable to that of the X' and X chromo-
somes of Sciara. In either case the geno-
type would not be considered a primary sex
factor, profound as the effect on sex may
ultimately be.

Sciara also has strains or species where
the progeny of single matings are of both
sexes, bisexual, yet the chromosomal elimi-
nation mechanisms remain as described
above save for the fact that they now op-
erate within broods rather than between
broods. Sex ratios for the families within
these bisexual strains are extremely varia-
ble, 1:0 or 0:1 ratios not being infrequent
and 1 : 1 ratios being the exception. An in-
stance of a bisexual strain arising from a
unisexual line has been analyzed by Reyn-
olds (1938). The females were found to be
3X in their germ line and produced 2X eggs
which developed into daughters and X eggs
which became sons. Loss of the extra X
from this line resulted in the line's return to
unisexual reproduction.

In general, bisexuality as contrasted with
unisexuality seems to be caused by differ-
ences in the X'X-XX mechanisms, either
specific for the mechanism, or induced by
modifying genes which cause different em-
bryos within the same progeny to cast out
sometimes one or sometimes two X chromo-
somes from all their somatic nuclei. This
casting out occurs uniformly for all nuclei
of a given embryo as is shown by the fact
that sex mosaics or gynandromorphs are as
seldom observed in bisexual as unisexual
strains.

Some species of Sciara have large chro-
mosomes which are in essence "limited" to
the germ line since they are lost from the
somatic nuclei at the 5th or 6th cleavage
division. Other species lack these chromo-
somes, yet agree with the rest in the be-
havior of their remaining chromosomes and
in sex determination. So far as is now
known the "limited" chromosomes seem
empty of inheritance factors influencing ei-
ther sex or unisexual vs. bisexual broods or
for that matter other characteristics (their
persistence seems to make this latter un-
likely I. Their presence does lead to a ques-
tion of the efficacv of desoxvribonucleic acid

FOUNDATIONS FOR SEX

31

(DNA) as the all inclusive agent in inherit-
ance for they are clearly DNA-positive.

These cytogenetic observations clear up
most of the events leading to sex differentia-
tion and the delayed time in embryologic
development when it takes place, but, as
INIetz points out, other puzzling questions
are raised. Observed chromosome extrusions
from cell nuclei are rarities today even
though discovered for Ascaris 60 years ago.
Why should this mechanism be so well de-
fined in Sciara? Why should the mode of
elimination be so accurately timed and yet
be different in method and time for the
soma and germ cell lines? Two of the three
X chromosomes in the fertilized egg are
sisters from the father and should be identi-
cal. What is special in the 7th or 8th cleav-
age that causes elimination in the soma
plasm but not in the germ plasm? Why
should the casting out of one of these same
paternal X's in the germ plasm of both
sexes be reserved for a much later stage in
cleavage and by a different procedure?

The observations of Grouse (1943, 1960a),
obtained from translocations and species
crosses, are critical in showing that the
chromosome first moA'ing in its entirety to
the first differentiating pole of the second
spermatocyte division is the X and is the
one taking part in the later postfertilization
sex chromosome elimination of the 7th or
8th cleavage. The X heterochromatic end of
the chromosome and not the centromere re-
gion seems to be responsible for the unique
behavior of this chromosome (Grouse,
1960b). Induced nondisjunctional eggs of
X'X (female-producing) mothers having as
a consequence no sex chromosomes, when
fertilized by the normal sperm bringing in
XX chromosomes, develop into embryos
eliminating one of these X's at the 7th or
8th cleavage, as expected of the eggs of
X'X mothers, and become "exceptional"
males with X + 2A soma and XX paternal
germ line. Imagoes of these embryos become
functional but partially sterile males trans-
mitting the paternal X, a reversal of the
normal condition. Similarly the nondis-
junctional eggs retaining both X's, XX
from male-producing mothers, and having
4 X's on fertilization eliminate the two pa-
ternal X's, as occurs normally in XXX em-
l)ryos, to become "exceptional" females with

2X or sometimes in species hybrids 3X
soma.

Phenotypic sex in Sciara on this evidence
depends on the adjustment of the X chro-
mosome numbers and their contained genes
with this adjustment regularly taking place,
not at fertilization as in most forms like
Drosophila, but at the 7th or 8th cell cleav-
age of embryologic development. The X' vs.
X chromosomes of the mother predispose
her haploid egg cytoplasm to time specific
chromosome elimination. The germ line cells
on the other hand may regularly have other
chromosome numbers than the soma or ex-
ceptionally tolerate other numbers as extra
X's (Grouse, 1943), or extra sets (Metz,
1959).

Triploids of 3X and 3A soma have not
been found in pure Sciara species. Salivary
gland cells of a triploid hybrid S. ocellaris x
aS. reynoldsi have two sets of S. ocellaris and
one of S. reynoldsi chromosomes (Metz,
1959). Ghromosome banding in 2X -I- 3A,
3X + 3A larval glands showed the »S. ocel-
laris homologues were heterozygous. The
3X + 3A type chromosomes were of typical
female appearance. The 2X:3A-type had
X chromosomes of typical male type, very
thick, pale and diffuse. The mosaic sali-
varies were a mixture of the two types of
cells. These results suggest that the 3X:3A
would be a typical female. The intersexes
2X -I- 3A conceivably have a range in male
and female organ development depending
on how the S. ocellaris and S. reynoldsi
chromosomes act. If the chromosomes are
equivalent, the gene expression would be
expected to be that of a 2X to 3A or a
phenotypic intersex. If they act separately
the S. ocellaris genes would be in balance for
female determination, whereas the S. reyn-
oldsi autosomal genes would have no X S.
reynoldsi genes to balance them and pre-
sumably would be overwhelmingly male in
effect. The phenotypic outcome would be
in doubt.

The maternally governed cleavage and
chromosomal sequence occurring before the
8th embryonic cell division, although re-
lated to the subsequent events leading to the
particular sex, is unique to Sciara and its
relatives. Phenotypic sex expression appar-
ently starts with the somatic nuclei having
either X + 2A or XX + 2A as their chro-

32

BIOLOGIC BASIS OF SEX

mosomal constitutions. Sciara is like Dro-
sophila and many other genera in its basic
sex-determining mechanism. There is a sex-
indifferent period when the maternal geno-
type may influence development through
cytoplasmic substances. This is followed by
active differentiation when the sex pheno-
types may be influenced by various agencies
but when passed becomes fixed for pheno-
typic sex. The indifferent period may repre-
sent a fairly large number of early cell gen-
erations as in amphibia or fish. Somatic
and germ cells are labile but are normally
susceptible to control only by forces de-
veloped within the organism as present
knowledge indicates for Sciara. In Sciara,
only the somatic cells conform in chromo-
some number to their expected phenotypes.
The germ cells show not only a regulated in-
stability of their chromosome number but
the number is different from that of the
supporting somatic cells. The fact that sex
mosaics in Sciara may develop both an
ovary and testis in the same animal (Metz,
1938; Grouse, 1960a, b) shows localized
phenotypic sex determination occurring
fairly late in development rather than the
generalized determination that would oc-
cur if sex were established at fertilization.

The germ line chromosomal differentia-
tion from that of the somatic tissue has
parallels in a number of fairly widely dis-
persed species. The significance of these
types to soma line determination and to the
reversibility or irreversibility of differen-
tial gene activation in differentiated cells
subsequent to embryogenesis has been con-
sidered by Beermann (1956). Although
these types are yet in the fringe areas of
our knowledge surrounding the general
paths taken in the evolution of sex deter-
mining mechanisms, they promise to have
more generality as clarification of gene ac-
tion becomes more exact.

B. APIS AND HABROBRACON

Hymenopteran interi:)retations of sex de-
termination have largely turned on studies
of the honey bee. Apis mellijera Linnaeus
and Habrohracon juglandis Ashmead. Dzi-
erzon early established that the drones of
honey bees come from unfertilized eggs,
whereas the females, queens, and workers
come from fertilized eggs so that sex differ-

entiation is marked by an N set of chromo-
somes for the male and 2N for the female.
This mechanism was satisfactory until it
was realized that on a balanced theory the
doubling of a set of chromosomes, if truly
identical, should not change the gene bal-
ance and consequently not the sex. Further
doubt was cast on the N-2N hypothesis for
sex determination by the discovery of dip-
loid males by Whiting and "Whiting (1925).
These difficulties were resolved by Whiting
(1933a, b) with the suggestion that the two
genomes in the female were not truly alike
but actually contained a sex locus linked
with the gene for fused which was occupied
by different sex alleles as Xa and Xb . In
this A'iew the heterozygous condition would
lead to the production of females, whereas in
the haploid or homozygous condition either
of these genes would react to produce males.
Difficulties with this hypothesis became evi-
dent in that the ratios of males to females in
certain crosses were significantly divergent
from those which were expected. Evidence
was collected and trials made to interpret
these difficulties (Whiting, 1943a, b) by as-
suming that on fertilization by Xa-bearing
sperm the polar body spindle would turn
so as to cast out the Xa chromosome and re-
tain the complementary Xb in the majority
of cases instead of in a random half. Snell
(1935) offered the hypothesis that there
could be several loci for sex genes, such that
the X locus might be a master locus, but
when this locus was in homozygous condi-
tion the sex would then be controlled by
genes in other loci in the genome. As pointed
out by Bridges (1939) this hypothesis would
satisfy that of sex gene balances postulated
in Drosophila and of change in emphasis on
loci for sex genes as observed by Winge
(1934) in Lebistes. However, the work of
Bostian (1939) called both hypotheses into
question. Consideration of these further re-
sults led to the multiple allele hypothesis of
complementary sex determination for
Habrobracon (Whiting, 1943a) taking a
similar form to that of the sterility relations
observed for a single locus in the white
clover of INIelilotus. Under this system the
sex locus would be occupied by multii)le al-
leles, any one of which or any combination
of identical alleles would be male-producing,
but any combination of two different al-

FOUNDATIONS FOR SEX

83

Iclcs would l»e female determining. By hav-
ing a sufficient number of these genes the
jiroduction of diploid males would be cur-
tailed to a point at which their various fre-
quencies could be explained. In support of
this liyi)othesis, multiple alleles to the num-
ber of at least 9 were found to exist for
thi.> single sex locus.

In principle at least, the honey bee could
have the Habrobracon scheme of sex de-
termination. Rothenbuhler (1958) has re-
cently collected the researches which test
this possibility. Tests of the multiple allele
hypothesis as applied to the honey bee were
made by Mackensen (1951, 1955) who in-
terpreted evidence for inviable progeny pro-
duced by mating of closely related individ-
uals as proof that this species as well as
Habrobracon juglandis follows the multiple
allelic system. The discovery of male tissue
of bii)arental origin in mosaic bees from re-
lated i)arents was considered as further evi-
dence for the multiple allelic theory of sex
determination (Rothenbuhler, 1957).

Most recent cytologic evidence supports
the concept that there are 16 chromosomes
in the gonadal cells of the male and 32 in
those of the female (Sanderson and Hall,
1948, 1951; Ris and Kerr, 1952; Hachinohe
and Onishi, 1952; Wolf, 1960». Hachinohe
and Onishi (1952) found 16 chromosomes
as characteristic of the meiosis in the drone.
Wolf observed a nucleus in both bud and
spermatocyte of the only maturation (equa-
tional) division.

The greatest progress has been made in
understanding the mechanisms of sex mosai-
cism in the Hymenoptera species. These
mosaics, although ordinarily of rather rare
occurrence, have a direct bearing on sex
determination and development. In Apis,
iiolyspermy furnishes the customary basis
for their formation. One sperm fertilizes
the haploid egg nucleus and another sperm,
which has entered the egg instead of degen-
erating as it ordinarily does, enters into
mitotic cleavage and eventually forms is-
lands of haploid cells of paternal origin
among the diploid cells derived from the
fertilized egg (Rothenbuhler, Gowen and
Park, 1952). Evidence showing that genetic
influences affect the sperm nucleus toward
stimulating its independent cleavage is
found to exist in Apis material (Rothen-

buhler, 1955, 1958). Tliousands of gynan-
dromorphs have been observed in Apis, all
but a small number of which have been pro-
duced in this manner. This method of initi-
ating sex mosaics also exists for Habrobra-
con (Whiting, 1943b) but is rather rare.
In Habrobracon the frequent mode has a
different origin. The gynandromorph is
formed from the cleavage products of the
normal fertilization of the egg nucleus com-
bined with those of a remaining nuclear
product of oogonial meiosis. Under these
conditions the female tissue is 2N of bi-
parental origin and the male tissue is N of
maternal origin (Whiting, 1935, 1943b).
This type is less frequent in Apis but one
specimen has been described by Mackensen
(1951).

A number of other ways in which sex
mosaics may occur are occasionally ex-
pressed in these species. Three different
kinds of male tissue have been observed in
individual honey bee mosaics produced by
doubly mated queens — haploid male tissue
from one father, haploid male tissue from
the other genetically different father, and
diploid male tissue of maternal-paternal
origin. In other cases, the diploid, biparental
tissue was female and associated with two
kinds of male tissue (Rothenbuhler, 1957,
1958) . Cases where the haploid portions of
the sex mosaics are of two different origins,
one paternal and the other maternal, while
the female portion is representative of the
fertilized egg, are known in Habrobracon
(Whiting, 1943b). Similarly, Taber (1955)
observed females which were mosaics for
two genetically different tissues and which
he accounted for as the result of binucleate
eggs fertilized by two sperm. Mosaic drones
of yet another type were observed by
Tucker (1958) as progeny of unmated
queens. They were interpreted as the cleav-
age products from two of the separate nu-
clei formed in meiosis. These cases represent
a number of the possible types that arise
through meiotic or cleavage disfunctions
under particular environmental or heredi-
tary conditions.

Tucker (1958) studied the method by
which impaternate workers were formed
from the eggs of unmated queens. For this
purpose he used genetic markers, red, char-
treuse, ivory, and cordovan. Observations

34

BIOLOGIC BASIS OF SEX

Avore nuult' i)ii 237 workers from hotcrozy-
gous mothers. For the chartreuse loeus. 12 to
20 per cent were homozygous, for x\w i\ory
locus 1.8 per cent, and (he cordovan locus
on lesser numbers per t-ent. An egg which
is destined to become an automictic worker,
a gynandromorph with somatic male tissue
or a mosaic male, is retained within the
queen for an unusually long time during
which meiosis is suspended in anaphase 1.
Normal reorientation of first division
spindle is i)ossibly inhibited by this aging
so that after the egg is laid meiosis II occurs
with two second division spindles on sejia-
rate axes as Goldschmidt conjectured for
rarely fertile rudimentary Drosophila
(19")7i. Two polar l)odies and two egg
prontich^i are formed. The polar bodies take
no fiu'ther part in develoinnent. In most of
the unusual eggs the two egg pronuclei
unite to form a diploid cleavage nucleus
which develops into a female. Rarely the
two egg pronuclei develop separately as two
haploid cleavage nuclei to form a mosaic
male. Two unlike haploid cleavage nuclei,
one descending from each of the two sec-
ondary oocytes after at least one cleavage
di\ision. unite to form a dijiloid cleavage
nucleus which develops together with the
remaintler of the haploid cleavage nuclei
to produce a gynandromorph with mosaic
male tissues. The male and female tissues
within these unusual gynandromor[ihs or
female types were identical with normal
drone or normal female tissues so were
probably haploid and diploid resiiectively.
Genetic segregation observed within the
mothers of automictic workers allows the
estimation of the distance between the locus
of the gene and its centromere. "With random
recombination and "central union" Tucker
estimates this distance for the chartreuse
locus to centromere as 28.8 units and for the
ivory to its centromere 3.6 units. Four lines
of bees of diverse origin all showed a low
percentage of automictic or gynandro-
morphic types produced from queens in each
line. Various chance environmental condi-
tions apparently influence the rate of pro-
duction of these types. However, there were
some females in two of the lines with higher
frequencies inciicating that innate factors
may have significant eft'ects on their fre-
quencies. Observations on Drosophila spe-

cies — I), porthenogeuetica (Stalker, 1956b),
I). ))ia)njabeirai (INIurdy and Carson, 1959)
and D. nielanogaster (Goldschmidt, 1957) —
strongly suppt)i-t this A-iew.

The [irohleni of si'x determination in Hal)-
rol)i'acon presently stands as a function of
multii)le alleles in one locus, the heterozy-
gotes being female and the azygotes and
homozygotes male. The occasionally diploid
males are regularly produced from fertilized
eggs in two allele crosses after inbreeding.
These diploid males are of low viability and
are nearly sterile. Their few daughters are
triploids, their sperm being dijiloid. Apis
probably follows the same scheme, as a
few cases of mosaics with diploid male tis-
sue are known and close inbreeding results
in a sufficient number of deaths in the egg
to account for diploid males which might be
formed. ^lormoniella (Whiting, 1958), how-
ever, shows that this scheme for sex determi-
nation does not hold for all Hymenoptera.
In this form diploid males may occur
through some form of mutation. They may
then develop from unfertilized eggs laid by
triploid females. In contrast to Habrobra-
con the dijiloid males are highly viable and
fertile. Their sperm are diploid and their
numerous daughters triploid. Virgin triploid
females produce 6 kinds of males, 3 haploid
and 3 dij^loid. Similarly ]\Ielittobia has still
a different and as yet unexplained form of
sex determination. Haploid eggs develop
into males. After mating many eggs are laid
which develop into nearly 97+ per cent fe-
males. The method of reproduction is close
inbreeding but no dijiloid males or "bad"
eggs are formed. The prol)lem of the sex
determining mechanism remains open
(Schmieder and Whitiuii, 1947).

The silkworm, Bomby.r mori, differs from
Drosophila, Lymantria and the species thus
far discussed in having a single region in the
^^' chromosome (Hasimoto, 1933; Tazima,
1941, 1952) occupied by a factor or factors
of high female potency. The strong female
potency has thus far been connnon to all
races. The chromosome patterns of the
sexes are like those of Abraxas and Lyman-
tria: males ZZ + 2A and females ZwV 2A.
The diploid chromosome number is 56 in
both sexes. Extensive, well executed studies

FOUNDATIONS FOR SEX

35

liave revealed no W chruniosome loci for
genes expressed as morphologic traits. From
radiation-treated material it has been pos-
sible to pick up a translocation of chromo-
some II to the W chromosome as well as a
cross-over from chromosome Z. This chro-
mosome together with tests of hypoploids
and hyperploids have materially aided in
understanding how the normal chromosome
complexes determine sex. The sex types re-
sulting from different chromosome arrange-
ments have been summarized by Yokoyama
1 1959) and are presented in Table 1.3.

Whenever the W chromosome was ab-
sent a male resulted. Extra Z or A chromo-
somes did not influence the result. Similarly
with a W chromosome in the fertilized egg
a female developed. Again extra Z or A
chromosomes did not influence the result.

A full Z chromosome was essential to sur-
vival. Hypoploids deficient for different
amounts of the Z chromosome in the pres-
ence of a normal W chromosome all died
without regard to the portion deleted. Hy-
perploids for the Z chromosome, on the other
hand, when accompanied with a W chromo-
some all lived and showed no abnormal sex-
ual cliaracteristics. Parthenogenesis led to
the pioduction of both sexes, although the
males were more numerous than the fe-
males. Diploidy was necessary for the em-
l)ryo to go beyond the blastoderm stage.
Triploid and tetraploid cells were often
found. High temperature treatments led to
merogony (Hasimoto, 1929, 1934). The ex-
ceptional males were homozygous for a sex-
linked recessive gene and were explained
by assuming that the egg nuclei were in-
activated by the high temperature and
the exceptional males developed from the
union of two sperm nuclei. This conclusion
was supported by cytologically observed
l)olyspermy (Kawaguchi, 1928) and by
cytologic observation of the union of two
sperm nuclei by Sato ( 1942 ) . Binucleate eggs
were also believed to occur, which when
fertilized by different sperm may each con-
struct half of the future body. This type of
mosaicism was influenced by heredity
iGoldschmidt and Katsuki, i927, 1928,
1931 ). Polar body fertilization was also be-
lieved to occur, one side of the embryo orig-
inating from the ordinary fertilized egg nu-
cleus and the other side from the union of

TABLE 1.:^
Sex in Bombyx iitori
(Summarized by T. Yokoyama, 1959.)

Sex

Chromosome Types and Numbers

W

z

A

Male

W
II.W.ZL

w
w

WW
WW

zz

zz

zzz

z

z

zz
zzz
zz
zz

AA

Male

Male

Female

Female

Female

Female

Female

Female

AAA
AAA

AA

AA

AAA

AAAA

AAA

AAAA

nuclei of two of the polar bodies. Similarly,
dispermic merogony was noted following
the formation of one part of the body from
the fertilization nucleus, the other part from
the union of two sperm nuclei, the result
being a gynandromorph or mosaic.

VI. Sex Determination in Dioecious
Plants

\. MELAXDRUM t LYCHNIS]

Over the last 20 years studies on several
species of dioecious plants have made not-
able advances in unclerstanding the mecha-
nisms by which sex is determined.

Melandrium album has been shown to
have the same chromosome arrangement as
Drosophila. The male has an X and Y plus
22 autosomes, whereas the female has XX
plus 22 autosomes. Sex-linked inheritance
is known for genes borne in the X chromo-
somes as well as for genes born in the Y
chromosome. The X and Y chromosomes
are larger than any of the autosomes with
the Y chromosome about 1.6 times that of
the X in the materials studied by Warmke
( 1946) . Separate male and female plants are
characteristic of the species. By use of
colchicine and other methods, Warmke and
Blakeslee (1939), Warmke (1946), and
Westergaard (1940) have made various
l^olyploid types from which they could de-
rive other new X, Y and A chromosome
combinations from which information was
obtained on the location of the sex deter-
mining elements. The Y chromosome carries
the male determining elements, the X chro-

36

BIOLOGIC BASIS OF SEX

mosome the female determining elements.
The guiding force of the elements in the Y
chromosome during development is suffi-
cient to override the female tendencies of
several X chromosomes. Data derived by
each of these investigators are shown in
Table 1.4.

From these data Warmke (1946) con-
cluded that the balances between the X and
the Y chromosomes essentially determined
sex with the autosomes of relatively little
importance. Where no Y chromosomes were
present but the numbers of X chromosomes
ranged from 1 to 5, only females were ob-
served, even though the autosomes varied in
number from two to four sets. When a Y
chromosome was present the individual was
of the male type unless the Y was balanced
by at least 3 X chromosomes when an occa-
sional hermaphroditic blossom was formed.

TABLE L4

Numhers of X, Y, chromosomes and A, autosome sets

and the sex of the various Melandrium plants

(Data from H. E. Warmke, 1946; and

M. Westergaard, 1953.)

Ratio

Chromosome

Warmlce

Westergaard

Constitution

X/A

4A 5X

Female

1.3

4A 4X

Female

1.0

Female

4A 3X

Female

0.8

4A 2X

Female

0.5

3A 3X

Female

1.0

Female

3A 2X

Female

0.7

Female

2A 3X

Female

1.5

2A 2X

Female

1.0

Female

Bisexual*

X/Y

4A 4X Y

4.0

Male

4A 3X Y

Malet

3.0

Male

4A 2X Y

Malet

2.0

Male

4A X Y

Male

1.0

3A 3X Y

Malet

3.0

3A 2X Y

Malet

2.0

Male

3A X Y

Male

1.0

2A 2X Y

Malet

2.0

2A X Y

Male

1.0

Male

4A 4X YY

Malet

2.0

4A 3X YY

Male

1.5

4A 2X YY

Male

1.0

Male

2A X YY

Male

0.5

* Occasional staminate but never carpellate
blossom.

t Occasional licrina])hr(i(lit ic blossom.

When 4 X chromosomes were present to-
gether with a Y, the plants were hermaphro-
ditic but occasionally had a male blossom.
Two Y chromosomes almost doubled the
male effect. Two Y chromosomes balanced
4 X chromosomes to give a majority of male
plants. Only an occasional plant showed an
hermaphroditic blossom. Autosomal sex
effects, if present, were only observed when
plants had 4 sets and 3 or 4 X chromosomes
balanced by a Y chromosome. Warmke used
the ratio of the numbers of X to Y chromo-
somes as a scale against which to measure
clianges from complete male to hermaphro-
ditic types. No mention is made of quanti-
tative measures of the sex character changes
with increasing X chromosome dosages.
This is of interest since in many forms
changes in chromosome balance are accom-
panied by changes of phenotype which are
unrelated to sex. That such phenotypic
changes do accompany changes in autosomal
balance in Melandrium are proven, how-
ever, by further observations of Warmke in
4 trisomic types coming from crosses of
triploids by diploids. Of 36 such trisomies
analyzed, 5 or 6 of them were of different
growth habits and morphologic types. These
differences did not affect the sex patterns
since all were females. Warmke and Blakes-
lee in 1940 observed an almost complete
array of chromosome types from 25 to 48 in
progeny derived from crosses of 3N x 3N,
4N X 3N, and 3N x 4N. Out of about 200
plants studied, only 4 were found to show
indications of hermaphroditism. These types
were 2XY and 3XY. As noted from the
table, even the euploid plants would occa-
sionally be expected to have an hermaphro-
ditic blossom. Of the 200 plants, all with a
Y (XY, 2XY, 3XY) were males and all
plants without the Y (2X, 3X, 4X) were
females. In an 8-year period up to 1946.
Warmke was able to observe only one male
trisomic. From these facts he concluded
that the autosomes are unimportant in the
sex determining mechanism utilized by this
species. In their crosses they were unsuc-
cessful in getting a 5XY plant, the point at
which the female factor influence of the X
chromosomes might be expected to nearly
equal or slightly surpass that of the single
Y. From the j^hysiologic side the obscrva-

FOUNDATIONS FOR SEX

37

tion of Strassburger in 1900, as quoted by
both Warmke and Westergaard, that the
fungus Ustilago violacea when it infects
Melandrium will cause diseased plants to
produce mature blossoms with well devel-
oped stamens (filled with fungus spores) as
well as fertile pistils, shows that these fe-
males have the potentialities of both male
and female development. The case suggests
that sex hormone-like substances may be
produced by the fungus which acting on the
developing Melandrium sex structures
cause sex reversions. Should this be true,
Melandrium cells would have a parallel
with those of fish where sex hormones incor-
|)orated in the developing organism in suf-
ficient quantities can cause the soma to
develop a phenotype opposite to that ex-
pected of their chromosomal type. For other
aspects see Burn's chapter and Young's
chapter on hormones.

Westergaard's studies (1940) with Euro-
pean strains of Melandrium were in prog-
ress at the same time as those of Warmke
and Blakeslee. In their broad aspects both
sets of data are concordant in showing the
l^rimary role of elements found within the
Y chromosome in determining the male sex
and of elements in the X chromosomes for
the female sex. Examination of Table 1.4
shows that the strain used by Westergaard
has a Y chromosome containing elements of
greater male sex potentialities than the
strain used by Warmke.

A similar difference appeared in the sex
potencies of the autosomes of European
strains. Instead of obtaining essentially
only male and female plants in crosses in-
volving aneuploid types, Westergaard ob-
tained from 3N females (3A + 3X) x 3N
males (3A + 2XY) 10 plants which were
more or less hermaphroditic, 21 females, and
15 males. Studies of the offspring of these
hermaphrodites through several generations
showed that their sex expression required
effects by both the X chromosomes and cer-
tain autosome combinations which under
special conditions counterbalanced the fe-
male suppressor in the Y chromosome. In-
creasing the X chromosomes from 1 to 4
increased the hermaphrodites from to 100
per cent in the presence of a Y chromosome.
However, in euploids these types would be

all males. The significance of the autosomes
is further shown by the fact that among 205
aneuploid 3XY plants, 72 were males and
133 were hermaphrodites.

As pointed out for Drosophila, quanti-
tative studies on the effects of sex chromo-
somes and autosomes in Melandrium are
handicapped by not having a suitable scale
for the evaluation of the different sex types.
The data presented by Westergaard and by
Warmke make this difficulty become partic-
ularly evident. In the interest of quantizing
the X, Y, and A chromosome on sex the
author has assigned a value of 1 for the
male type, 3 for the female type, and 2
when the types are said to be hermaphro-
ditic. When the types are mixed, as for ex-
ample, in the data of Warmke where he says
a particular type is male with a few blos-
soms, the type is assigned a value of 1.05 or
1.10, depending on the numbers of these
blossoms. His bisexual type which comes
as a consequence of Y, 4X and 4A chromo-
some arrangement is given a value of 2, al-
though possibly the value should be some-
what higher as it may well be that the fully
bisexuals are further along in the scale to-
ward female development than the hermaph-
roditic types. The data are treated on the
additive scale both as between chromosomal
types and within chromosomal type. This is
apparently unfair if we examine the work
of Westergaard in which it looks as if par-
ticular autosomes rather than autosomes in
general make a contribution to sex determi-
nation. The results, when these methods are
used, are as follows:

Westergaard in Tables 1 to 5 of his 1948
paper gives information on sex types with
a determination of the numbers of their
different kinds of chromosomes. Analysis
of these data by least square methods shows
that the sex type may be predicted from
the equation

Sex type = - 1.37 Y + 0.10 X

+ 0.01 A + 2.34

This equation fits the data fairly well con-
sidering that the correlation between the
variables and the sex type is 0.87. This
analysis again shows that the Y chromo-
some has a strong effect toward maleness.
The X chromosomes are next in importance

38

BIOLOGIC BASIS OF SEX

with an effect of each X only about 1/13
that of the Y and in the direction of female-
ness, the autosomes have one tenth the effect
of the X chromosomes but they too have
a composite effect toward femaleness. It is
to be remembered that the Y chromosome
variation is limited to 2 chromosomes
whereas the X chromosomes may total 4,
and the autosomes may range from 22 up
to 42, so that the total effect of the auto-
somes is definitely more than their single
effects. These data are for aneuploids. Ex-
amining Westergaard's data for 1953 for the
euploids and assigning the value of 1.5 for
the type observed when there was one Y
chromosome, four X, and four sets of auto-
somes, we have the following equation:

Sex value = -1.29 Y + 0.10 X - 0.01

(autosome sets) + 2.53

In these data, as distinct from those above,
the autosomes are treated as sets of auto-
somes since they are direct multiples of each
other so the value of the individual auto-
some is but 1/11 that given in the equation.

This equation shows no pronounced dif-
ference from that when the aneuploids were
utilized. The Y chromosomes have slightly
less effect toward the male side. The X
chromosomes have practically identical ef-
fects but there has been a shift in direction
of the autosomal effects on sex, although
the value is small. The constants are sub-
ject to fairly large variations arising
through chance.

In Table 10 of Westergaard's 1948 paper
he presented data on the chromosome con-
stitution and sex in the aneuploids which
carried a Y chromosome. These data are of
particular interest as the plants are counted
for the i)roi)ortions of those which are male
to those which are hermaphroditic. The
plants with a Y chromosome plus an X are
all males. Those which have either one, two
or three Y chi'omosomes balanced by two
X chromosomes have 89 per cent males.
The plants with three X chromosomes and
one oi- two Y's have 36 p(T cent males and
those which have one or two ^■ chroiiiosoincs
and four X chromosomes have no males.
The woik iiiv()l\-cd in getting these data is,
of course, large indeed and is definitely
handicapped by the diiiicuHies in obtaining

certain types. Thus the XXYYY and the
4X + 2Y types depend only on one plant.
There are eight observations but the fitting
of the data for the X and Y constitutions
eliminates three degrees of freedom from
that number so that statistically the obser-
vations are few. The data do have the ad-
vantage that the sex differences can be
measured on an independent quantitative
scale. The equation coming from the results
is:

Percentage of males = 13.2 Y - 36.4 X

+ 134.4

These results show that the Y chromosome
increases the proportion of males and the
X chromosome increases the proportion of
intersexes. The data are not comparable
with those analyzed earlier as these data
are describing simply the ratio between the
males and intersexes, instead of the rela-
tions betw'een the males, intersexes, and fe-
males. The equation fits the observations
rather well, as indicated by the fact that
the correlation between the X's and Y's
and the percentages of males is 0.98, but
there are large uncertainties.

As a contrast to these data we have those
presented by Warmke (1946) in his Tables
2 and 3. These data give the numbers of X
and Y chromosomes found within the plants
but not the numbers of autosomes, the auto-
somes being considered as 2, 3, and 4 gen-
omes. Analyzing these data in the same
manner as those of Westergaard's Table 1,
we find that the

Sex type = -1.05 Y + 0.22 X -0.04 A

-f2.25

As indicated for Westergaard's data, the A
effect is now in terms of the diploid type
e(iualing 2, the trijiloid 3, and the tetra-
l)loid 4.

The Y chromosome has a i)ronounccd ef-
fect toward maleness, the effect l)eing some-
W'hat less in Warmke's data than that of
Westergaard's. The X chromosomes on the
other hand, have nearly twice the female
infhience in Warmke's data that they do in
the phiiits grown by Westergaartl. A differ-
ence in sign exists for the effect of the A
rhi'oinosnnie genom(\s as well as a difference
in [\\v (|uantitati\'e effect. Th(^ values for

FOUNDATIONS FOR SEX

39

both sets of data are small and toward the
male side. As Westergaard points out, the
strains used by these investigators are of
different geographic origins. The evolution-
ary history of the two strains may have a
bearing on the lesser Y and greater X ef-
fects on the sex of the American types.
Chromosome changes seem to have occurred
in the strains before the studies of Warmke
and Westergaard and will be discussed.

The location of the sex determiners has
been studied by both investigators utilizing
techniques by which the Y chromosome be-
comes broken at different places. These
breakages may occur naturally and at fairly
high rates in individuals which are Y + 2X
+ 2A. These facts suggest that the break-
age of the Y chromosome occurs in meiosis
since the breakage comes in selfed individ-
uals of highly inbred stocks where heterozy-
gosity is not to be expected, in the Y
chromosome which has no homologue thus
does not synapse, and in the second meiotic
<li vision. Plants showing these initial break-
ages seem to have the same constitution in
all of the somatic cells of different organs as
well as in the germ cells, again pointing to
meiosis as the time of breakage.

In AVarmke, Davidson and LeClerc's
(1945) material, the normal offspring which
resulted from selfing 2A, 2XY plants, were
2A, 2XY (male hermaphrodites), 2A, 2X
(female), 2A, XY (male), and 2A, X2Y
(suiiermales) , and in addition two abnormal
hermaphroditic classes. These were: (1) a
type in which the female structures were
liighly developed, essentially as well de-
veloped as in 2A, 2X females and with nor-
mal stamens; and (2) a type in which there
was a complete failure of stamen develop-
ment shortly after meiosis. Cytologic ex-
amination showed these types to be as-
sociated with the Y chromosome breaks.
The first type occurred when the homol-
ogous (synaptic) arm of the Y was de-
ficient. The deficiencies ranged in size
from a short terminal loss to one which
seemed to include the entire, or nearly en-
tire, homologous arm. It was of importance
that the degree of abnormality was not pro-
l)ortional to the length of the deficiency.
Once a small terminal segment was lost the
change in sex type occurred and larger

losses seemed to have no more pronounced
effect. Chromosomes of this type showed
complete asynapsis of the Y chromosome,
indicating that its synaptic element compar-
able with the X was lost. Loss of as much as
one-fifth to one-fourth of the arm prevented
synapsis. These results seemed to indicate
that the Y chromosomes had lost elements
which acted as suppressors to the female de-
velopment.

The second type observed by Warmke
was associated with a break in the differ-
ential arm of the Y chromosome. A
small terminal loss of this differential
arm was sufficient to cause male devel-
opment to be arrested and sterility to re-
sult. The plants that had lost as little as
one-fourth the differential arm were there-
fore male sterile and were also indistin-
guishable from plants that had lost both
arms. The altered X chromosomes retained
their centromeres and were carried through
mitotic growth divisions to every cell of the
])lant. Only in rare cases, and with very
small fragments, was there evidence that
somatic loss may have occurred. From these
results Warmke concluded that the Y chro-
mosome contained at least three gene com-
plexes which operated in the development
of maleness. First there was one near the
centromere and present in the smallest frag-
ment of the Y chromosome which initiated
male development. The stamens developed
but only just past meiosis. The second fac-
tor was found near the end of the differ-
ential arm of the Y chromosome and in-
fiuenced complete male development. When
the entire differential arm of the Y chromo-
some was present, full male development
resulted. The third element appeared on the
terminal fourth of the homologous arm of
the Y chromosome and suppressed female
development. Whether this was in the pair-
ing segment or close to it was uncertain. In
individuals with entire Y chromosomes,
plus two X chromosomes, the female struc-
tures were underdeveloped with only a
small percentage of the blossoms capable
of setting capsules with seed. When the
homologous arm was deficient the female
development was complete and every blos-
som produced seed-filled capsules, again
supporting the conclusion that this part of

40

BIOLOGIC BASIS OF SEX

the Y chromosome acted as a positive sup-
pressor of female determining regions in the
X chromosomes. That these regions were
strictly located and the effect not due to the
quantities of the Y chromosome which may
have been present or absent was indicated
by crosses of the different types of frag-
ments. In these crosses the resulting total Y
chromatin may have been considerably
greater in size than a normal single Y chro-
mosome, yet the observed changes in sex
characteristics of the specific regions lost
were present. The results indicated that the
sex elements located in the Y chromosomes
were qualitatively distinct from one another
in their action and cannot be substituted
for another in quantitati^'e fashion. The
causes of the differential changes in each
case could rest on single gene differences or
possibly a closely associated nest of such
genes.

Westergaard's studies showed that his
plants behaved differently in some particu-
lars from those of Warmke. His search for
the male determining elements in the Y
chromosome of his Danish plants empha-
sized these differences. He was able (1946a,
b) to divide the Y chromosome into four
different regions, a region corresponding to
the X chromosome in which there was
synapsis and three regions containing vari-
ous sex initiating elements. When these ele-
ments were compared with those of
Warmke's it was found that they were com-
parable in action but differed in their order
within the Y chromosomes. The Danish
l)lants had the female suppressor region in
the end opposite the pairing region at the
extremity of the differential region. The ele-
ments initiating anther development were
found near the centromere, but toward the
pairing region. The element which com-
l)leted development was found near the
l)airing region or homologous section. When
compared with Warmke's results the i^osi-
tion of the different elements in the Wester-
gaard material was the reverse of that in
the American material. Westergaard ex-
plained this difference on the assumption of
a centric inversion in the Y chromosome re-
sulting in the change of positions. As he
suggested, it would certainly be interesting
to know the geographic distribution of

these two types and what would result in
progeny of crosses between them.

Westergaard (1948) summarized his
views on the sex-determining mechanism in
Alelandrium as follows. A trigger mecha-
nism is built up by an absolute linkage be-
tween the female suppressor region and at
least two blocks of essential male genes in
the Y chromosome. This trigger mechanism
operates with the X chromosomes and auto-
somes in which the X chromosomes have fe-
male potencies and certainly the autosomes
contribute to them. The action of these two
types can only be demonstrated through
the breaking of the normal balance by pol-
yploidy or aneuploidy. As yet, the female
])otentials of the X chromosome and certain
of the autosomes have not been analyzed to
the extent of showing whether they contain
major female sex genes or flocks of modify-
ing genes. W^estergaard favors the hy-
pothesis of modifying genes.

B. RUMEX

Rumex studies on sex determination took
their origin as with most dioecious plants
in chromosome examinations of the different
members of this genus (Kihara and Ono,
1923; Kihara, 1925; Ono, 1930, 1935;
Kihara and Yamamoto, 1935). These
studies showed that the species Rumex ace-
tosa had the normal diploid female comple-
ment of 14 chromosomes consisting of a pair
of X chromosomes and 6 pairs of autosomes
and the male had one X chromosome op-
posed by 2 Y chromosomes with 6 pairs of
autosomes. Occasionally intersexes found in
nature had 2 X chromosomes, 2 Y chromo-
somes, and 3 sets of autosomes. Sex deter-
mination in this earlier data, as summarized
in Tables 3 and 4 of Yamamoto's excellent
1938 paper, when analyzed by us utilizing
the metliods of least squares, showed tiiat
X chromosomes had large female effects in
both euploids and aneuploids whereas the
net effects of the Y chromosomes and auto-
some sets were but one-sixth to one-tenth
as great and in the male direction. Since
that time great advances have been made
through the studies of Yamamoto (1938),
Love (1944), Smith (1955), Love and
Sarkar ( 1956) , and Love (1957). As Bridges
foi'ctold in 1939, "It may now be suggested

FOUNDATIONS FOR SEX

41

from genetic studies that the occurrence of
translocations is responsible for (a) the
production of multiple elements from orig-
inally single elements, for (b) the frequent
change in type of sex chromosome configur-
ation in closely related forms, such as in the
various species of Rumex and Humulus, and
for (c) the associations and non-random
segregation of compound elements." The
]n-edictions have been borne out in the com-
l)lex chromosomal and genetic systems ob-
served in some of the more recent studies.

Polyploids and trisomies of R. acetosa
were studied, particularly by Ono (1935)
and Yamamoto (1938). Yamamoto identi-
fied each chromosome found in each sex
type. He showed that the 6 pairs of auto-
somes were not ecjually balanced toward the
promoting of the male sex. The chromo-
somes called ai , a4 , and ae , had net effects
toward the males, whereas chromosome
pairs denoted by ao and as had net effects
toward female determination. Different bal-
ances of the different chromosome types and
pairs lead to the production of types named
after those of Bridges, supermales, males,
intersexes, females, triploids, and superfe-
males.

In his studies of euploid types, Yamamoto
set up ratios similar to those used by
Bridges in Drosophila, except that he gave
the X chromosome a weight of 100 and
each set of autosomes a weight of 60. Like
Bridges he considered Y chromosome empty
of sex genes. By using these weights he was
able to arrange the sex types in a consistent
series in which the so-called supermales had
an index of 0.56, the males 0.83, the inter-
sexes 1.11 to 1.43, females 1.67, and super-
females 2.50. As in Drosophila the assigning
of these different values was handicapped by
the lack of any really quantitative measure
of the sex evaluations.

If Yamamoto's carefully tabulated data
are assigned 1 for male, 3 for female, 2 for
intersex, 3.5 for superfemale, and 0.5 for
supermale and then analyzed by least
squares for the effects of the different chro-
mosomes on sex, the resulting equation is

Sex value = 1.96 + 1.09 X - 0.1 8Yi

- O.27Y2 - 0.28ai , + 0.06ao - O.OSa.,

- 0.23a4 + 0.12a.5 - 0.23a6 .

The X chromosomes contribute a strong
female influence and each Y a less effective
male influence. The autosomes ai , Sn and
ae are somewhat more potent toward the
male type than the Y chromosomes. Chro-
mosomes &2 , Siz , and as have their sex genes
almost in balance. As may be noted, this
form of quantitative analysis leads to con-
clusions in agreement with those of Yama-
moto.

In another section of the genus, Rumex
paucifolms, Love and Sarkar (1956) have
analyzed a tetraploid type with 28 chromo-
somes. The sex chromosomes were suggested
as of the XXXX and XXXY types, the
male being heterogametic. The X chromo-
somes were the longest whereas the Y was
the smallest chromosome in the comple-
ment. They concluded that the mechanism
of sex determination in this species is de-
pendent on the Y chromosome's having
strongly epistatic male determinants. This
conclusion was based on the fact that the
species is dioecious and the belief that the
plants are polyploids so that the sex mecha-
nism must be based on strong male deter-
minants in the Y chromosomes. The strength
of these male determinants is suggested to
be less than to allow the production of
dioecious hexaploids, inasmuch as the tetra-
ploid included not only true females and
males, but also a low frequency of androgy-
nous individuals (Love, 1957).

In another group of species classified by
Love (1957) in the subgenus Acetosella
there were 5 species: 2 diploid, 1 tetraploid,
1 hexaploid, and 1 octoploid. The diploid
species have the XX and XY arrangement.
The natural tetraploid, R. tenuijolius, shows
about the same degree of pairing at meiosis
as do hybrids between it and experimentally
produced panautotetraploids of R. angio-
carpus. The natural tetraploids show 4 X's
or 3X + Y for the females and males. Hexa-
ploids derived by alloploidy from the diploid
R. angiocarpus and the tetraploid R. tenui-
folius have 6 X chromosomes for the female
and 5X + Y for the male. Similarly the
octoploid R. graminifolius is an autotetra-
ploid oi R. tenuifolius with 8 X chromosomes
in the female and 7X + Y in the male. Only
in this stage do slightly intersexual individ-
uals occur as 2N = 57 or 58 chromosomes

42

BIOLOGIC BASIS OF SEX

instead of 56. At least one of the extra
chromosomes is an X. From these observa-
tions Love (1957) concluded "detailed stud-
ies and comparisons of the sex chromosomes
and their pairing in natural and experi-
mentally produced polyploids lead to the
conclusion that the sex mechanism in this
group must be based on the evolution of a
strong male determinant in the Y chromo-
some, of much the same kind as in Melan-
drium, but stronger."

Within this one genus, Rumex, species are
present which seem to have the male de-
termining elements (a) located in the auto-
somes as in Drosophila, and (b) in the Y
chromosomes as in man. When contrasted
with the female-determining elements of the
X chromosome these male elements seem
to vary in their sex-determining capacity
in the different species.

C. SPINACIA

Spinach is dioecious l)ut the X and Y
•chromosomes are cytologically indistin-
guishable from each other in at least some
species. Spinacia oleracea has 6 chromo-
some pairs. Recent work on sex-determining
mechanisms for this species has been con-
ducted by Bemis and Wilson ( 1953) , Janick
and Stevenson (1955a, b), Dressier (1958),
and Janick, Mahoney and Pfahler (1959).
Dressier has indicated that each pair of the
6 chromosomes can be identified by different
morphology although most other investi-
gators have been unable to make these
separations within their own material. He
assigns the role of sex differentiation to
the chromosome pair having the largest
size. The Y chromosome bears a satellite,
whereas the X chromosome does not. Janick
and Ellis (1959) located the sex chromo-
some pair through the use of the six primary
trisomies each of which is differentiated
morphologically. These trisomies have been
obtained as progeny from triploid pistillate
XXX bred to diploid staminate XY. Five of
the crosses between staminate plants of the
six trisomies mated with pistillate diploids
gave the one male to one female sex ratio
indicative of independence of the sex com-
plex from the particular trisomies. The sixth
cross utilizing the reflex trisomic gave a one
male to two female ratio indicative of the
sex complex being within the chromosome

pair which in triploid condition showed the
reflex type. Each of the morphologic triso-
mies was associated with one of the six
chromosomes. The chromosome associated
with reflex trisomic, the sex chromosome, is
the longest chromosome and is characterized
by submedian centromere. In the somatic
cells Janick and Ellis were unable to ob-
serve obvious heteromorphism in this
chromosome pair. These results, although
not agreeing in detail with those of Zoschke
(1956) and of Dressier (1958) confirmed
the existence of races which differ with re-
spect to morphology of the chromosomes
containing the XY factors. Janick and
Stevenson (1955b) considered that the
monoecious character did not depend on
unaltered balance between the X and Y
factors but seemed to be caused by an allele
as well as by other modifying genes. They
found that in polyploidy the sex expression
in spinach indicated that a single Y factor
was male-determining even when opposed
by three doses of the X. In their results only
the XX, XXX, and XXXX formed pistil-
late flowers, whereas the XY, YY, XXY,
XXXY, and XXYY were of the staminate
type. Extra doses of the Y may have fur-
ther effects as illustrated by the fact that
YY plants do not produce seed, whereas
sometimes the XY staminate progenies ob-
tained from selfing staminate plants do,
indicating that staminate plants come to
their fullest expression with the YY geno-
type. Chromosome recovery in the progeny
from crosses of 2N females mated to 3N
(XYY and XXY) males revealed that the
functional gametes from staminate triploids
were not confined to N and N 4- 1 types
(Janick, IVIahoney and Pfahler, 1959). The
progeny produced contained 49 per cent
diploids, 18 per cent trisomies, 0.5 per cent
14 chromosomes, 1.2 per cent 16 chromo-
somes, 11 per cent 17 chromosomes, 19 per
cent triploids, and 0.8 per cent 19 chromo-
somes. The reciprocal cross in which the
female was of the 3N type and the male
2N gave distinctly higher ratios of ancu-
ploids. Of the progeny, 28 per cent were 12
chromosomes, 36 per cent were 13, 7 per
cent wTre 14, 0.9 per cent were 15, 2.8 per
cent were 16, 13 per cent were 17, 13 per
cent were 18, or 60 per cent of the total
progeny were aneuploids, whereas in the

FOUNDATIONS FOR SEX

43

reciprocal cross the value was 31. Aside
from some indication of preferential X-YY
segregation in the triploid staminate parent
when mated to the 2N female, the data fit
expectations for the segregations of the Y
chromosomes rather well. Staminate flowers
were unaffected by such environmental
variables as temperature and day length,
whereas the monoecious and some pistillate
types tended to increase in maleness at the
high temperature of 80°F. and short day
length (Janick, 1955 L

D. ASPARAGUS

Asparagus officinalis plants are ordinarily
staminate or pistillate with occasional rudi-
mentary organs of the opposite sex appear-
ing in both the staminate and pistillate
flowers. The staminate and pistillate plants
ordinarily represent approximately equal
numbers. The rudimentary organs of the
male sex sometimes develop and form seed.
Rick and Hanna (1943) have showed that
when such seeds were planted, 155 males to
43 females were produced. The data sug-
gested that maleness depends on a dominant
gene with the homozygous and heterozygous
types indistinguishable. This conclusion was
supported by Sneep (1953). The data fur-
ther showed that the proportion of stami-
nate plants producing seeds was apparently
influenced by both heredity and environ-
ment, in that seed production in these stam-
inate plants was enhanced in some inbred
lines but at the same time showed rather
wide variability from plant to plant.

E. HUMULUS

Humulus sex chromosomes have been
identified by Jacobsen (1957) in two spe-
cies, H. lupidus and H. japonicus. H. lupulus
has a complement of 20 chromosomes in
mitosis. The X chromosome is identifiable
and separable from the Y chromosome in
both mitosis and meiosis. The X chromo-
some is longer than the Y but is of medium
size compared with the autosomes. H. ja-
ponicus has 17 chromosomes in the male
and 16 chromosomes in the female at mi-
tosis. There are two Y chromosomes Yi and
Y2 and an X in the male in both mitotic
and meiotic divisions. The female has two
X chromosomes. In both species the sex
chromosomes in prophase and prometaphase

are differentiated to show the position and
extent of the homologous and differential
segments. The Y chromosomes are highly
heterochromatic. Based on Ono's 1940 work
on triploids derived from crosses of diploids
and colchicine induced tetraploids in H. ja-
poni'cMs, Westergaard (1958j concludes that
the preliminary evidence suggests that sex
determination in H. japonicus follows the
arrangement of Drosophila in that the X
chromosomes are female determining and
the autosomes carry the male inheritance.
Sex differentiation in other sex dimorphic
higher plants follows patterns like those
represented by one or another of the plant
species discussed above.

VII. Mating Types

Species without morphologic sex dif-
ferences, which occur as unicellular and as
haploid forms, often show differences in
their behavioral relations to each other. The
types may be alternate. Blakeslee (1904)
first called attention to these reactions in
heterothallic fungi in which the opposite
mating types were so indistinguishable mor-
phologically that opposite types were desig-
nated as + and — . Preliminary criteria, to
be met in assigning the + and — types,
were: (1) the individuals studied should be
shown to be in fact sexually dimorphic and
not merely hermaphrodites or sex inter-
grades; (2) tests should include a large num-
ber of races in order to show that the dif-
ferences are truly related to behavioral
differences in reproduction and are not sec-
ondary characters peculiar to a race; and
(3) the strengths of the reactions should be
graded in order to correlate them with any
sex differences that might be observed (Sa-
tina and Blakeslee, 1925). From more than
a quarter century of study, Blakeslee and his
group working on some 2000 races included
in 30 species of 15 genera came to the belief
that over this large group of heterothallic
mucors strict sexual dimorphism was the
rule. Similar systems have extended the con-
cept far beyond this group of fungi.

In fungi Raper (1960) emphasizes the
role of gene differences in the control of sex.
Genetic mutations have furnished evidence
to show that future sexual capacity follows
segregation of genetic factors at meiosis.
These factors impose changes in differentia-

44

BIOLOGIC BASIS OF SEX

tion Avhich affect type compatibilities. The
work of Esser and Straub (1958) on 21 ir-
radiation mutant strains is cited. Of the 21
strains, 18 were sterile when grown alone.
These mutants could be divided into quali-
tatively different groups in terms of the
stages at which sexual isolation was
achieved. Three strains developed only vege-
tative mycelia; 3, ascognia but no proto-
perithecia; 6, few to many protoperithecia
but no perithecia; 6, sterile perithecia; and
3, fertile perithecia with nondischarging asci.
Four mutant types were observed more than
once. The results suggested single factor
changes each affecting a single essential fac-
tor product significant to further sexual de-
velopment. Restoration occurred by pairing
with wild type alleles in heterokaryons de-
rived from hyphal fusions between the two
defective strains. Fertility is sometimes
noted in pairing of self-fertile strains ap-
parently mediated through the cytoplasm.
Diffusible agents, hormones and surface
agents may all play a part in fertility. In
some features the results in fungi suggest the
wide range in fertility phenotypes of gene
origin so frecjuently isolated in Drosophila
experiments. Together with much other ma-
terial they furnish further evidence for a
multigenic basis for sex-controlled charac-
teristics.

Physiologic differentiation of individuals
of a species into diverse mating types was
notably extended by Sonneborn's (1937)
discovery that in Paramecium aurelia two
classes of individuals exist. Members of dif-
ferent classes unite for conjugation; mem-
bers of the same classes do not. Information
on these types has accumulated rapidly dur-
ing the last few years (Jennings, 1939; Son-
neborn, 1947, 1949, 1957). As a rule P. au-
relia has two and only two interbreeding
mating types in a variety. In general mating
types from different varieties do not react
to each other. Death or low viability follows
in a few cleavages for some of the rarely
interbreeding mating types. In P. bursa n'a
Jennings and his group discovered six vari-
eties each of them comprising a set of mat-
ing types. A system of at least four mating
types is known for varieties I, III and VI;
eight are found in variety II; IV has two
and V is represented by but one. Simihir

mating type systems are found in other
genera, i.e., Euplotes (Kimball, 1939) .

Mating systems have been demonstrated
for some species of bacteria. A strain in
order to show conjugation followed by
transfer of separate genetic entities requires
at least two different types of individuals.
Ravin (1960) has recently reviewed this
subject for bacteria. The F+ and the F"
cell types when intermixed will conjugate
and show detectable rates of interchange for
genetic materials in tests of subsequent
progenies. The F~ and F~ when mixed do
not show recombination. The F+ and F +
when mixed may show low rates of inter-
change which have been suggested as due to
the presence of physiologic F~ variants
sometimes found in genetically F+ isola-
tions. The postulated F+ and F~ factors
suggest relations similar to those of some of
the genes in higher animals or plants. The F+
factor has a property that may set it apart
from these genes. In the presence of F +
cells, F~ cells adopt the F+ mating type.
The reaction is suggestive of that taking
place within the male sex in Bonellia as
observed by Baltzer.

In the absence of visible morphologic dif-
ferences which enforce exchanges and re-
combinations of genetic materials, the phys-
iologic or submicromorphologic conditions
found in either haploid or diploid unicellular
organisms accomplish the same objectives
by establishing mating types. AVhether these
differences are really comparable with sex
and sex determination is still an open ques-
tion. There may be some ground for think-
ing that there are precursors which assist in
the develoiiment of such systems.

Vin. Environmental Modifications
of Sex

A. AMPHIBIA

Amphibian sex chromosomes, Amby-
stoma, Siredon, Pleurodeles, Triton, Tri-
turus,Rana, and Xenopus as found in nature
are ZZ for the males and WZ for the fe-
males. Sex reversal of the normal sex pheno-
types has been particularly successful in
this class of animals and under a variety of
conditions (for further information see
Chapter l)y Burns). The haploid chromo-
sonu" nuiuhers foi' the males of these various

FOUNDATIONS FOR SEX

45

genera are 12, 14, and 18, with no sex
chromosomal differentiation. Types having
haploid, diploid, triploid, tetraploid, penta-
ploid, hexaploid, heptaploid, and various
aneuploid chromosome numbers have been
observed under experimental conditions
(Humphrey and Fankhauser, 1956; Fank-
hauser and Humphrey, 1959). Humphrey
( 1945) induced sex reversal by grafting tes-
ticular jirimordia on to embryonic genetic
female gonads. He showed that males hav-
ing the genotype WW were viable, as were
females having the same constitution. The
somatic cells could be modified to either sex
phenotype whereas the germ cells retained
their genotypic constitutions as observed
later under natural and hormone control as
in fish.

The improvement of estrogenic hormones
facilitated further studies on this problem.
Under natural conditions the sex genes are
effective sex determiners as normally they
guide the developing organism into one or
the other of the definite sex phenotypes.
When the embryonic forms of newts or
toads were exposed to relatively high con-
centrations of sex hormone-like substances
of the other sex, growth and development
were guided toward that sex rather than
toward that expected from the chromosomal
constitution of the somatic cells (Gallien,
1954, 1955, 1956; Chang and Witschi, 1955,
1956). When the male genotype was con-
verted to the female phenotype through this
treatment and was later bred to a normal
male, the progenies as expected on the basis
of the chromosomal constitution were all
normal males. When the female WZ was
sex-reversed to the male phenotype and then
bred to a normal female the progenies were
of three types, from a chromosome stand-
point ZZ, WZ, and WW, the ratio being
1 male to 3 females. The derived WW type
was then of female constitution showing
that this chromosome carries genes which
normally guide the organism to the female
phenotyjie. WW individuals had an ade-
quate gene content for normal development
of either sex in this class of organisms. The
observations on sex reversal in amphibia
were reminiscent of the experimental anal-
yses of sex differentiation by Baltzer on
Bonellia (1935) in which he quoted Har-
rison (1933) as saying, "A score of different

factors may be involved and their effects
most intricably interwoven. In order to
resolve this tangle we have to inquire un-
der as great a variety of experimental con-
ditions as is possible to impose. Success will
be assured by the implicity, precision, and
completeness of our descriptions rather than
by a specious facility in ascribing causes to
particular events." Bonellia showed the
way for synthetic hormone in that contact
of the embryonic form with the female was
sufficient to direct development into the
male type.

Sex determination in fish as in Amphibia
seems to be of such a nature that the results
of hormone treatments are revealed more
clearly than in Drosophila or mammals.
Winge (1922) found 46 chromosomes in
both male and female guppies. Lebistes
reticulatus has 22 pairs of autosomes and 2
X chromosomes in the female and 22 pairs
of autosomes and an XY chromosome set in
the male. Normally the Y chromosome is
found only in the male and is transmitted to
only male progeny. The genetic factors con-
tained in the differential segment of this
chromosome are not found in the females,
but are transmitted to all the young of the
male sex. The pairing segment on the other
hand crosses over with the X. Whether the

Y chromosome itself also is sex deciding is
uncertain. As Winge said in 1922, "experi-
ences gained from the Drosophilia re-
searches have proved that one must indeed
take care not to state anything certain on
this subject."

A fairly large number of genes had been
demonstrated in the X, Y, and autosomes of
this species by 1934. Genes which showed
X linkage showed crossing over to the Y
chromosome and vice versa (Winge, 1934;
Winge and Ditlevsen, 1948) . The amount of
crossing over showed some variation de-
pending on what genes were present in the
two chromosomes. A locus was found in the

Y chromosome containing a set of allelo-
morphic genes which could not cross over
to the X chromosome. These alleles were
completely linked with a male-determining
clement found in this Y chromosome and
located near its end. The pattern of inheri-
tance for chromosome sex determination

46

BIOLOGIC BASIS OF SEX

was apparently XX for the female and
XY for the male. Events observed in selec-
tion experiments for sex completely altered
this behavior. A pair of autosomes took over
the role of the sex chromosomes in the ex-
perimentally produced race. The males as
well as the females were found to have the
XX chromosome arrangement. In effect the
X chromosomes became autosomes and the
X-linked genes were transmitted auto-
somally. Pure males and pure females were
generally obtained in the progeny for this
race, but sex differentiation was so weak
that the sex percentage was subject to wide
variations betweeen broods and the ex-
pected 50 per cent males to females was ob-
served only during the spring months.
Winge interpreted these results as indicating
that there were both masculine and feminine
elements present in the autosomes as well
as in the XY chromosomes which con-
tributed to the determination of sex. Selec-
tion sorts out different proportions of these
elements and a change occurs in the mecha-
nism forming the sex phenotype. XY fe-
males were produced which when crossed
to XY male segregates gave as expected 3
males to 1 female. On this basis the YY
individuals were found. The YY males on
crossing with normal females produced only
male offspring just as previously the XX
males when crossed to normal females gave
only female offspring.

Changes of a similar type have been
found in nature. Gordon (1946, 1947) ob-
served wild stocks of the platyfish, Platy-
poecilus maculatus, from Mexico which
were XX for the female and XY for the
male. Platyfish from rivers in British Hon-
duras on the other hand were WZ for the fe-
male and ZZ for the male. The W and Y
chromosomes have many common char-
acteristics. Breeding data on domesticated
platyfish uncovered similar exceptional
chromosomal types with corresponding dif-
ferences in the sex transmission. Brcidci"
(1942) observed an exceptional WZ male
which when mated to a normal female WZ
had 51 daughters and 13 sons in the prog-
eny. The ratio is such as to indicate that the
females were a mixture of WW and WZ
genotypes. That this was probable was in-
dicated by the work of Bellamy and Qucal
(1951). Again an exceptional WZ male in

niatings with WZ females had female prog-
enies of two types, WW and WZ, the males
being ZZ. The WW females were proved by
mating to normal males, ZZ, and obtaining
progenies which were entirely females. The
results indicated that the WW females were
less fertile and so would soon be replaced in.
nature by the WZ type. Aida (1921, 1936)
located genes in the X and the Y chromo-
somes of another genus of fishes, the med-
aka, Oryzias latipes, and studied the effects
of these chromosomes on sex determination
and sex reversal. The chromosome number
for the diploid was apparently 48 with no
prominent morphologic difference between
the X and Y chromosomes. The males were
XY and the females XX. By selection for
high male ratios, lines were established in
which the male offspring far exceeded those
which were female. Females having the
genotype XY were isolated which on cross-
ing to normal males gave the ratio of 3'
males to 1 female. One-third of the male off-
spring were of the YY constitution so that
as with the platyfish this type was viable.
In interpreting these results the primary
sexual characteristics are held to be de-
termined by respective genes distributed
throughout the autosomes and set into ac-
tivity by stimulating genes. The female
genes were held to require greater stimu-
lation than the male genes to be active and
produce their jihcnotypes. Sex was viewed
as determined by the differences in quan-
tity of the stimulating genes. Between
these two quantities a threshold was postu-
lated above which the female and below
which the male genes were stimulated. Sex
I'eversal and differences in sex ratios among
the offspring of these fish from sex reversed
males were explained as due to fluctuations
of the stimulating jjower or potency of the
X chromosome.

Yamamoto (1953, 1959a, b) showed tlie
l)ossibility of reversing the phenotypic sex
by incorporating sex hormone-like sub-
stances in the diets of the developing young.
Functional sex reversal of the male geno-
types XY to those having female pheno-
tyi)es was accomplished by introducing
estrone or stilbestrol into the diet for ft
to 10 weeks to the extent of 50 /i.g. per gm.
diet from 1-day-old fry to those 11 to 16
mm. in length. Mating these estrone sex:

FOUNDATIONS FOR SEX

47

reversed mothers XY, to normal males XY,
resulted in progenies containing 1 female to
2.2 to 2.4 males where the expected theoretic
relation would be 1 female to 3 males. The
male progenies were submitted to test
crosses. It was shown that the YY individ-
uals were, as expected, males. With the re-
moval of hormone feeding the normal sex-
determining mechanism reestablished itself
as XX for the female and XY for the male in
the following generation. The hormone feed-
ing had apparently, through its excess fe-
male-stimulating growth capacities, caused
the somatic tract of the fish to develop
throughout as a female l)ut did not in any
way influence the fundamental genetic con-
stitution of the cells. In at least one case
(1957j, Yamamoto has shown that true in-
tersexes can be produced having the geno-
type XY. The secondary sex characters were
intermediate between both sexes. The gonads
became ovotestes, testicular elements in the
anterior and ovarian components in the pos-
terior region. By use of the same technique,
but substituting methyl testosterone as the
hormonal additive to the diet at the begin-
ning of the indifferent gonad stage and con-
tinuing through sex differentiation, it has
been possible where quantities of 50 ;ag. per
gm. diet were fed in the diet to cause both
genetic sexes XX or XY to differentiate into
males with rudimentary testes which even-
tually become neuters on becoming full-
grown fish. Intermediate dosages resulted
in XX individuals becoming males and
in 3 cases intersexes. These phenotypic
males of the XX type became fertile,
])roducing spermatozoa which on fer-
tilization of eggs of normal females gave
all female progeny. Again the effect of
the sex hormone was temporary in that
only the treated generations showed the sex
reversal, their progeny returning to the
customary XX female and XY male types.
Sex reversals have been accomplished on
fish that were themselves progeny of sex-
reversed parents. Genetic analyses showed
that the sex-reversed males were all XY
genotypes rather than YY genotypes. This
was accounted for through the low viability
of the males of the YY genotype.

These observations are similar to those
observed in some of the Amphibia and are
also of interest in connection with the regu-

lar sex mechanisms which have developed
for Sciara or for those of occasional abnor-
mal types as those observed in Drosophila
and man. These cases make it evident that
phenotypic sex may be derived in a quite
different manner from that adduced by the
sex promoting genes carried through the
germ line, even though the germ line may
be nourished by the products of the pheno-
typic somatic cells. The time in development
when the presence of hormone in excess of
and external to the organism's own gene
initiated sex organizers, if it is to be ef-
fective, would seem to be important. The
brief embryonic stage before the determi-
native changes in sex primordia have oc-
curred, the neutral stage or stage of bisexual
potentiality, is likely to be most influenced
by external agencies that redirect sex dif-
ferentiation. Developraentally speaking,
this period is rather short for the primary
sex organs, although possibly longer in
terms of some secondary sex characters such
as the breasts in man.

IX. Sex and Parthenogenesis in Birds

Sex in birds follows the general ZW -|-
2A for the female and ZZ + 2A for
the male; sex-linked genes follow this
pattern. Although breeding results in gen-
eral follow expected orthodox lines, the
birds are subject to much mosaic variation
in their color patterns as well as significant
sterility relations in species hybrids. Sec-
torial mosaics are prominent. Some of these
can be accounted for by nondisjunction,
polyspermy, binucleate eggs fertilized by
difi"erent sperm, development of super-
numerary sperm and other known genetic
means, whereas others still lack adequate
information. Gynandromorphs have been
described, but are subject to question in
view of the plumage characteristics observed
in mosaics and because of the fact that the
sex hormones are such that striking plumage
differentiations may occur if for any reason,
for instance disease, the hormone-producing
organs are removed from the birds. In
ordinary fowl the ovary produces hormones
which suppress male plumage, whereas in
the seabright male a gene controlled change
in the testis has taken over this functional
attribute. Mottling and flecking are also
common, particularly in the plumage pat-

48

BIOLOGIC BASIS OF SEX

terns although seldom in the structural ele-
ments of the body. This may be interpreted
in a variety of ways, including mutation
and various types of chromosomal inter-
change in either somatic cells or those in-
corporated into the germ cell tract. The
permanence of feather type differentiation
in grafts has been demonstrated by Dan-
forth (1932) and others. Conditions for sex
variability have certainly been demon-
strated as present in the bird (Hollander,
1944). Crosses between species of birds
have led to sterile hybrids and to rather
extensive discussions of sex reversal in the
development of such forms. In their hybrids
betweeen pigeons and ring doves. Cole and
Hollander (1950j presented a summary of
their evidence as well as a review of this
literature. Female hybrids rarely resulted
from pigeon sires mated to dove females
but were readily produced in the reciprocal
crosses. Surviving female hybrids produced
no eggs. Male hybrids produced abundant
sperm but varied greatly in the proportion
of sperm which seemed normal. In the back-
crosses to pigeons practically no fertility
was shown, but in back-crosses to doves
over 2 per cent gave fertile eggs and 9 of
these eggs hatched. All back-cross specimens
were males and have proven sterile, but tes-
tes and semen were not examined. With mi-
nor exceptions the expression of some 20 mu-
tant genes studied was not different from
that of the pure species. Recessives were not
expressed in the hybrids whereas most of
the dominants gave their customary pheno-
types. Sex-linked mutants were transmitted
as expected for each sex. This was particu-
larly important as the result of the inherit-
ance of these sex-linked characteristics
showed that sex reversal did not occur.

Eggs from virgin hens are known occa-
sionally to undergo some development of the
germinal discs. In dark Cornish chickens
Poole and Olsen (1958) observed that some
parthenogenetic development was present
in 57 per cent of the eggs from their total
flock. Factors stimulatory to parthenogene-
sis were noticeably higher in hens of some
strains than in those of others. Birds within
the flock differed in jmrthenogenetic rates
from 100 per cent to 43 per cent. After incu-
bation for a 9- to 10-day period, 3.9 per cent
of the A strain, 0.7 per cent of the B strain,

0.4 per cent of the C strain showed some
parthenogenetic development. In the ex-
perience of these observers these birds
showed more development than is found
generally in the domestic fowl. Yao and Ol-
sen (1955) showed that, in 95 per cent of
the instances when parthenogenetic de-
velopment was encountered, the growth con-
sisted solely of extra embryonic membranes.
In the remaining 5 per cent growth was
more advanced, ranging from the presence
of blood and blood vessels only, to the pres-
ence of well formed embryos in others. The
cells were diploid and were able to repro-
duce themselves by mitotic division. When
work on the turkey was begun in 1952, 17
per cent of the eggs began a limited cleav-
age on being incubated. Two embryos at-
tained the size of normal 3-day embryos.
In 1958, among 7269 eggs, 15 per cent were
found to contain blood of embryonic origin.
Embryos of various ages were encountered
in 9 per cent. Fifteen poults of partheno-
genetic origin were hatched in the 1958 sea-
son. No multinucleated or polyploid cells
were found in these turkeys. These poults,
as with others, have always been males.
Survival of parthenogenetic types generally
ranges from a few hours to several days.

In ex]:>eriments with fowl pox it has been
shown that the number of eggs developing
parthenogenetically increases considerably
following vaccination. The factors leading
to parthenogenesis are considered to be the
genetic characteristics of the strain of birds
and the presence of an activating agent or
agents in the blood stream of the hens.

The parthenogenetic forms are of particu-
lar interest to the problem of sex determina-
tion. The females should l)e producing two
types of oocytes Z -I- A and W -h A of
which presumably the Z + A alone survive
since the embryos capable of being sexed
are all males. The embryos are also diploids.
The 2Z -I- 2A could be derived from a fusion
of the Z -H A polav body nuclei as noted
earlier or possibly chromosome doubling
coming later in the early cleavage. A ge-
netic element seems partially to control the
parthenogenetic process. Chromosome dou-
bling would lead to cells with identical pairs
of chromosomes. The gene would be homo-
zygous. Inbreeding of poultry leads to a con-
tinning and rai^id loss in the vialiility of

FOUNDATIONS FOR SEX

49

most strains of chickens. A greater loss
would be expected for truly homozygous
chickens or poults as birds are known for
the large numbers of sublethal genes they
carry. In fact, it is surprising that any sur-
vive to the adult stage.

The doubling of the W and A type would
result in individuals lacking the Z chromo-
some. From what was observed in Amphibia
and fish the WW + 2A, individual if it
survived, would be expected to be female.
Since this type has not as yet been detected
it may be inferred that it is inviable be-
cause of loss of certain essential genes in the
Z chromosome.

X. Sex Determination in Mammals

A. GOAT HERMAPHRODITES

Goat hermaphroditism as reported by As-
dell (1936), Eaton (1943, 1945) and Kondo
(1952, 1955a, b) is of particular interest
when comjjared with human hermaphrodit-
ism as observed by Overzier (1955) and of
testicular feminization as reported by Ja-
cobs, Baikie, Court Brown, Forrest, Roy,
Stewart and Lennox (1959) and others.
In each species the phenotypic range in sex-
ual development extended from nearly per-
fect female to nearly perfect male, with the
most frequent class as an intermediate. Ex-
ternal appearance of each was partially cor-
related with internal structure. When inter-
nal female structures as the INIiillerian ducts
were present, the external appearance was
more female-like. When the male structures
AVolffian ducts were developed, the external
api^earance was more male-like. The pres-
ence of the dual systems within certain of
these hermaphroditic types indicates, as in
Drosophila, that there is independence of
development of each system without a so-
called turning point calling for differentia-
tion of the female sex followed by that of
the male sex or vice versa.

In goats the hermaphroditic types were
traced to the action of a recessive autosomal
gene (Eaton, 1945; Kondo, 1952, 1955a, b).
This gene apparently acts only on the fe-
male zygote. In homozygous condition the
eml)ryos bearing them develop simultane-
ously toward the male as well as toward the
female types. This development resembles
closely that of the Hr gene in Drosophila,

because, although Hr is dominant and the
one in goats is recessive, they both operate
only on the female type and both tend to
develop jointly both male and female sys-
tems in sexual development.

One jarring note comes in relating the
cytologic basis for sex determination in
goats with that for the intersexes. The sex
ratios for the different crosses clearly place
the hermaphrodites as genetic females ex-
pected to have the XX chromosome con-
stitution. The XX constitution would then
also agree with that found for human
hermaphrodites as discussed later in this
paper. Makino (1950) has shown for one
case of the intersexual goat that its sex
chromosomes were of the male type. Ma-
kino's excellent studies with other species
made this observation of particular signifi-
cance as it was contrary to the other mor-
phologic and genetic evidence on these
hermaphrodites. The implications were fully
realized by Makino when the cytologic ob-
servations were made so that as far as pos-
sible the observations should be critical on
this point. However, there are several
sources of cell variation that suggest the
desirability of further checks. The chromo-
some number of the goat is large, normal
mitoses rarely appear in the gonads of the
intersexes, and the chromosomes of the
goat's spermatogenesis are so small as to
make difficult details of structure or identi-
fication. Some of the difficulties possibly
could be avoided by making tissue cultures
and determining the somatic chromosome
numbers of their cells.

Kondo (1955b) has shown that under the
breeding conditions of Japan when the sire
was heterozygous, the percentage of inter-
sexes actually approached the expected
value 7.3 per cent. When the sires were
homozygous recessive individual matings
showed 14.6 per cent hermaphrodites as was
expected. Continued mating of homozygotes
should show 25 per cent of the total kids
hermaphrodites, or the equivalent of 50 per
cent of the female progeny.

Hermaphroditism in goats has a further
advantage in that the locus is apparently
linked closely to the horned or polled condi-
tion. The horned condition, in consequence,
becomes a valuable indicator marking the
presence of the hermaphroditic factor in the

50

BIOLOGIC BASIS OF SEX

otherwise indistinguishable male types.
With these characteristics the goat types
have remarkable advantages over other
species for the solution of problems of
hermaphroditism.

The gene for goat hermaphroditism has
even more interest when it is contrasted
with that of another gene, tra, discovered
by Sturtevant (1945). Tra is recessive wdth
no distinguishable heterozygous effect. In
the homozygous state it converts the zygotic
female into a form with completely male
genitalia and internal reproductive tract
with no evidence of the female sexual repro-
ductive system. The gene effects in Dro-
sophila are more extreme than those in
goats but are concordant in showing that
there are loci in the autosomes which may
be occupied by recessive genes having direct
effects on phenotypic development of the
genotypic female. This evidence indicates
the significance of these genes rather than
the happenstance of their being in the
autosome, X or Y chromosome.

B. SEX IN THE MOUSE

The mouse has the XY + 38 A chromo-
somal arrangement for the males and XX +
38 A for the females. Similar karyotype pat-
terns have been reviewed for some Amphibia
and fish. Other Amphibia and fish may have
their karyotypes reversed as both forms are
found in nature or observed in breeding
studies. Similar reversals may be made ex-
perimentally in the phenotypes even though
the genotypes remain unaltered. Birds show
the sex differentiating arrangement of ZW
for the females and ZZ for the males. Par-
thenogenesis seems to lead to males of ZZ
type in domestic fowl and turkeys. In an
evolutionary sense the mammals could have
originated from and perpetuated either of
the major karyotype sex arrangements.

Mice and men are alike in that the X has
female-determining properties and the Y
male potencies. How much part the genes in
the autosomes have in sex develojoment is
not yet clear. Welshons and Russell (1959)
have shown that mice of the presumed X()
constitution are females and arc fertile.
They have 39 as the modal number of
chromosomes found in their bone marrow
cells, wiiereas the genetically proven XX
types have 40 cln'omosomcs. X ('hroinosoinc-

linked genes' behavior substantiate the
chromosomal constitutions of XO and XX
as females and XY as males.

These results are further supported by the
breeding behavior of the X-linked recessive
gene, scurfy (Russell, Russell and Gower,.
1959). This gene is lethal to the hemizygous
males before breeding. The genetics of the
scurfy females have been analyzed by trans-
planting the ovaries to normal recipient fe-
males and obtaining offspring from them. In
the scurfy stock the XO type occurred as 0.9
per cent of the progeny. The YO progeny
w^ere not identified and probably die pre-
maturely. Nondisjunction of the X and Y
chromosomes in the males could result in
sperm carrying neither X nor Y chromo-
somes. These sperm on fertilization of the
X egg would give an XO + 2A type individ-
ual. Because the result is a female, this
would support the Y chromosome as of male
potency. The mouse arrangement may then
be expected to be like Melandrium in which
a well worked out series of types is known.

Sex ratio in mice is strain dependent over
what has thus far proven to be a 10 to 15
per cent range. Weir (1958) has shown that
for two strains of mice established by select-
ing for low and high pH, the sex ratio figures
were 33 and 53 per cent for artificially in-
seminated mice and 41 and 52 per cent for
natural matings of these respective strains.
The differential pH values for the bloods of
the low line were 7.498 ± 0.006 and for the
high line 7.557 ± 0.007 as of the sixth
generation of selection. The parents with
the more alkaline bloods tended to have
greater percentages of males in their prog-
enies. These results direct attention to the
genotype dependent phenotypic factor
which may be of some importance for
variations in sex ratios.

C. SEX AND STERILITY IN THE CAT

The tortoiseshell male cat has long inter-
ested geneticists because it has seemed that
by theory it should not be. However, nature
has wonderful ways of circumventing best
laid hypotheses, sometimes when they are
fals(\ sometimes when they have not been
probed dee])ly enough. The yellow gene for
coat color in cats is sex-linked. This gene
operates on an autosomal background of
^(■lu's for black oi' tabby. Tlu^ females may

FOUNDATIONS FOR SEX

51

be phenotypically orange as the double dose,
0/0, covers up the effects of the other coat
color genes; or tortoiseshell, 0/+; or black
or tabby, +/+. The males may be orange,
'0/, black or tabby, +/, and the type unex-
pected tortoise. The tortoiseshell males are
timid, keep away from other males, and are
generally sterile. Testes are of much reduced
size and solid consistency. Exceptionally,
tortoiseshell males may mate and offspring
presumed from the matings may be born.
Active study of these males commenced as
early as 1904. Komai (1952) has offered a
unified hypothesis for their origin. Komai
and Ishihara (1956) have contributed added
information and a review of the literature
to which the reader is referred.

The cat has 38 chromosomes including an
X-Y pair for the males. The tortoise males
agree in having this arrangement (Ishihara,
1956) , the X being 3 or 4 times the length of
the Y in all cytologic preparations from
Japanese cats. Komai (1952) visualizes the
cat X chromosomes as composed of a pairing
segment containing the kinetochore and gene
loci among which is that for the orange gene
and a differential segment, not found in the
Y chromosome, containing the factor-com-
plex for femaleness. The Y chromosome is
visualized as having a segment containing
the kinetochore and capable of pairing with
the X chromosome. This segment may cross
over with the X so that it may acquire the
locus for orange or its wild type. The Y
chromosome is viewed as containing two
differential segments. The one carrying the
factor complex for maleness is located to
correspond with the X differential segment
carrying the female sex factor. The second Y
differential segment is at the other end of the
chromosome and contains the male fertility
complex. The tortoiseshell sterile males are
interpreted as caused by a Y chromosome
crossing over with the X chromosome to in-
corporate the male segment and the gene
in the resulting Y chromosome but with the
loss of the male fertility segment. The gam-
ete carrying this modified Y fertilizing an
egg with a normal X chromosome containing
the wild type instead of the gene develops
into the sterile tortoiseshell male. The data
show that the probability of these events oc-
curring is small. Komai records as reliable
■65 tortoiseshell male cats where the inci-

dence of the O gene in the whole population
of Japanese cats is 25 to 40 per cent. Of the
65, 3 were apparently fertile. These cases
and the few others found in the literature are
regarded as caused by those rare occasions
when the Y chromosome incorporates the
gene but retains the male fertility complex
as might occur in double crossing over. The
hypothesized factor locations and crossing
over arrangements also may explain the un-
expected black females which are known to
occur in some matings. Although not men-
tioned, black males and orange males show-
ing the same sterility features as the sterile
tortoiseshell males should also be found in
the cat poi)ulation. If found they would
further strengthen the hypotheses.

It is difficult to understand why, even with
its low initial frequency, the fertile tortoise-
shell male would not establish itself in the
Japanese cat population, inasmuch as they
are so admired and sought after by all the
people if any tortoiseshell males became as
fertile as the tortoiseshell male "lucifer"
(Bamber and Herdman, 1932) known to
have sired 56 kittens.

Ishihara's work (1956) seems to close the
door on another attractive hypothesis to ex-
plain the origin of these unexpected cat
types. Tortoiseshell male reproductive or-
gans include small, firm testes showing re-
duced spermatogonial development. To-
gether with the interaction of the gene
with the wild type allele they suggest the
human types XXY + 2A which may arise
from nondisjunction. However, the chromo-
some type is shown to be XY -f- 2A = 38
which is fatal to this hypothesis.

It is of interest that Komai in 1952 postu-
lated the male complex and fertility factors
in the Y chromosome of a mammal. The case
has a further parallel in the plant Melan-
drium in that the work of both Westergaard
(1946) and Warmke (1946) indicated the
Y chromosomes of this plant to contain such
factor complexes although in differing ar-
rangements.

D. DEVIATE SEX TYPES IN
CATTLE AND SWINE

As a caution in the mushrooming of cyto-
logic interjiretations of sex development, at-
tention may be directed to the freemartin
types known particularly from the work of

52

BIOLOGIC BASIS OF SEX

Keller and Tandler (1916), Lillie (1917),
and the researches stimulated by their ob-
servations on cattle twins. The freemartin
in cattle develops in the same uterus with its
twin male. The blood circulations anasto-
mose so that blood and the products it con-
tains are common to both fetuses during de-
velopment. The development of the female
twin is intersexual, presumably because of
substances contributed by the male twin to
the common blood during uterine growth.
The freemartin intersexuality may be graded
into perfectly functioning fertile females to
types with external female genitalia and
typically male sex cords except germ cells
are absent, vasa efferentia, and elements of
the vasa deferentia. The conditions are simi-
lar to those discussed for amphibia, fish, and
rabbits in which early sex development
passes through neutral stages during which
it may be directed toward one sex or the
other by the right environmental stimuli.

Intersexes in swine have been interpreted
as owing to similar causes (Hughes, 1929;
Andersson, 1956) although the resulting
phenotypes may not be quite as extreme.
The resulting intersexes for both cattle and
swine presumably are not caused by chromo-
somal misbehavior but to the right environ-
mental stimuli operating on suitable gene
backgrounds. The observations of Johnston,
Zeller and Cantwell (1958) on 25 intersexual
pigs all from one breeding group of York-
shires suggest significant inheritance effects.
The intersexes were of two types, "male
pseudohermaphrodites" and "true hermaph-
rodites," but there was some intergrading of
their phenotypes suggesting that they may
be the products of like causes. Common or-
gans between the two groups included uteri,
vulvae, vaginae, testes, epididymis, and
penis or enlarged clitori. The "true hermaph-
rodites" were separated on the basis of no
prostates, bulbo-urethral glands, or seminal
vesicles as well as having testes or ovotestes
with ovaries. A similar case was described
by Hammond (1912) but, as in one of the
above cases, the supposed ovaries when sec-
tioned seemed to be lymphatic tissue. Favor-
able nerve tissue^ from 6 of the Yorkshire
pigs was examined foi- nuclear chromatin.
The cases were found chromatin positive.
Phenotypically these cases also have paral-
l(>ls in mice and man.

E. .SEX-iN man: chromosomal basis

A surprise even to its discoverers, Tjio
and Levan (1956), came with the observa-
tion that the somatic number of chromo-
somes in cultures of human tissue was 46
rather than the previously supposed 48.
Search for the true number has been going
on for more than half a century. In early
investigations the numbers reported varied
widely. Difficulties of proper fixation and
spreading of the chromosomes of human
cells accounted for most of this variation
and the numerous erroneous interpretations.
Among the observations that of de Wini-
warter (1912) was of particular interest in
showing the chromosome number as 46
autosomes plus one sex chromosome with
the Y being absent. This number was also
found later by de Winiwarter and Oguma
(1926). Observations by Painter (1921,
1923) showed 46 chromosomes plus an X
and a Y, a total of 48. This number was
subsequently reported by a series of able
investigators, Evans and Swezy (1929),
Minouchi and Ohta (1934), Shiwago and
Andres (1932), Andres and Navashin
(1936), Roller (1937), Hsu (1952), Mitt-
woch (1952), and Darlington and Haque
(1955). As Tjio and Levan indicated, the
acceptance of 48 as the correct number,
with X and Y as the sex chromosome
arrangement, was so general that when
Drs. Eva Hanson-Melander and S. Kul-
lander had earlier found 46 chromosomes
in the liver cells of the material they
were studying they temporarily gave up the
study. In the few years since 1956, the ac-
ceptance of 46 chromosomes as the normal
complement of man has become nearly
universal. There are 22 paired autosomes
plus the X and Y sex chromosomes.

The reasons which have warranted this
change of viewpoint are no doubt many,
but three improvements in technique are
certainly significant. The first came as a
consequence of simplifying the culture of
human somatic cells. The second followed
Hsu's (1952) recognition that pretreatment
of these cells before fixation with hypotonic
solutions tended to better spreads of the
chromosomes on the division plates when
subsefiuently stained by the squash tech-
niciuo. Pretreatment of the cultures with

FOUNDATIONS FOR SEX

53

colchicine made the studies more attractive
by increasing the numbers of usable cells
that were in the metaphase of cell division.

Ford and Hamerton (1956) in an inde-
pendent investigation, closely following
that of Tjio and Levan, observed that the
human cell complement contained 46 chro-
mosomes. They, too, agreed with Painter
and others that followed him that the male
was XY and the female XX in composition.
A flood of confirming evidence soon fol-
lowed: Hsu, Pomerat and Moorhead (1957),
Bender (1957), Syverton (1957), Ford,
Jacobs and Lajtha (1958), Tjio and Puck
(1958), Puck (1958), Chu and Giles (1959),
and a number of others.

In most instances the results of the dif-
ferent investigators were surprisingly con-
sistent in showing that the individual cell
chromosome counts nearly always totaled
46. This was no doubt due in part to the
desirability of single layers of somatic cells
for identifying and separating the different
chromosomes into distinct units. Chu and
Giles' results illustrate this consistency.
For 34 normal human subjects, including
29 American whites and 4 American Ne-
groes, and one of unknown race, and re-
gardless of sex, age, or tissue, the diploid
chromosome number of the somatic cells
was overwhelmingly 46. In only five indi-
viduals were other numbers observed in
isolated cells. Out of 620 counts, 611 had
46 chromosomes; two individuals, whose
majority of cells showed 46, had 3 cells with
45 chromosomes; three other individuals,
the majority of whose cells showed 46, had
6 cells with 47 chromosomes. Average cell
plates counted per individual was nearly 20.

The only recent observations at variance
with these results were those of Kodani
(1958) who studied spermatogonial and
first meiotic metaphases in the testes from
15 Japanese and 8 whites. In these studies
at least several good spermatogonial meta-
phases in which the chromosomes could be
counted accurately, and secondly at least
15 spermatocyte metaphases in which the
structure of individual chromosomes could
be observed clearly, were made on each
specimen. The numbers of cells studied in
metaphase were generally above these num-
bers, one reaching 60 metaphases. Some var-
iation was noted within individuals. Among

individuals, numbers of 46, 47, and 48 were
observed. Among 15 Japanese, 9 had 46, 1
had 47, and 5 had 48 chromosomes, whereas
among the whites 7 had 46, and 1 had 48.
Sixteen of the 23 individuals had 46 chro-
mosomes. Karyotype analyses indicated
that the numerical variation was caused by
a small supernumerary chromosome. On the
basis of these observations it would appear
that individuals within races may vary in
chromosome number and yet be of normal
phenotype. However, in view of the exten-
sive observations by others, it seems un-
likely that the variation between individ-
uals is as large as that indicated. It will
require much further study to establish any
other number than 46 as the normal karyo-
type of man. This is particularly true in
view of the work of Makino and Sasaki
(1959) and Alakino and Sasaki cited by
Ford (1960), in which they studied the hu-
man cell cultures of 39 Japanese and found
without exception 46 chromosomes, and the
earlier work of Ford and Hamerton (1956)
on spermatogonial material where they, too,
found 46 chromosomes in that tissue. The
best features of these human chromosome
studies will come in the identification of
the individual chromosomes making up the
human group. The chromosome pairs may
be ordered according to their lengths. The
longest chromosome is about 8 times the
length of the smallest. The chromosomes
may be classified according to their centro-
mere positions. The chromosomes are said
by most observers to be fairly easily sepa-
rated into 7 groups. Separation of the indi-
vidual chromosome pairs from each other
and designation of the pairs so that they
can be identified by trained investigators
in all good chromosome preparations is not
possible according to some ciualified cytolo-
gists and admitted difficult by all students.
However, standardized reporting in the
rapidly growing advances in human cell
studies should refine observations, reduce
errors, and encourage better techniques.
With this in mind, 17 investigators working
in this field met in Denver in 1959 in what
has come to be called the "Denver confer-
ence" (Editorial, 1960). From an examina-
tion of the available evidence on chromo-
some morphologies an idiogram was set up
as a standard for the somatic chromosome

54

BIOLOGIC BASIS OF SEX

CD
LO

00
CM

^

m

CM
CM

'li

m

^

CM

i|

m

m

8

1^

9m

c

£

II

CM

C30

£

Sg

^

*o

1

«.

C^

►5

'^

^

9^

*-

-f =

O

^

•^

CD
in

^ i

<7>

^D

4M

.£ i -^

CD

«

«

>t

J 2 ?:

C^

-•

«

CO

cording to
not easily ]
and Thood(

€

<

CM

CM

)ered ac
ich are
H. Tjio

C

<

S

s numl
tes \vh
I of J.

n

«

CM

Ill

• CD

^ m-m %J7

#c #• *—

iO

CM

in

\^ 00

2 >e #c i£

CM

s s S c
£ £ ^-^i

a>

d

-e

LO

S J^ £

PI

O q; i

CO

o

^ct

v£

Fig. 1.1. Id

r number, th
chromosom

r^

00

FOUNDATIONS FOR SEX

55

complement of the normal human genome.
A reproduction of this standard is presented
in Figure 1.1, as kindly loaned by Dr. Theo-
dore T. Puck for this purpose.

The autosomes were first ordered in re-
lation to their size and such attributes as
would help in their positive identification.
Numbers were given to each chromosome as
a means of permanent identification. Bas-
ically, identification is assisted by the ratio
of the length of the long arm to that of the
short arm; the centromeric index calculated
from the ratio of the length of the shorter
arm to the whole length of the chromosome ;
and the presence or absence of satellites.
Classification is assisted by dividing the
chromosome pairs into seven groups.
Groups 1-3. Large chromosomes with ap-
proximatel}^ median centromeres. The
three chromosomes are readily distin-
guished from each other by size and
centromere position.
Group 4-6. Large chromosomes with sub-
median centromeres. The two chromo-
somes are difficult to distinguish, but
chromosome 4 is slightly longer.
Group 6-12. Medium sized chromosomes
with submedian centromeres. The X
chromosome resembles the longer chro-
mosomes in this group, especially chro-
mosome 6, from which it is difficult to
distinguish. This large group is the one
which presents major difficulty in iden-
tification of individual chromosomes.
Group 13-15. INledium sized chromosomes
with nearly terminal centromeres ("ac-
rocentric" chromosomes). Chromosome
13 has a prominent satellite on the
short arm. Chromosome 14 has a small
satellite on the short arm. No satellite
has been detected on chromosome 15.
Group 16-18. Rather short chromosomes
with approximately median (in chro-
mosome 16) or sul>median centromeres.
Group 19-20. Short chromosomes with ap-
proximately median centromeres.
Group 21-22. Very short, acrocentric chro-
mosomes. Chromosome 21 has a satel-
lite on its short arm. The Y chromo-
some belongs to this group.
Separations of the human chromosome
pairs into the seven groups is not as difficult
as designating the pairs within groups
(Patau, 1960). The svstem is a notable ad-

vance in summarizing visually the current
information in the hope that availability of
such a standard will promote further refine-
ments, lessen misclassification, and contrib-
ute to a better understanding of the problems
by cytologists and other workers in the field.

1. Xuclear Chromatin, Sex Chromatin

Sexual dimorphism in nuclei of man
(Barr, 1949-59) and certain other mammals
may be detected by the observable presence
of nuclear chromatin adherent to the inner
surfaces of the nuclear membrane. The ma-
terial is about 1 /x in diameter. It frequently
can be resolved into two components of
equal size. It has an affinity for basic dyes
and is Feulgen and methyl green positive.
Nuclear chromatin can be recognized in 60
to 80 per cent of the somatic nuclei of fe-
males and not more than 10 per cent of
males. It is known to be identifiable in the
females of man, monkey, cat, dog, mink,
marten, ferret, raccoon, skunk, coyote,
wolf, bear, fox, goat, deer, swine, cattle, and
opossum, but is not easily usable for sex
differentiation in rabbit and rodents be-
cause these forms have multiple large par-
ticles of chromatin in their nuclei. The tests
can be made quickly and easily on skin
biopsy material or oral smears. Extensive
utilization of the presence or absence of
nuclear chromatin in cell samples of man
has been made for assigning the presumed
genetic sex to individuals who are pheno-
typically deviates from normal sex types.
(See also chapters by Hampson and Hamp-
son, and by Money.) Numerous studies on
normal individuals seem to support the
test's high accuracy. However, in certain
cases involving sexual modification, ques-
tions have arisen which are only now being
resolved. In male pseudohermaphroditism,
sex, determined by nuclear chromatin, is
male, thus agreeing with the major aspects of
the phenotype. For female pseudohermaph-
roditism, individuals with adrenal hyper-
plasia or those without adrenal hyperplasia
give the female nuclear chromatin test. For
cases listed as true hermaphrodites Grum-
bach and Barr (1958) list 6 of the male type
and 19 of the female type. For the syndrome
of gonadal dysgenesis they list 90 as male
and 12 as female among the proved cases
and 15 more as female among those that are

56

BIOLOGIC BASIS OF SEX

suspected. In the syndrome of seminiferous-
tubule dysgenesis where there is tubular
fibrosis, 9 are listed as male and 18 female.
Where there is germinal aplasia, 15 are
listed as male and 1 as female. The seeming
difficulties in assigning a sex constitution
to some of these types are now being dis-
sipated through the study of the full chro-
mosome complements which are responsible
for these different disease conditions. As ob-
servations on different chromosome types
have been extended, evidence has accumu-
lated to show that the numbers of sex nuclear
chromatins, for at least some of the nuclei
making up the organism, often equals
(n — 1) times the number of X chromo-
somes. The majority of male XY nuclei are
chromatin negative as are most of the Tur-
ner XO type. Female nuclei XX have a sin-
gle chromatin positive element as do the
XXY and XXYY types. The XXX and
XXXY have 14 and 40 per cent respectively
with two Barr bodies in cases for which
quantitative data are available. However, a
child with 49 chromosomes, but whose cul-
tured cell chromosomes appear as single
heteropycnotic masses making identification
of the individual chromosomes difficult,
showed 50 per cent of the cell nuclei with
three Barr elements (Fraccaro and Lindsten,
1960) . The chromosome constitution of these
nuclei was interpreted as trisomic for 8, 11,
and sex chromosomes. Sandberg, Cross-
white and Gordy (1960) report the case of a
woman 21 years old having various somatic
changes which does not fit this sequence. The
chromosome number was 47 and the nuclei
were considered trisomic for the sixth largest
chromosome. Two chromatin positive bodies
were ])rosent in the nuclei.

2. Chrotnosome Complement und Phenotijpe
in Man

Experience of the past 50 years has em-
phasized that genes and trisomies or other
types of aneuploid chromosome complexes
may lead to the development of abnormal
phenotypes expressing a variety of charac-
teristics. Drosophila led the way in illustrat-
ing how the different gene or chromosome
arrangements may affect sex expression. In-
vestigations of human abnormal types, par-
ticularly those with altered sex differentia-
tion, have reccntlv .^liown that man follow.-^

other species in this regard. The Y carries
highly potent male influencing factors. Gene
differences often lead to characteristic phe-
notypes of unique form.

3. Testicular Feminization

The testicular feminization syndrome il-
lustrates one of these types. As described by
Jacobs, Baikie, Court Brown, Forrest, Roy,
Stewart and Lennox (1959), "In complete
expression of this syndrome the external
genitalia are female, pubic and axillary hair
are absent or scanty, the habitus at puberty
is typically female, and there is primary
amenorrhoea. The testes can be found either
within the abdomen, or in the inguinal
canals, or in the labia majora, and as a rule
the vagina is incompletely developed. An
epididymis and vas deferens are commonly
present on both sides, and there may be a
rudimentary uterus and Fallopian tubes.
The condition is familial and is transmitted
through the maternal line." A sex-linked
recessive, a sex-limited dominant, and chro-
mosome irregularities of the affected per-
sons have been postulated as mechanisms
causing the apparent inheritance of this
condition. Chromosome examinations of the
cells of affected persons have shown 46 as
the total number and X and Y as the sex
complement. The karyotype analysis agrees
with the Barr nuclear chromatin test in
that the cells are chromatin-negative but
both are at variance with the sex pheno-
types in the sense that aside from sup-
pressed testes the patients are so completely
female. Genetically, Stewart (1959) has de-
scribed two color-blind patients with the
testicular feminization syndrome in the first
five patients he reported. The limited data
from these cases suggest that the genie basis
for this condition is either independent or
but loosely linked with color blindness. This
evidence does not exclude sex-linkage but
does make it less probable. The third hy-
pothesis of autosomal inheritance may take
one of several forms. A recessive gene which
affects only the male phenotypes when in
homozygous condition is apparently un-
tenable because the matings from which
these individuals come are of the outbreed-
ing type and the ratios apparently do not
differ from the one-to-one ratio expected of
a heterozygous dominant instead of that re-

FOUNDATIONS FOR SEX

57

quired for an autosomal recessive. The hy-
pothesis advanced by Witschi, Nelson and
Segal (1957), that the presence of an auto-
somal gene in the mother converts all her
male offspring into phenotypes of more or
less female constitution, in a manner com-
parable to that of the Ne gene in Dro-
sophila (Gowen and Nelson, 1942) which
causes the elimination of all the female type
zygotes, is also made unlikely by the ratios
of normal to testicular feminization pheno-
types observed in the progenies of these
affected mothers. The evidence favors a
simple autosomal dominant, acting only in
the male zygotes and perhaps balanced by
some genes of the X chromosome, which
have sufficient influence on the developing
male zygote to guide it toward an inter-
mediate to nearly female phenotype. The
observations of Puck, Robinson and Tjio
( 1960) indicate that the action of a gene for
this condition may not be entirely absent
in the female, because in heterozygous con-
dition in an XX individual it seemed to
delay menarche as much as 8 years. If this
delay be diagnostic for the heterozygote, it
will further assist in the genetic analysis of
this problem. Evidence on this point should
be a part of the genetic studies.

Cases closely similar to those described
by Jacobs, Baikie, Court Brown, Forrest,
Roy, Stewart and Lennox (1959) are
presented by Sternberg and Kloepfer
(1960). The patients show no trace of mas-
culinity. They are remarkably uniform in
anatomic expression. Except for failure to
menstruate due to lack of uteri they un-
dergo normal female puberty. Cryptorchid
testes, usually intra-abdominal, if removed
precipitate menopause symptoms. Four un-
related cases were found in this one study
with 7 additional cases traced through pedi-
gree information. A total of 11 affected in-
dividuals was found in 6 sibships having
26 siblings of whom 5 were normal males.
In each kindred the inheritance was com-
patible with that of a sex-linkecl recessive
gene. A chromosomal study of a thyroid
tissue culture from one case revealed 46
chromosomes with normal XY male con-
figuration. The individuals observed were
designated as ''simulant females."

4- Superfemale

The human superfemale has been recog-
nized by Jacobs, Baikie, Court Brown, Mac-
Gregor, Maclean and Harnden (1959) in
a girl of medium height and weight, breasts
underdeveloped, genitalia infantile, vagina
small, and uterocervical canal 6 cm. in
length. Ovaries appeared postmenopausal
with normal stroma, and as indicated by a
biopsy specimen, deficient in follicle forma-
tion. Menstruation was thought to have
begun at age 14, but was irregular, occurring
every 3 to 4 months and lasting 3 days. The
last spontaneous menstruation was at 19.
Estrogen therapy caused some development
of the breasts and external genitalia, vagina,
and uterus with slight uterine bleeding. The
patient's parents were above 40 years of
age, mother 41, at time of her daughter's
birth.

Examination of sternal marrow cultures
showed 47 chromosomes in over 80 per cent
of the cells examined. The extra chromo-
some was the X, the chromosomal type
being XXX plus 22 pairs of autosomes.
Buccal smears showed 47 per cent of nuclei
contained a single chromatin body and 14
per cent contained 2 chromatin bodies as
expected of a multiple XX or XXX geno-
type. In comparison, 25 smears from 20 nor-
mal women had 36 to 51 per cent chromatin
positive cells but none of these contained
2 chromatin bodies. Two chromatin bodies
were seen in some cells of the ovarian stro-
mal tissue. The patient showed a lack of
vigor, mentally was subnormal, was under-
developed rather than overly developed in
the phenotypic sexual characteristics. Ex-
amination of the patient's mother showed
her to be XX plus 22 pairs of autosomes,
the normal 46 chromosomes.

Other cases show that types with XXX
plus 22 pairs of autosomes are of female
l)henotype but may vary in fertility and
development of the secondary sexual char-
acteristics from nonfunctional to functional
females bearing children ( Stewart and San-
derson, 1960; Eraser, Campbell, MacGilli-
vray, Boyd and Lennox, 1960). The triplo
X condition in man has a greater range of de-
velopment and fertility than in Drosophila.
In man ovaries may develop spontaneously.
In Drosophila they require transplantation

58

BIOLOGIC BASIS OF SEX

to a diploid female host where they may at-
tach to the oviducts and release eggs for
fertilization (Beadle and Ephrussi, 1937).

These cases present confirmation of two
facts already mentioned for Drosophila.
They show that when the X chromosome
has primarily sex determining genes, the
organism generally becomes unbalanced
when 3 of these X chromosomes are matched
against two sets of autosomes. The re-
sulting phenotypes are female but relatively
undeveloped rather than overdeveloped.
The second is that the connotations evoked
by the prefix "super" are by no means ap-
plicable to this human type or to the Dro-
sophila type.

The characteristics of the patient also
suggest that the autosomes may be carrying
sex genes opposing those of female tenden-
cies as observed in both Drosophila and
Rumex genie imbalance.

5. Klinefelter Syndrome

In the Klinefelter syndrome there is male
differentiation of the reproductive tracts
with small firm descended testes. Meiotic or
mitotic divisions are rare, sperm are ordi-
narily not found in the semen. The type is
eunuchoid in appearance with gyneco-
mastia, high-pitched voice, and sparse fa-
cial hair growth. Seminiferous tubules show-
ing an increased number of interstitial cells
are atrophic and hyalinized. Urinary excre-
tion of pituitary gonadotrophins is generally
increased, whereas the level of 17-keto-
steroids may be decreased. The nuclear
chromatin is typically female. Of the dozen
or more cases studied (Jacobs and Strong,
1959; Ford, Jones, Miller, Mittwoch, Pen-
rose, Ridler and Sha])iro, 1959; Bergman
and Reitalu quoted by Ford, 1960), only
one, having but 5 metaphase figures, had
less than 47 chromosomes in the somatic
cells and XXY sex chromosomes. That case
was thought to have typical female chro-
mosomes XX + 22 AA. Two other cases were
of particular interest as indicating further
chromosome aberration. Ford, Jones, Miller,
Mittwoch, Penrose, Ridler and Shapiro
(1959) studied one patient who displayed
both the Klinefelter and Mongoloid syn-
dromes. The chromosome number was 48,
the sex chromosomes being XXY and the
48tli chromosoinc being small acrocentric.

This individual had evidently developed
from an egg carrying 2 chromosomal aberra-
tions, one for the sex chromosomes and the
second for one of the autosomes. The other
case, Bergman and Reitalu as cited by
Ford (1960), had 30 per cent of its cells
with an additional acrocentric chromosome
which had no close counterpart in the nor-
mal set.

Data where the Klinefelter syndrome oc-
curs in families showing color blindness
(Polani, Bishop, Ferguson-Smith, Lennox,
Stewart and Prader, 1958; Nowakowski,
Lenz and Parada, 1959; and Stern, 1959a)
further test the XXY relationship and give
information on the possible position of the
color blindness locus with reference to the
kinetochore. Polani, Bishop, Ferguson-
Smith, Lennox, Stewart and Prader (1958)
tested 72 sex chromatin-positive Klinefelter
patients for their color vision and found
that none was affected by red-green color
blindness. Nowakowski, Lenz and Parada
( 1959) tested 34 cases and detected 3 af-
fected persons, 2 of whom were deutera-
nomalous and one protanopic. Stern (1959a)
l^oints out that these cases and their ratios
are compatible with the interpretation of
the Klinefelter syndrome as XXY. One of
the deuteranomalous cases had a deutera-
nomalous mother and a father with normal
color vision. This case could have originated
from a nondisjunctional egg carrying 2
maternal X chromosomes fertilized by a
sperm carrying a Y chromosome. The other
two cases had normal fathers with hetero-
zygous mothers. There are several explana-
tions by which the color-blind Klinefelter
progenies could be obtained. The hetero-
zygotes might manifest the color-blind con-
dition. The second hypothesis, which is
favored, is that of crossing over between the
kinetochore and the color-blind locus at the
first meiotic division to form eggs each
carrying 2 X chromosomes, one homozygous
for color blindness, and the other for normal
vision. An equational nondisjunction would
form eggs homozygous for color blindness
which on fertilization by the Y chromo-
somes of the male would give the necessary
XXY constitution for the color-blind male
which is Klinefelter in phenotype. A third
possibiHty is that these exceptions may
arise without crossing over as the result of

FOUNDATIONS FOR SEX

59

nondisjunction at the second meiotic divi-
sion.

If the hypothesis of crossing over is ac-
cepted, the color-blind locus separates freely
from its kinetochore and would suggest that
the position of the locus is at some distance
from the kinetochore of the X chromosome.

A disturbed balance between the X and
the Y chromosomes alters the sexual type.
A single Y chromosome, contributing fac-
tors important to male development, is able
to alter the effects of two sets of female
influencing X chromosomes. Yet two Y
chromosomes in a complex of XXYY plus
44 autosomes seem to have little or no more
influence than one Y (Muldal and Ockey,
1960). The locations of the sex-influencing
genes in man are thus more like those of the
plant Melandrium than of Drosophila in
which the male-determining factors occur
in the autosomes. The relative potencies of
the male sex factors compared with those of
the female, however, are much less than
those in Melandrium.

6. Turner Syndrome

Turner's syndrome or ovarian agenesis
further substantiates the female influence
of the X chromosomes. The cases occur as
the developmental expression of accidents in
the meiotic or mitotic divisions of the chro-
mosomes. These accidents lead to adults
unbalanced for the female tendencies of the
X chromosome. The gonads consist of con-
nective tissue. The rest of the reproductive
tract is female. Growth stimuli of puberty
are lacking, resulting in greatly reduced fe-
male secondary sexual development. Pa-
tients are noticeably short and may be ab-
normal in bone growth. In its more extreme
form, designated as Turner's syndrome, the
individuals may show skin folds over the
neck, congenital heart disease, and subnor-
mal intellect, as well as other metabolic
conditions. Earlier work (Barr, 1959; Ford,
Jones, Polani, de Almeida and Briggs,
1959) shows that 80 per cent of the nu-
clear chromatin patterns are of the male
type. Evidence from families having both
this condition and color blindness suggested
that at least some of the Turner cases would
be found to have 45 chromosomes, the sex
chromosome being a lone X (Polani, Lessof
and Bishop, 1956). Work of Ford, Jones,

Polani, de Almeida and Briggs, (1959)
has confirmed this hypothesis and added the
fact that some of these individuals are also
mosaics of cells having 45 and 46 chromo-
somes. The 45 chromosome cells had but one
X, whereas the 46 had two X's. This finding
may explain the female-chromatin cell type
observed in about 20 per cent of the cases
having the Turner syndrome. Such mosaics
of different chromosome cell types could
also be significant in reducing the severity
of the Turner syndrome and in increasing
the range of symptoms which characterize
this chromosome-caused disease as con-
trasted with those characterizing Turner's
disease. Further cases observed in other
investigations, Fraccaro, Kaijser and Lind-
sten (1959), Tjio, Puck and Robinson
(1959), Harnden, and Jacobs and Stewart
cited by Ford (1960) have all shown 45
chromosome cells and a single X chromo-
some. As with the XXX plus 44 autosome
super females, the Turner type, X plus 44
autosomes, also shows a rather wide range
in development from sterility with extensive
detrimental secondary effects to nearly nor-
mal in all respects. Bahner, Schwarz, Harn-
den, Jacobs, Hienz and Walter (1960) re-
port a case which gave birth to a normal
boy. Other cases have been described (Hof-
fenberg, Jackson and jVIuller, 1957; Stewart,
19601 in which menstruation was estab-
lished over a period of years. The XO type
in man and Melandrium is morphologically
female. In Drosophila on the other hand,
the XO type is phenotypically nearly a
perfect male. It is further to be noted that
the X chromosome of Drosophila appears to
have a less pronounced female bias than
that of man when balanced against its as-
sociated autosomes, inasmuch as the XO +
2A type in Drosophila is male as contrasted
with the XO + 2A type in man which is
female. At the same time it seems that the
autosomes in the human may be influential
in that the female gonadal development is
suppressed instead of going to completion
as it does in the XX type.

7. Hermaphrodites

Hermaphroditic phenotypes in man, to
the number of at least 74 (Overzier, 1955),
have been observed and recorded since
1900. Types with a urogenital sinus pre-

60

BIOLOGIC BASIS OF SEX

dominated. The uteri were absent in some
cases, even when complete external female
genitalia were present. Ovotestes were found
on the right side of the body in over half
the cases; separate left ovaries or testes
were about equally frequent ; in three cases
separate testis and ovary were indicated.
The left side of the body showed a different
distribution of gonad types; about one-
fourth had ovotestes, another fourth ova-
ries, and one-twelfth testes. Unilateral dis-
tribution of gonad types was most frequent.
The presence or absence of the prostate
seemed to have significance because it is
sometimes absent in purely female types. In
recent literature similar cases have been
called true hermaphrodites. This is an ex-
aggeration in terms of long established
practice in plants and animals where true
hermaphroditism includes fully functioning
gametes of each sex.

Hungerford, Donnelly, Nowell and Beck
(1959» have reported on a case of a Negro
in which the culture cells had the chromo-
some complement of a normal female 46,
with XX sex chromosomes. Unfortunately,
the possibility that this case may be a chro-
mosome mosaic was not tested by karyo-
type samples from several parts of the body.

Harnden and Armstrong (1959) estab-
lished separate skin cultures from both sides
of the body of another hermaphroditic type.
The majority of the cells were apparently of
XX constitution with a total of 46 chromo-
somes. However, in one of the 4 cultures
established, some 7 per cent of cells had an
abnormal chromosome present, suggesting
that the case might involve a reciprocal
translocation between chromosomes 3 and
4 when the chromosomes were ordered ac-
cording to size. All the other cell nuclei were
normal. The fact that the majority of the
cells in these two cases were XX and with
46 chromosomes seems to predicate against
the view that either changes in chromosome
number or structure of the fertilized egg are
necessary for the initiation of hermaplu'o-
dites.

Ferguson-Smith (1960) describes two
cases of gynandromorphic type in which the
reproductive organs on the left side were
female and on the right side were male. The
recognizable organs were Fallopian tube,
ovary with primordial follicles only, imma-

ture uterus in one case, none in the other,
rudimentary prostate, small testis and epi-
didymis, vas deferens, bifid scrotum, phal-
lus, perineal urethra, pubic and axillary
hair, breasts enlarging at 14 years. Testicu-
lar development with hyperplasia of Leydig
cells, germinal aplasia, and hyalinization of
the tubules was suggestive of the Klinefelter
syndrome. Nuclear-chromatin was positive
in both cases. Modal chromosome number
was 46. The sex chromosomes were inter-
preted as XX. The 119 cell counts on one
patient showed a rather wide range; 7 per
cent had 44 chromosomes, 13 per cent had
45, 62 per cent had 46, and 18 per cent had
47 chromosomes. The extra chromosome
within the cells containing 47 chromosomes
was of medium size with submedian kineto-
chore as generally observed for chromo-
somes of group 3.

It is surprising that the male differentia-
tion in these four and other hermaphroditic
cases (Table 1.5) is as complete as it is.
Other observations show that the Y chro-
mosome contains factors of strong male
potency, yet in its absence the hermaphro-
dites develop an easily recognized male
system. It is not complete but the degree of
gonadal differentiation is as great as that
observed in the XXY 4- 2A Klinefelter
types. The bilateral sex differentiation in
hermaphrodites would seem to require other
conditions than those heretofore considered.

Another case of hermaphroditism is that
presented by Hirschhorn, Decker and
Cooper (1960). The patient's j:)henotype
was intersexual with phallus, hypospadias,
vagina, uterus. Fallopian tubes, two slightly
differentiated gonads in the position of ova-
ries. The child was 4 months old. Culture
of bone marrow cells showed that the indi-
vidual was a mosaic of two types. About
60 per cent of the cells had 45 chromosomes
of XO karyotype, and 40 per cent had 46
chromosomes with a karyotA'^pe XY. The
Y chromosome when present was larger
than Y chromosomes of normal individuals.
The change in size may be related to the
association of the XO and XY cells and
be similar etiologically to the case discussed
by yivtz (1959) in Sciara triploids.

There are mosaics in Drosophila formed
from the loss by the female in some cells of
one of lioi- X chromosomes, as for instance

FOUNDATIONS FOR SEX

61

in ring chromosome types, which may dis-
play primary and secondary hermaphroditic
development. For this to happen the altered
nuclei apparently find their way into the
region of the egg cytoplasm which is to
differentiate into the reproductive tract. As
seen in the adults, organ tissue of one chro-
mosome type is cell for cell sharply differ-
entiated from that of the other chromosome
type with regard to sex. These observations
indicate that for these mosaics the basic
chromosome structure of the cell itself
determines its development. In fact most
mosaics of this species show this cell-re-
stricted differentiation. Several problems
arise when these well tested observations
are considered in comparison with those
now arising in the chromosome mosaics of
the sex types in man. It would seem unlikely
that the bone marrow cells or for that
matter any somatic cells not a part of the
reproductive tract would operate to modify
the adult sex or a part thereof. Rather the
developmental secjuence should start from
cell differences within the early developing
reproductive tract. Circulatory cells or sub-
stances would be of dubious direct signifi-
cance from another viewpoint. All cells of
the body would ultimately be about equally
affected by any cells or elements circulating
in the blood. With strong male elements
and strong female elements the result ex-
pected would be a reduction in sex develop-
ment of either sex instead of the sharply
differentiated organ systems which are ob-
served. This raises the question, are the
chromosomally differentiated cell sex mo-
saics primary to or secondarily derived
from the tissues of the ultimate hermaphro-
dites? Study of the cell structure of the sex
organs themselves as well as much other
information will be necessary to clear up
this problem.

There are, however, other types of con-
trolled sex development, as by various
genes, which lead to the presence of both
male and female sexual systems. Genes
for these phenotypes are relatively rare, but
once found are transmitted as commonly
expected. The inherited hermaphroditic
cases in Drosophila are certainly relevant
to the testicular feminization syndrome in
man. Are they equally pertinent to the
highly sporadic hermaphroditic forms just

considered for man? If so, they indicate a
genie basis for these types which would
probably be beyond the range of the micro-
scope to detect. The low frequency of true
hermaphrodites in the human, together with
lack of information on possible inheritance
mitigates against the genie explanation ; al-
though genie predisposition acting in con-
junction with rare environmental events as
occurs in our Balb/Gw mice (Hollander,
Go wen and Stadler, 1956 ) could explain the
rare hermaphrodites observed in that par-
ticular line of mice and a limited number of
its descendants.

8. XX XY + U Autosome Type

The XXXY -I- 44 autosome type in the
human has been studied by Ferguson-Smith,
Johnson and Handmaker (1960) and
Ferguson-Smith, Johnston and Weinberg
(1960). The two cases described were char-
acterized by primary amentia, micro-
orchidism and by two sex chromatin bodies
in intermitotic nuclei. The patients were
similar in having disproportionately long
legs; facial, axillary, and abdominal hair
scant; pubic hair present; penes and scrota
medium to well developed ; small testes and
prostates; vasa deferentia and epididymides
normally developed on both sides of the
body and no abnormally developed Miiller-
ian derivatives. Testes findings were like
those in Klinefelter cases with chromatin-
positive nuclei, small testes with nearly
complete atrophy, and hyalinization of
seminiferous tubules and islands of abnor-
mal and pigmented Leydig cells in the
hyalinized areas. The few seminiferous
tubules present were lined with Sertoli cells
but were without germinal cells. Nuclear
chromatin was of female type. About two-
fifths of the nuclei had double and two-
fifths single sex chromatin. The modal chro-
mosome count for bone marrow cells was
48, 75 per cent of the cells having this
number. Chromosome counts spread from
45 to 49. This type, XXXY plus 44 auto-
somes, may be looked upon as a superfemale
plus a Y or a Klinefelter plus an X chro-
mosome. In either case the male potency of
the genes in the Y chromosome is able to
dominate the female tendencies of XXX to
develop nearly complete male phenotypes.

Both cases had severe mental defects but

TABLE 1.5
Chromosome kinds and numbers for different recognized sex types in man

External
Type

Male

Female

Female (rare he-
mophilic)

Eunuchoid female
Female

Female
Female
Female

Female
Female

Male .
Male.
Male.

Male.
Male.

Male . . .
Female.
Female .

Female .
Female .

Male.
Male.

Hermaphrodite

Hermaphrodite.
Intersex

Numbers of
Chromosomes

chromoso-
il rearrange-

1 + X
fragment

1
2
2

2
2
Interpreted a.s
XX trisomic
for Sand 11 or
as XXXX
4X + Yl

3X+ Yj
1
3
3

2l
/mosaic

Missing
or Extra
Chromo-
somes

? Chr
mosome 3

orT(X;A)

Small
auto-
some

? Large
chromo-
some or

T(X;A)

X or Y

(X or Y)
+21

46,47,48
cliromo-
somes
X, Y
X or Y
-fA set

Sex Chromatm

Negative
Positive

Negative
Negative
Positive

Negative
Positive
Negative

Negative

Negative
Negative

Negative ±

Negative
Positive
Positive

Triple positi

Negative
Double posit
Double posit

Negative

Double posit
Negative

Negative
Negative

Designating Term

Normal male
Normal female

Pure gonadal dysgenesis
CJonadal dysgenesis

Testicular feiuinizati

Turner

Turner type female

Tinner? gave birth to boy
Turner

Turner

Klinefelter

Klinefelter mongoloid

Klinefelter
Klinefelter

Klinefelter

Klinefelter
Sujierfemale

Superfemale gave birtli to
children

Testicular deficiency

Precocious puberty

Triploid

Hermaphrodite or inter-
sex dei)cnding on defini-
tion

Investigators*

2
3, 4, 5,

5, 12, 13, 14, 40, 41

41

16, 17, 18, 40, 41

19

40

23, 40

24, 25

41

26

29, 30, 31,. 32, 33
35, 39, 40

34

41

Tjio and Levan, 1956.

Nilsson, Bergman, Reitalu and Waldenstrom, 1959.
Harnden and Stewart, 1959.
Stewart, 1960b.
Stewart, 1960a.

Elliott, Sandler and Rabinowitz, 1959.

Jacobs, Baikie, Court Brown, Forrest, Roy, Stewart and Len-
nox, 1959.
Stewart, 1959

Lubs, Vilar and Bergenstal, 1959.
Sternberg and Kloepfer, 1960.
Puck, Robinson and Tjio, 1960.
Ford, Jones, Polani, de Almeida and Briggs, 1959.
Fraccaro, Kaijser and Lindsten, 1959.
Fraccaro, Kaijser and Lindsten, 1960a.

15. Bahner, Schwarz, Harnden, Jacobs, Hienz and Walter, 1960.

16. Jacobs and Strong, 1959.

17. Bergman, Reitalu, Nowakowski and Lenz, 1960.

18. Nelson, Ferrari and Bottura, 1960.

19. Ford, Jones, Miller, Mittwoch, Penrose, Hidlor and Shapiro,

1959.

20. Ford, Polani, Briggs and Bishor), 1959.

21. Crooke and Hayward, 1960.

22. Muldal and Ockey, 1960.

23. Jacobs, Baikie, Court Broun, MacCJregor, Maclean and

Harnden, 1959.

24. Eraser, Campbell, .MacCiillivray, Boyd and Lennox, 1960.

25. Stewart and Sanderson, 1960.

26. Jacobs, Harnden, Court Brown, Cohl.stein, Close, Mac-

Gregor, Maclean and Strong, 1960.

62

FOUNDATIONS FOR SEX

63

27. Ferguson-Smith, Johnston and Handiiiaker, 196(

28. Book and Santesson, 1960.

29. Harnden and Armstrong, 1959.

30. Hungerford, Donnelly, Nowell and Beck, 1959.

31. Ferguson-Smith, Johnston and Weinberg, 1960.

32. deAssis, Epps, Bottura and Ferrari, 1960.

33. Gordon, O'Gorman, Dewhurst and Blank, 1960

34. Hirschhorn, Decker and Cooper, 1960.

35. Sasaki and Makino, 1960.

36. Bloise, Bottura, deAssis, and Ferrari, 1960,

37. Fraccaro, Kaijser and Lindsten, 1960c.
37a. Fraccaro and Lindsten, 1960.

38. Fraccaro, Ikkos, Lindsten, Luft and Kaijser, 1960.

39. Harnden, 1960.

40. Ferguson-Smith and Johnston, 1960.

41. Sandberg, Koepf, Crosswhite and Hauschka, 1960.

42. Hayward, 1960.

it should be remembered that they were
sought in institutions for which this is a
criterion of admittance. Their mental abil-
ity was distinctly less than that of Kline-
felter XXY cases which have come under
study. The pattern of the XXXY effects on
the reproductive tract, however, was com-
parable with that observed in the XXY
genotypes. The effects of one Y chromosome
were balanced by either two or three X
chromosomes to give nearly equal pheno-
typic effects.

9. XXY + 66 Autosome Type

XXY + 66 autosome type was estab-
lished by Book and Santesson (1960) for
an infant boy having several somatic anom-
alies which may or may not be relevant to
the sex type. Externally the genitalia were
normal for a male of his age, penis and
scrotum with testes present in the scrotum.
Again the Y chromosome demonstrates its
male potencies over two X's even in the
presence of three sets of autosomes. The
case is of particular significance since fur-
ther development may indicate what male
potencies an extra set of autosomes may
possess.

10. Summary of Types

Other types of sex modifying chromo-
somal combinations and their contained
genes have been observed particularly as
mosaics or as chromosomal fragments added
or substracted from the normal genomes.
No doubt other types will be discovered
during the mushroom growth of this period.
Time can only test the soundness of the
observations for the field of human chro-
mosomal genetics and cytology is difficult
at best requiring special aptitudes and ex-
perience. Mistakes, no doubt, will be made.
The status of the subject is summarized
in Table 1.5.

11. Types Unrelated to Sex

Other cases not related to sex or only
secondarily so were scrutinized during the

course of these studies. The information
gained from them is valuable as it strength-
ens our respect for the mechanisms in-
volved. The sex types which are dependent
on loss or gain of the X and/or Y chromo-
somes belong to the larger category of
monosomies or trisomies. Numbers of auto-
somal monosomic and trisomic syndromes
have also been identified in the course of
these investigations. Similarly, not all cases
that have been studied have turned out to
be associated with chromosomal changes.
This in itself is important since it lends
confidence in those that have, as well as
redirects research effort toward the search
for other causes than chromosomal mis-
behavior. The first trisomic in man was
identified through the study of Mongolism.
The condition affects a number of primary
characteristics but not those of sex, for
males and females occur in about equal
numbers. The broad spectrum of these ef-
fects points to a loss of balance for an
equally extensive group of genes in the two
sexes. The common association of charac-
teristics making up these Mongoloids, to-
gether with their sporadic appearance and
their change in frequency with maternal
age, all suggest the findings which Lejeune,
Gautier and Turpin (1959a, b) and Lejeune,
Turi)in and Gautier (1959a, b) were able
to demonstrate so successfully. They estab-
lished that the tissue culture cells of Mongol-
oid imbeciles had 47 chromosomes and that
the extra chromosome was in the small
acrocentric group. Lejeune, Gautier and
Turpin (1959a, b) have now confirmed
these observations on not less than nine
cases. Jacobs, Baikie, Court Brown and
Strong (1959), Book, Fraccaro and Lind-
sten (1959) and Fraccaro (cited by Ford,
1960) as well as later observers have sub-
stantiated the results on more than ten
other cases. The well known maternal age
effect, whereby women over 40 have a
chance of having IMongoloid offspring 10
to 40 times as frequently as those of the
younger ages, would seem to point to non-

64

BIOLOGIC BASIS OF SEX

disjunction in oogenesis as the most im-
portant cause of this condition. Some women
who have had previous JMongoloid progeny
have an increased risk of having others.
This is an important consideration in that
genetic factors may materially assist in
bringing about nondisjunction in man as
they are known to do in Drosophila (Gowen
and Gowen, 1922; Gowen, 1928). The prod-
ucts of the nondisjunctions approach those
expected on random distribution of the
chromosomes (Gowen, 1933) so that occa-
sionally more than one type of chromosome
disjunction will appear in a given indi-
vidual. Such a case is that illustrated by
Ford, Jones, Miller, Mittwoch, Penrose,
R idler and Shapiro (1959) in which the
nondisjunctional type included not only
that for the chromosome important to Mon-
golism but also the sex chromosomes sig-
nificant in determining the Klinefelter
condition. This individual showed 48 chro-
mosomes, 22 pairs of normal autosomes, 3
sex chromosomes XXY, and a small acro-
centric chromosome matching a pair of
chromosomes, the 21st, within the smallest
chromosomes of the human idiogram. The
analysis of Mongolism showed the way for
the separation of the various human sex
types through chromosome analyses.

Chromosome translocations furnish an-
other means of establishing an anomaly
that may then continue on an hereditary
basis as either the male or female may
transmit the rearranged chromosomes. Po-
lani, Briggs, Ford, Clarke and Berg (1960),
Fraccaro, Kaijser and Lindsten (1960b),
Penrose, Ellis and Delhanty (1960) and
Carter, Hamerton, Polani, Gunalp and
Weller (1960) have studied Mongoloid
cases which they interpreted in this manner.
In some cases the rearranged chromosomes
have been transmitted for three generations.
Several of the translocations were con-
sidered to include chromosomes 15 and 21.

Another trisomic autosomal type was rc-
])ortcd by Patau, Smith, Therman, Inhorn
and Wagner (1960). The patient was fe-
male and had 47 chromosomes. The extra
chromosome was a medium-sized acrocen-
tric autosome belonging to the D group.
Despite extensive malformations affecting
several organs the patient lived more than
a year. Another female iiortraying the same

syndrome has since been found, so other
cases may be expected. Among the charac-
teristics are mental retardation, minor mo-
tor seizures, deafness, apparent micro or
anophthalmia, horizontal palmar creases,
trigger thumbs, Polydactyly, cleft i)alate,
and hemangiomata.

The third trisomic type was also reported
by Patau, Smith, Therman, Inhorn and
Wagner (1960). Six individuals have been
observed. The characters affected are men-
tal retardation, hypertonicity (5 patients),
small mandible, malformed ears, flexion of
fingers, index finger overlaps third, big toe
dorsiflexed (at least 4), hernia and/or dia-
phragm eventration, heart anomaly (at
least 4), and renal anomaly (3). The sexes
were two males and four females. The ex-
tra chromosome was in the E group and was
diagnosed as number 18. Edwards, Harnden,
Cameron, Crosse and Wolff (1960) have
described a similar case but they consider
the trisomic to be number 17. Ultimate com-
parisons of these types no doubt will decide
if this is a 4th trisomic or if all the cases
belong in the same group.

The Sturge-Weber syndrome apparently
is caused by another trisomic. Locomotor
and mental abilities are retarded. Hayward
and Bower (1960) interpret the 3 chromo-
somes responsible as the smallest autosomes,
number 22, of the human group.

Trisomic frequencies should be matched
by equal numbers of monosomies. Turpin,
Lejeune, Lafourcade and Gautier (1959)
have reported polydysspondylism in a child
with low intelligence, dwarfing, and multi-
l)le malformations of spine and sella turcica.
The somatic cell chromosome count was
only 45 but one of the smallest acrocentric
chromosomes appeared to have been trans-
located, the greatest part of this chromo-
some being observed on the short arm of one
of the 3 longer acrocentric chromosomes.
Th(> condition appears to be unique and not
likely to be found in other unrelated fami-
Vws. However, the phenotyj^ic effects were
so severe that all members of the proband's
family would seemingly be worthy of care-
ful sur\'('y for their chromosome character-
istics.

The comi)lex pattern of multiple anom-
alies renders each syndrome distinct from
the otliei's. Chromosome losses or gains from

FOUNDATIONS FOR SEX

Go

the normal diploid would be expected to
lead to the complex changes. Mongolism is
influenced by age of the mother and prob-
ably to some extent by her inheritance. It
is to be expected that the other trisomies
may show parallel relations. Other trisomies
may be expected although, as the chromo-
somes increase in size, a group of them will
have less opportunity to survive because of
loss of balance with the rest of the diploid
set. Thus far most of these conditions af-
fect the sex phenotypes. This is in accord
with the results in Drosophila. Changes in
the balance of the X chromosomes are less
often lethal than the gain or loss of an auto-
some. Other animals show like effects. In
plants, loss or gain of a chromosome, al-
though generally detrimental, often causes
less severe restrictions on life. Harmful ef-
fects are observed but do not cause early
deaths. This may be because many aneu-
ploids are within what are presumably
polyploid plant species.

Ford (1960) has collected the data on 13
different phenotypes that could come under
suspicion of chromosomal etiology as ex-
amined by a number of workers. Careful
cytologic examination of patients suffering
from one or another of these diseases has
shown that the idiograms were normal in
both number and structure of the chromo-
somes. The disease conditions were:
acrocephalosyndactyly, arachnodactyly

(Marfan's syndrome), chondrodystrophy,
Crouzon's disease, epiloia, gargoylism, Gau-
cher's disease, hypopituitary dwarfism,
juvenile amaurotic idiocy, Laurence-Moon-
Biedl syndrome. Little's disease, osteogene-
sis imperfecta, phenylketonuria, and anen-
cephalic types. To this list Sandberg, Koepf ,
Crosswhite and Hauschka (1960) have now
been added neurofibromatosis, Lowe's syn-
drome, and pseudohypoparathyroidism.

F. SEX RATIO IN MAN

Sex ratio studies on human and other ani-
mal populations have always been large in
volume. The period since 1938 is no excep-
tion. Geissler's (1889) data on family sex
ratios, containing more than four million
births, have been reviewed and questions
raised by several later analysts. Edwards
(1958) has reanalyzed the clata from this
population and considered these points and

reviewed the problems in the light of the
following questions: (1) Does the sex ratio
vary between families of the same size? (2)
Do parents capable of producing only uni-
sexual families exist? (3) Can the residual
deviations in tlie data be satisfactorily ex-
plained? Probability analyses were based
on Skellam's modified binomial distribution,
a special case of the hypergeometrical. The
following conclusions were drawn. The
probability of a birth being male varies
between families of the same size among
a complete cross-section of this 19th century
German population. There is no evidence
for the existence of parents capable of pro-
ducing only unisexual families. With the
assumption that proportions of males vary
within families, the apparent anomalies in
the data appear to be explicable. These
studies have a bearing on the variances ob-
served in further work dealing with family
differences such as that of Cohen and Glass
( 1959) on the relation of ABO blood groups
to the sex ratio and that of Novitski and
Kimball (1958) on birth order, parental age,
and sex of offspring. Novitski and Kim-
ball's data are of basic significance, for the
interpretations are based on a large volume
of material covering a one-year period in
which improved statistical techniques were
utilized in the data collection, in showing
that within these data sex ratio variation
showed relatively little dependence on age
of mother, whereas it did show dependence
on age of father, birth order, and inter-
actions between them. These observations
have direct bearing on the larger geographic
differences observed in sex ratios as dis-
cussed by Russell (1936) and have recently
been brought to the fore through the studies
of Kang and Cho (1959a, b). If these data
stand the tests for biases, they are of signifi-
cance in showing Korea to have one of the
highest secondary sex ratios of any region,
113.5 males to 100 females, as contrasted
with the American ratio of about 106 males
to 100 females. Of similar interest is the
lower rate of twin births, 0.7 per cent in
Korea vs. about 1 per cent in Caucasian
populations and the fact that nearly two-
thirds of these tW'in births in Korean peoples
are identical, whereas those in the Cauca-
sian groups are only about half that num-
ber. The reasons for these differences must

66

BIOLOGIC BASIS OF SEX

lie in the relations of the human X and Y
chromosomes and autosomes and the bal-
ance of their contained genes. Little or
nothing is known about how these factors
operate in the given situations.

Acknowledgment. In formulating and
organizing the material on which this paper
is based I have been fortunate in the helpful
discussions and analytical advice contrib-
uted so generously by Doctors H. L. Cai'son,
K. W. Cooper, H. V. Grouse, C. W. Metz,
S. B. Pipkin, W. C. Rothenbuhler, and H. D.
Stalker, and others having primary research
interests in this field. To them, and particu-
larly to my research associates Doctors S.
T. C. Fung and Janice Stadler, our secre-
tary Gladys M. Dicke and to my wife,
]\Iarie S. Gowen, I tender grateful acknowl-
edgment for their contributions to the manu-
script. Direct reference is made to some 300
investigations in the body of the paper but
the actual number, many of which could
equally well have been incorporated, that
furnished background to these advances is
much larger, more than 1600. To this vast
effort toward understanding the foundations
for sex we are all indebted.

XL References

AiDA, T. 1921. 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.

AiDA, T. 1936. Sex reversal in Aplocheilus latipes
and a new explanation of sex differentiation.
Genetics, 21, 136-153.

Andersson, T. 1956. Intersexuality in pigs.
Kungl. Skogs-och Lantbruksakad. tidskrift Arg.,
95, 257-262.

Andke-s, A. H., AND Navashin, M. S. 1936. Mor-
phological analysis of human chromosomes.
Proc. Maxim Gorkv Med. Genet. Res. Inst.,
Moscow, 4, 506-524."

Aronson, M. M. 1959. New evidence for genetic
activity of the Y chromosome in D. melano-
gasler. Records Genet. Soc. Am., 28, 59.

AsDELL, S. A. 1936. Hermaphroditism in goats.
Dairy Goat J., 14, 3-4.

Bahner, F., Schwarz, G., H.arnden, D. G., J.\cobs,
P. A., HiENz, H. A., AND Walter, K. 1960.
A fertile female with XO sex chromosome con-
stitution. Lancet, 2, 100-101 (7141).

Baker, W. K. 1956. Chromosome segregation
in D. virilis males with multiple sex chromo-
somes. Genetics, 41, 907-914.

Baker, W. K., and Spofford, J. B. 1959. Hetero-
chromatic control of position-effect variega-

tion in Drosophila. Biol. Contr., Univ. Texas
Publ., 5914, 135-154.

Baltzer, F. 1935. Experiments on .sex-develop-
ment in Bonellia. The Collecting Net, 10, 3-8.

Bamber, R. C, and Herdman, E. C. 1932. A re-
port on the progeny of a tortoiseshell male cat,
together with a discussion of his gametic con-
stitution. J. Genet., 26, 115-128.

Barr, M. L. 1959. Sex chromatin and phenotype
in man. Science, 130, 679-685.

Barr, M. L., and Bertram, E. G. 1949. A mor-
phological distinction between neurones of
the male and female, and the behavior of the
nucleolar satellite during accelerated nucleo-
protein synthesis. Nature, London, 163, 676-
677.

Be.\dle, G. W. 1930. Genetical and cytological
studies of Mendelian asynapsis in Zea mays.
Cornell Univ. Agric. Exper. Sta., 129, 3-23.

Beadle, G. W., and Ephrussi, B. 1937. Ovary
transplants in Drosophila melanogaster: meio-
sis and crossing over in superfemales. Proc.
Nat. Acad. Sc, 23, 356-360.

Beerm.^nn, W. 1956. Nuclear differentiation and
functional morphology of chromosomes. Cold
Spring Harbor Symposia Quant. Biol, 21,
217-232.

Bellamy, A. W., and Queal, M. L. 1951. Hetero-
somal inheritance and sex determination in
Platypoecilus maculatus. Genetics, 36, 93-107.

Bemis, W. p., and Wilson, G. B. 1953. A new
hypothesis explaining the genetics of sex de-
termination in Spinacia oleracea L. J. Hered.,
44, 91-95.

Bender, M. A. 1957. X-ray induced chromosome
aberrations in normal human tissue cultures.
Science, 126, 974-975.

Berg, R. L. 1937. The relative frequency of
mutations in different chromosomes of Dro-
sophila melanogaster. II. Sterility mutations.
Genetics, 22, 241-248.

Bergman, S., Reitalu, J., Nowakowski, H., and
Lenz, W. 1960. The chromosomes in two
patients with Klinefelter syndrome. Ann. Hu-
man Genet., 24, 81-88.

Berry, R. O. 1941. Chromosome behavior in
the germ cells and development of llic gonads
in Sciara ocellaris. J. Morphol., 68, 547-576.

Blakeslee, a. F. 1904. Se.xual reproduction in
the mucorineae. Proc. Am. Acad. Sc, 40, 203-
322.

Bloise, W., Bottura, C, de As.sis, L. M., and
Ferrari, I. 1960. Gonadal dysgenesis (Tur-
ner's .syndrome) with male phenotype and XO
chromosomal constitution. Lancet, 2, 1059-60
(7159).

Book, J. A., Fraccvro, M., and Lindsten, J. 1959.
Cvtogenetical ()l).scrvations in Mongolism.
Acta Paediat., 48, 453-468.

Book, J. A., and Santesson, B. 1960. Malfor-
mation syndrome in man associated with
triploidv (69 chromosomes). Lancet, 1, 868
(7129).

Bostian, C. H. 1939. Multiple alleles and sex

FOUNDATIONS FOR SEX

67

determination in Habrobracon. Genetics, 24,
770-776.

Breider, H. 1942. ZW-Mannchen iind WW-
Weibchen bei Platypoecilus maculatus. Biol.
Zentralbl., 62, 183-195.

Bridges, C. B. 1916. Non-disjunction as proof
of the chromosome theory of heredity. Genet-
ics, 1, 1-52; 107-163.

Bridges, C. B. 1921. Triploid intersexes in Dro-
sophila melanogaster. Science, 54, 252-254.

Bridges, C. B. 1922. The origin of variations
in sexual and sex-limited characters. Am.
Naturalist, 56, 51-63.

Bridges, C. B. 1932. The genetics of sex in
Drosophila. In Sex and Internal Secretions,
1st ed., E. Allen, Ed., pp. 55-93. Baltimore:
The Williams & Wilkins Company.

Bridges, C. B. 1939. Cytological and genetic
basis of sex. In Sex and Internal Secretions,
2nd ed., E. Allen, C. H. Danforth, and E. A.
Doisy, Eds., pp. 15-63. Baltimore: The Wil-
liams & Wilkins Company.

Bridges, C. B., and Breh.me, K. S. 1944. The
mutants of Drosophila melanogaster. Carnegie
Inst. Washington Publ., 552, 1-257.

Brosseau, G. E., Jr. 1960. Genetic analysis of
the male fertility factors on the Y chromosome
of Drosophila melanogaster. Genetics, 45,
257-274.

Burdette, W. J. 1940. The effect of artificially
produced tetiaploid regions of the chromo-
somes of Drosophila melanogaster. Univ. Texas
Publ., 4032, 157-163.

Buzzati-Traverso, a. 1941. An extreme case of
sex-ratio in D. bilineata. Drosophila Informa-
tion Service, 14, 49.

Carson, H. L. 1956. A female-producing strain
of D. borealis Patterson. Drosophila Informa-
tion Service, 30, 109-110.

CARSON, H. L., Wheeler, M. R., and Heed, W. B.
1957. A parthenogenetic strain of D. man-
gabeirai Malogolowkin. Univ. Texas Publ.,
5721, 115-122.

Carter, C. O., Hamerton, J. L., Polani, P. E.,
GuNALP, A., AND Weller, S. D. V. 1960.
Chromosome translocation as a cause of
familial mongohsm. Lancet, 2, 678-680 (7152).

Cav.alcanti, a. G. L., and Falcao, D. N. 1954. A
new type of sex-ratio in Drosophila prosaltans
Duda. In Proceedings 9th International
Congress Genetics, Part II, pp. 1233-1235.
(Caryologia, Vol. VI. Suppl.)

Chang, C. Y., and Witschi, E. 1955. Breeding
of sex-reversed males of Xenopus laevis Dau-
din. Proc. Soc. Exper. Biol. & Med., 89, 150-
152.

Chang, C. Y., .and Witschi, E. 1956. Genetic
control and hormonal reversal of sex differenti-
ation in Xenopns. Proc. Soc. Exper. Biol. &
Med., 93, 140-144.

Chu, E. H. Y., and Giles, N. H. 1959. Human
chromosome complements in normal somatic
cells in culture. Am. J. Human Genet., 11,
63-79.

Cohen, B. H., .and Glass, B. 1959. Further ob-

servations on the ABO blood groups and the
sex ratio. Am. J. Human Genet.. 11, 274-278.

Cole, L. J., and Hollander, W. F. 1950. Hybrids
of pigeon by ringdove. Am. Naturalist, 84, 275-
308.

Cooper, K. W. 1952. Studies on spermatogenesis
in Drosophila. Am. Philos. Soc. Yrbk., 1951,
146-147.

Cooper, K. W. 1956. Phenotypic effects of Y
chromosome hyperploidy in Drosophila mel-
anogaster, and their relation to variegation.
Genetics, 41, 242-264.

Cooper, K. W. 1959. Cytogenetic analysis of
major heterogametic elements (especially Xh
and Y) in Drosophila melanogaster and the
theory of "heterochromatin." Chromosoma,
10, 535-588.

Crooke, a. C, and Hayward, M. D. 1960. Mosa-
icism in Klinefelter's Svndrome. Lancet, 1,
1198 (7135).

Crouse, H. V. 1943. Translocations in Sciara;
their bearing on chromosome behavior and
sex determination. Missouri Agric. Exper. Sta.
Res. Bull., 379, 1-75.

Crouse, H. V. 1960a. The nature of the influ-
ence of X-translocations on sex of progeny in
Sciara coprophila. Chromosoma, 11, 146-166.

Crouse, H. V. 1960b. The controlling element
in sex chromosome behavior in Sciara. Genetics,
45, 1429-1443.

Crow, J. F. 1946. The absence of a primary sex
factor on the A" chromosome of Drosophila.
Am. Naturahst, 80, 663-665.

Danforth, C. H. 1932. Genetics of sexual di-
morphism in plumage. In Congress of Genetics,
Ithaca, vol. 2, pp. 34-36.

D.ARLINGTON, C. D., AND Janaki Ammal, E. K. 1945.
Chromosome Atlas oj Cultivated Plants. Lon-
don: George Allen & Unwin, Ltd.

Darlington, C. D., and Haque, A. 1955. Chro-
mosomes of monkevs and men. Nature, Lon-
don, 175, 32.

de Assis, L. M., Epps, D. R., Bottura, C, and
Ferrari, I. 1960. Chromosomal constitution
and nuclear sex of a true hermaphrodite.
Lancet, 2, 129-130 (7142).

Demerec, M. (Editor). 1950. Biology of Dro-
sophila. New York: John Wiley & Sons, Inc.

de Winiavarter, H. 1912. Etudes sur la sperma-
togenese humaine. I. Cellule de Sertoli. II.
Heterochromosome et mitoses de I'epithelium
seminal. Arch. Biol., 27, 91-190.

DE Winiavarter, H., and Oguma, K. 1926. Nou-
velles recherches sur la spermatogenese hu-
maine. Arch. Biol., 36, 99-166.

DOBZH.ANSKY, T., AND BRIDGES, C. B. 1928. The
reproductive system of triploid intersexes in
Drosophila melanogaster. Am. Naturalist 62,
425-434.

DoBZHANSKY, T., AND HoLZ, A. M. 1943. A re-
examination of the problem of manifold ef-
fects of genes in Drosophila melanogaster.
Genetics, 28, 295-303.

DoBZH.ANSKY, T., AND ScHULTZ, J. 1934. The
distribution of sex-factors in the X chromo-

68

BIOLOGIC BASIS OF SEX

some of Drosopliila melanogaster. J. Genet.,
28, 349-386.

DoBZH.^x.SKY, T., AND Spassky, B. 1941. Inter-
sexes in Drosophila pseiidoobscura. Proc. Nat.
Acad. Sc, 27, 556-562.

Dressler, O. 1958. Zytogenetische Untersu-
chungen an diploidem und polyploidem spinat
(Spinacia oleracea L.) unter besonderer Be-
riicksichtigung der Geschlechtsvererbung als
Grundlage einer Imzucht-Heterosis-Zuchtimg.
Ztschr. Pflanzenzucht., 40, 385-424.

DuBois, A. M. 1932. A contribution to the em-
bryology of Sciara (Diptera). J. Morphol.,
54, 161-195.

Eatox, O. N. 1943. An anatomical study of her-
maphroditism in goats. Am. J. Vet. Res., 4,
333-343.

Eaton, 0. N. 1945. The relation between polled
and hermaphroditic characters in dairy goats.
Genetics, 30, 51-61.

Editorial. 1960. A proposed standard system of
nomenclature of human mitotic chromosomes.
Lancet, 1, 1063-1065 (7133).

Edwards, A. W. F. 1958. An analysis of Geissler's
data on the human sex ratio. Ann. Human
Genet., 23, 6-15.

Edwards, J. H., Harnden, D. G., Cameron, A. H.,
Crosse, M., and Wolff, O. H. 1960. A new
trisomic syndrome. Lancet, 1, 787-790 (7128).

Elliott, G. A., Sandler, A., and Rabinowitz, D.

1959. Gonadal dysgenesis in three sisters. J.
Clin. Endocrinol., 19, 995.

Esser, K., and Straub, J. 1958. Genetische Unter-
suchungen an Sordaria macrospora Auersw..
Kompensation und Induktion bei genbedingten
Entwicklungsdefekten. Ztschr. indukt. Ab-
stammungs. Vererb., 89, 729-746.

Evans, H. M., and Swezy, O. 1929. The chro-
mosomes in man. Mem. Univ. California, 9,
1^1.

Fankhauser, G., and Humphrey, R. R. 1959.
The origin of spontaneous hcteroploids in the
progeny of diploid, triploid, and tetraploid
axolotl females. J. Exper. Zool., 142, 379-422.

Ferguson-Smith, M. A. 1960. Cytogenetics in
man. Arch. Int. Med., 105, 627-639.

Ferguson -Smith, M. A., and Johnston, A. W.

1960. Chromosome abnormalities in certain
diseases of man. Ann. Int. Med., 53, 359-71.

Ferguson-Smith, M. A., Johnston, A. W., and
Handmaker, S. D. 1960. Primary amentia
and micro-orchidism associated with an XXXY
sex chromosome constitution. Lancet, 2, 184-
187(7143).

Ferguson-Smith, M. A., Johnston, A. W., and
Weinberg, A. R. 1960. The chromosome
complement in true hermaphroditism. Lancet,
2,126-128(7142).

Ford, C. E. 1960. Human cytogenetics: its pres-
ent place and future possibilities. Am. J. Hu-
man Genet., 12, 104-117.

Ford, C. E., and Hamerton, J. L. 1956. Tlic
chromosomes of man. Natm-e, London, 178,
1020-1023.

Ford, C. E., J.'^cobs, P. A., and Lajtha, L. G.

1958. Human somatic clu-omosomes. Nature.
London, 181, 1565-1568.

Ford, C. E., Jones, K. W., Miller, O. J., Mitt-
wocH, U., Penrose, L. S., Ridler, M., and
Sh.-^piro, a. 1959. The chromosomes in a
patient showing both mongolism and the
Klinefelter syndrome. Lancet, 1, 709-710
(7075).

Ford, C. E., Jones, K. W., Polani, P. E., de
Almeida, J. C, and Briggs, J. H. 1959. A
sex-chromosome anomaly in a case of gonadal
dysgenesis (Turner's syndrome). Lancet, 1,
711-713 (7075).

Ford, C. E., Pol.ani, P. E., Briggs, J. H., and
Bishop, P. M. F. 1959. A presumptive hu-
man XXY/XX mosaic. Nature, London, 183,
1030-1032.

Fraccaro, M., Ikkos, D., Lindsten, J., Luft, R.,
and Kaijser, K. 1960. A new type of chro-
mosomal abnormality in gonadal dysgenesis.
Lancet, 2, 1144 (7160).

Fraccaro, M., Kaijser, K., and Lindsten, J.

1959. Chromosome complement in gonadal
dysgenesis (Turner's syndrome). Lancet, 1,
886(7078).

Fraccaro, M., Kaijser, K., and Lindsten, J. 1960a.
Somatic chromosome complement in continu-
ously cultured cells of two individuals with
gonadal dysgenesis. Ann. Human Genet., 24,
45-61.

Fraccaro, M., K.aijser, K., and Lindsten, J. 1960b.
Chromosomal abnormalities in father and
Mongol child. Lancet, 1, 724-727 (7127).

Fr.accaro, M., Kaijser, K., .\nd Lindsten, J. 1960c.
A child with 49 chromosomes. Lancet, 2, 899-
904 (7156).

Fraccaro, M., .-vnd Lindsten, M. 1960. A child
with 49 chromosomes. Lancet, 2, 1303 (7163).

Eraser, J. H., Campbell, J., M.^cGillivray, R. C,
Boyd, E., .^nd Lennox, B. 1960. The XXX
syndrome frequencj^ among mental defectives
and fertility. Lancet, 2, 626-627 (7151).

Frost, J. N. 1960. The occurrence of partially
fertile triploid metafemales in Drosophila mel-
anogaster. Proc. Nat. Acad. Sc, 46, 47-51.

Fung, S. T. C, and Gowen, J. W. 1957a. The
developmental effect of a sex-limited gene in
Drosophila melanogaster. J. Exper. Zool., 134,
515-532.

Fung, S. T. C, and Gowen, J. W. 1957b. Pig-
ment-inducing potentialities of testes, ovaries
and hermaphroditic (Hr) gonads. J. Exper.
Zool., 135, 5-18.

Fung, S. T., and Gowen, J. W. 1960. Role of
autosome IV in Drosophila melanogaster sex
balance. Rec. Genet. Soc. Am., 29, 70-71.

Gallien, L. 1954. Inversion experimentale du
sexe, sous Taction des hormones sexuelles,
chez le triton Pleurodeles waltlii Michah.
Analyse des consequences genetiques. Bull.
Biol. France et Belgique, 88, 1-51.

Gallien, L. 1955. Descendance unisexuee d'une
fcmelle de Xenopus laevis Daud. ayant subi,
pendent sa phase larvaire. Taction gynogenese

FOUNDATIONS FOR SEX

69

(ill benzoate d'oestradiol. Compt. rend. Acad.
Sc, 240, 913-915.

Gallien, L. 1956. Inversion experimentale du
.sexe chez im Anoure inferieur Xenopus laevis
Daudin. Bull. Biol. France et Belgique, 90,
163-183.

Geissler, a. 1889. Beitrage zur Frage des
Geschlects verhaltnisses der Geborenen. Ztschr.
Sachsischen Statist. Bureaus, 35, 1-24.

Gershenson, S. 1928. A new sex ratio abnor-
mality in Drosophila obscura. Genetics, 13,
488-507.

GoLDSCHMiDT, R. B. 1934. Lymantria. Bibl.
Genet., 11, 1-186.

GoLDSCHMiDT, R. B. 1948. New facts on sex de-
termination in D rosopliila mclanogaster. Proc.
Nat. Acad. Sc, 34, 245-252.

GoLDSCHMiDT, R. B. 1949. The Beaded Minute-
intersexes in Drosophila rnelanogaster Meig. J.
Exper.Zool., 112, 233-301.

GoLDSCHMiDT, R. B. 1951. The maternal effect in
the production of the Bd-M-intersexes in Dro-
sophila rnelanogaster. J. Exper. Zool., 117, 75-
110.

GoLDSCHMiDT, R. B. 1955. Theoretical Genetics.
Berkeley: Uni\ersitj- of California Press.

GoLDSCHMiDT, R. B. 1957. A remarkable action
of the mutant "rudimentary" in Drosophila
rnelanogaster. Proc. Nat. Acad. Sc, 43, 731-736.

GOLDSCHMIDT, R. B., AND Katsuki, K. 1927. Erb-
licher Gynandromorphismus imd somatische
Mosaikbildung bei Bombijx niori. Biol. Zen-
tralbl., 47, 45-54.

GOLDSCHMIDT, R. B., AND Katsuki, K. 1928. Cy-
tologie des erblichen Gynandromorphismus von
Bombyx mori L. Biol.Zentralbl., 48, 685-699.

GoLDSCHMiDT, R. B.,.'\ndK.'\tsuki, K. 1931. Vierte
Mitteilung liber erblichen Gynandromorphis-
mus und somatische Mosaikbildung bei Bom-
byx mori L. Biol. Zentralbl., 51, 58-74.

Goodrich, H. B. 1916. The germ-cells in Ascaris
incurva. J. Exper. Zool., 21, 61-100.

Gordon, M. 1946. Interchanging genetic mech-
anisms for sex determination in fishes under
domestication. J. Hered., 37, 307-320.

Gordon, M. 1947. Genetics of Platypoecilus
maculatus. IV. The sex determining mech-
anism in two wild populations of the Mexican
platyfish. Genetics, 32, 8-17.

Gordon, R. R., O'Gorman, F. J. P., Dewhurst, C.
J., .\ND Blank, C. E. 1960. Chromosome
count in a hermaphrodite with some features
of Klinefelter's svndrome. Lancet, 2, 736-739
(7153).

CrOWEN, J. W. 1928. On the mechanism of chro-
mosome behavior in male and female Dro-
sophila. Proc. Nat. Acad. Sc, 14, 475-477.

GowEN, J. W. 1933. Meiosis as a genetic char-
acter in Drosophila melanogaster. J. Exper.
Zool., 65, 83-106.

GowEN, J. W. 1942. On the genetic basis for
hermaphroditism. Anat. Rec, 84, 458 (Abst.).

GowEN, J. W. 1947. Sex determination in Dro-
sophila melanogaster. Rec Genet. Soc Am.,
16, 33 (Abst.).

GowEN, J. W., AND Fung, S. T. C. 1957. De-
termination of sex through genes in a major
sex locus in Drosophila mekmogaster. He-
redity, 11, 397-402.

GowEN, J. W., AND G.'Vy, E. H. 1933. Eversport-
ing as a function of the Y chromosome in
Drosophila melanogaster. Proc Nat. Acad. Sc,
19, 122-126.

GowEN, M. S., AND GowEN, J. W. 1922. Com-
plete linkage in Drosophila melanogaster. Am.
Naturalist, 56, 286-288.

GowEN, J. W., AND Nelson, R. H. 1942. Prede-
termination of sex. Science, 96, 558-559.

Grumbach, M. M., and Barr, M. L. 1958. Cyto-
logic tests of chromosomal sex in relation to
sexual anomalies in man. Recent Progr. Hor-
mone Res., 14, 255-324.

H.^CHiNOHE, Y., AND Onishi, N. 1952. On the
meiosis of the drone of Honey bee Apis mel-
liUca L. Bull. Nat. Inst. Agric. Sc. (Japan),
(G)3, 83-87.

Hammond, J. 1912. A case of hermaphroditism in
the pig. J. Anat. Physiol., 46, 307-312.

H.'Vnnah, A. M. 1955. The effect of aging the
maternal parent upon the sex ratio in Dro-
sophila melanogaster. Ztschr. indukt. Ab-
stammungs. Vererb., 86, 574-599.

Harnden, D. G. 1960. Abnormalities of chromo-
some constitution and defects of sex develop-
ment. Proc. Roy. Soc. Med., 53, 493-494.

Harnden, D. G., and Armstrong, C. N. 1959.
Chromosomes of a true hermaphrodite. Brit.
M. J., 2, 1287-1288.

Harnden, D. G., and Stewart, J. S. S. 1959.
Chromosomes in a case of pure gonadal
dysgenesis. Brit. M. J., 2, 1285-1287.

Hartmann, M. 1956. Das Wesen und die Grund-
setzlichkeiten des Geschlechts und der Gesch-
lechtsbestimmung im Tier-und Pflanzenreich.
In Die Sexualitdt, 2nd ed. Stuttgart: Gustav
Fischer.

H.\rtmann, M., and Bauer, H. 1953. Allgemeine
Biologic. Stuttgart: Gustav Fischer.

Hasimoto, H. 1929. Some experiments on the
mosaics of Bombyx mori. Japan. J. Genet.,
4(3), 113-114.

Hasimoto, H. 1933. Genetical studies on the
tetraploid female in the silkworm. I. Bull,
sericult. Exper. Sta. Japan, 8, 359-381.

Hasimoto, H. 1934. Genetical studies on the
tetraploid female in the silkworm. Bull,
sericult. Exper. Sta. Japan, 8, 505-523.

Hayward, M. D. 1960. Sex-chromosome mosaic-
ism in man. Heredity, 15, 219-235.

H.ayward, M. D., and Bower, B. D. 1960. Chro-
mosomal trisom\ associated with the Sturge-
Weber syndrome. Lancet, 2, 844-846 (7155).

Henking, H. 1891. LTntersuchungen libsr die
ersten Entwichlungsvorgange in den Eiern
der Insekten II. Uber spermatogenese und
deren Beziehung zur Eintwickelung bei Pyr-
rhocoris apterus M. Ztschr. wiss. Zool., 51,
685-736.

Hixtox, C. W. 1959. A cytological study of w''"
chromosome instability in cleavage mitoses of

70

BIOLOGIC BASIS OF SEX

Drosophila ynelanogaster. Genetics, 44, 923-
933.
HiRSCHHORN, K., Decker, W. H., and Cooper, H.
L. 1960. True hermaphroditism with XY/
XO mosaicism. Lancet, 2, 319-320 (7145).

HOFFENBERG, R., JaCKSOX, W. P. U., AND MULLER,

W. H. 1957. Gonadal dysgenesis with men-
struation. J. Chn. Endocrinol., 17, 902-907.

HoLL.ANDER, W. F. 1944. Mosaic effects in domes-
tic birds. Quart. Rev. Biol., 19, 285-307.

Hollander, W. F., and Gowen, J. W. 1959. A
single-gene antagonism between mother and
fetus in the mouse. Proc. Soc. Exper. Biol. &
Med., 101, 425-428.

Hollander, W. F., Gowen, J. W., and Stadler, J.

1956. A study of 25 gynandromorphic mice
of the Bagg albino strain. Anat. Rec, 124,
223-244.

Hsu, T. C. 1952. Mammalian chromosomes in
vitro. 1. The karyotype of man. J. Heredity,
43, 167-172.

Hsu, T. C, Po.MERAT, C. M., and Moorhe.ad, p. S.

1957. Mammalian chromosomes in vitro.
VIII. Heteroploid transformation in human
cell strain Maves. J. Nat. Cancer Inst., 19,
867-873.

Hughes, W. 1929. The freemartin condition in
swine. Anat. Rec, 41, 213-247.

Humphrey, R. R. 1945. Sex determination in
ambystomid salamanders : a study of the prog-
eny of females experimentally converted into
males. Am. J. Anat., 76, 33-66.

Humphrey, R. R., .\nd Fankhauser, G. 1956.
Structure and functional capacity of the ovaries
of higher polyploid (4N, 5N) in the Mexican
axolotl {Siredon or Amhystoma mexicanum).
J. Morphol., 98, 161-198.

Hungerford, D. A., Donnelly, A. J., Nowell, P.

C, AND Beck, S. 1959. The chromosome
constitution of a human phenotypic intersex.
Am. J. Human Genet., 11, 215-236.

Irwin, M. R., and Cole, L. S. 1936. Immuno-
genetic studies of species and of species hy-
brids from the cross of Columba livia and
Sireptopelia risoria. J. Exper. Zool., 73, 309-
318.

Ishihara, T. 1956. Cytological studies of tor-
toiseslu'll male cats. Cytologia, 21, 391-398.

Jacobs, P. A., Baikie, A. G., Court Brown, W. M.,
Forrest, H. Roy, J. R., STEW^'\RT, J. S. S., and
Lennox, B. 1959. Chromosomal sex in the
syndrome of testicular feminisation. Lancet,
2,591-592(7103).

J.\coBs, P. A., B.\iKiE, A. G., Court Brown, W. M.,
MacGrkgok, T. N., Maclean, N., and Harnden,

D. G. 1959. P]vidence for the existence of
the human "super female." Lancet, 2, 423-
425 (7100).

J.\coBS, P. A., Baikie, X. C;., Court Brown, W. M.,
AND Strong, J. A. 1959. The somatic chro-
mosomes in mongolism. Lancet, 1, 710 (7075).

Jacobs, P. A., Harnden, D. G., Court Brown, W.
M., Goldstein, J., Close, H. G., MacGregor,
T. N., Maclean, N., and Strong, J. A. 1960.
Abnormalities involving the A^-chromosome
in women. Lancet, 1, 1213-1216 (7136).

Jacobs, P. A., and Strong, J. A. 1959. A case of
human intersexuality having a possible XXY
sex-determining mechanism. Nature, London,
183, 302-303.

Jacobsex, p. 1957. The sex chromosomes in
Humulus. Hereditas, 43, 357-370.

Janick, J. 1955. Environmental influences on
sex expression in monoecious lines of spinach.
Proc. Am. Soc. Horticult. Sc, 65, 416-422.

Janick, J., and Ellis, J. R. 1959. Identification
of the sex chromosome in Sjyinacia oleracea.
(Abst.) Rec. Genet. Soc. Am., 28, 78.

Janick, J., Mahoney, D. L., and Pfahler, P. L.
1959. The trisomies of Spinacia oleracea. J.
Hered., 50, 47-50.

Jaxick, J., axd Stevexson, E. C. 1955a. The
effects of polyploidy on sex expression in
spinach. J. Hered., 46, 151-156.

Janick, J., and Stevenson, E. C. 1955b. Ge-
netics of the monoecious character in spinach.
Genetics, 40, 429-437.

Jennings, H. S. 1939. Genetics of Paramecium
hursaria. I. Mating types and groups, their
interrelations and distribution; mating be-
havior and self sterility. Genetics, 24, 202-
233.

Johnston, E. F., Zeller, J. H., and Cantwell, G.
1958. Sex anomalies in swine. J. Hered., 49,
255-261.

Kang, Y. S., and Cho, W. K. 1959a. Data on the
biology of Korean populations. Human Biol.,
31, 244-251.

K.\ng, Y. S., and Cho, W. K. 1959b. The sex ratio
at birth of the Korean population. Eugenics
Quart., 6, 187-195.

Kawaguchi, E. 1928. Zytologische Untersuchun-
gen am Seidenspinner und seine Verwandten.
I. Gametogenese von Bombyx mori L. und
Bom,byx rnandarina M. und ihre Bastarde.
Ztschr. Zellforsch., 7, 519-533.

Keller, K., and Tandler, J. 1916. Uber das Ver-
halten der Eihaeute bei der Zwellingstriichtig-
keit des Rindes. Wien. tieriirztl. Monatsschr., 3,
513-526.

KiHARA, H. 1925. Chromosomes of Rinncx accto-
sella L. Botan. Mag., Tokyo, 39, 353-360.

KiHARA, H., AND Ono, T. 1923. Cytological studies
on Rumex L. I. Chromosomes of Rumex
acetosa L. Botan. Mag., Tokyo, 37, 84-91.

KiHARA, H., AXD Yamamoto, Y. 1935. Chromo-
somenverhiiltnisse bei Aucidia chinensis Benth.
Agric. & Horticult., Japan, 10, 2485-2496.

Kimball, R. F. 1939. Mating tvpes in Euplotes.
Am. Naturalist, 73, 451-456.

KoDANi, M. 1958. Tlie supernumerary chromo-
some of man. .\in. J. Human Genet.. 10, 125-
140.

KoLLER, P. C. 1937. The genetical and mechani-
cal properties of sex chromo-somes. III. Man.
Proc. Roy. Soc. Edinburgh, 57, 194-214.

KoMAi, T. 1952. On the origin of the tortoiseshell
male cat — a correction. Proc. Japan Acad.,
Tokyo, 28, 150-155.

KoMAi, T., AXD Ishihara, T. 1956. On the origin
of the male tortoiseshell cat. J. Hered., 47, 287-
291.

FOUNDATIONS FOR SEX

71

IvoNDO, K. 1952. Studies on intersexuality in
milk goats. Japan. J. Genet., 27, 131-141.

KoNDO, K. 1955a. Studies on intersexuality in
milk goats. III. On the homozygous reces-
sive males and the methods of eliminating
intersexual gene. Bull. Nat. Inst. Agric. Sc,
Japan, ser. G, 10, 107-117.

KoNDO, K. 1955b. The frequency of occurrence of
intersexes in milk goats. Japan. J. Genet., 30,
139-146.

IvRivsHENKO, J. 1959. New evidence for the
homology of the short euchromatic elements
of the X and Y chromosomes of Drosophila
busckii with the microchromosome of Dro-
sophila melayiogaster . Genetics, 44, 1027-1040.

Lebedeff, G. a. 1934. Genetics of liermaphro-
ditism in Drosophila virilis. Proc. Nat. Acad.
Sc, 20, 613-616.

Lejeune, J., Gautier, M., and Turpin, R. 1959a.
Les chromosomes humaines en culture de tis-
sus. Compt. rend. Acad. Sc, 248, 602-603.

Lejeune, J., Gautier, M., and Turpin, R. 1959b.
Etude des chromosomes somaticjues de neuf
enfants mongoliens. Compt. rend. Acad. Sc,
248, 1721-1722.

Lejeune, J., Turpin, R., and G.^utier, M. 1959a.
Le Mongolisme, premier exemple d 'aberration
autosomique humaine. Ann. Genet., 2, 41-19.

Lejeune, J., Turpin, R., and Gautier, M. 1959b.
Le mongolisme, maladie chromosomique (tri-
somie). Bull. Acad. Nat. Med., 143, 256-265.

Lewis, D. 1942. Evolution of sex in flowering
plants. Biol. Rev., Cambridge Phil. Soc, 17,
46-67.

L'Heritier, p. 1951. The CO2 sensitivity prob-
lem in Dro.sophila. Cold Spring Harbor Sym-
posia Quant. Biol., 16, 99-112.

L'Heritier, P. 1955. Les \'irus integres et I'unite
cellulaire. Ann. Biol., 31, 481-496.

LiLLiE, F. R. 1917. The freemartin; a study of
the action of sex hormones in the foetal life of
cattle. J. Exper. Zool., 23, 371.

Love, A. 1944. Cj'togenetic studies on Rumex
subgenus Acetosella. Hereditas, 30, 1-136.

Love, A. 1957. Sex determination in Rumex.
Proc Genet. Soc. Canada, 2, 31-36.

Love, A., and S.\rk.ar, N. 1956. Cytotaxonomy
and sex determination of Rumex paucifolius.
Canad. J. Botany, 34, 261-268.

LuBS, H. A., Jr., Vilar, O., and Bergenst.al, D.
M. 1959. Familial male pseudohermaphro-
ditism with labial testes and partial feminiza-
tion: endocrine studies and genetic aspects. J.
Clin. Endocrinol., 19, 1110-1120.

M.ackensen, O. 1951. Viability and sex determi-
nation in the honev bee {Apis ynellijcra L.).
Genetics, 36, 500-509.

Mackensen, O. 1955. Further studies on a lethal
series in the honey bee. J. Hered., 46, 72-74.

Magni, G. E. 1952. Sex-ratio in Drosophila bi-
fasciata. 1st. Lombardi di Scienze e Lettere,
85,391^11.

Magni, G. E. 1953. 'Sex-ratio,' a non-Mendelian
character in Drosophila bifasciata. Nature,
London, 172,81.

Magni, G. E. 1954. Thermic cure of cytoplasmic

sex-ratio in Drosophila bifasciata. In Proceed-
ings 9th International Congress of Genetics,
Bellagio, pp. 1213-1216.

Magni, G. E. 1957. Analisi della trasmissione
delle particelle citoplasmatiche "sex-ratio" in
Drosophila bifasciata. Atti. 3 Riun. A.G.I.,
Ric Sci. 27, 67-70. (Suppl.)

Making, S. 1950. Contribution of the sex chro-
mosome in an intersex goat. A preliminary
note. Iden-Sogo-Kenkya, 1, 1-3.

Making, S. 1951. Chromosome Numbers in Ani-
mals, 2nd ed. Ames: Iowa state College Press.

Making, S., and Sa.saki, M. 1959. On the chro-
mosome nimiber of man. Proc. Japan Acad ,
35, 99-104.

Malogolowkin, C. 1958. Maternally inherited
sex-ratio conditions in Drosophila unllistoni
and Drosophila paulistorum. Genetics, 43,
276-286.

Malogolowkin, C, and Pgulson, D. F. 1957. In-
fective transfer of maternally inherited abnor-
mal sex -ratio in Drosophila willistoni. Science,
125, 32.

Malogolowkin, C, Poulson, D. F., and Wright,
E. Y. 1959. Experimental transfer of mater-
nally inherited abnormal sex-ratio in Droso-
phila uiillistoni. Genetics, 44, 59-74.

M.\mpell, K. 1941. Female sterility in interra-
cial hj'brids of Drosophila pseudoobscura.
Proc Nat. Acad. Sc, 27, 337-341.

McClung, C. E. 1902. The accessory chromo-
some — sex determinant? Biol. Bull., 3, 43-84.

Metz, C. W. 1938. Chromosome behavior, in-
heritance and sex determination in Sciara. Am.
Naturalist, 72, 485-520.

Metz, C. W. 1959. Chromosome behavior and
cell lineage in triploid and mosaic salivary
glands of species hvbrids in Sciara. Chromo-
soma, 10, 515-534.

MiNoucHi, O., and Ohta, T. 1934. On the num-
ber of chromosomes and the type of sex chro-
mosomes in man. Cytologia, 5, 472^90.

Mittwoch, U. 1952. The chromosome comple-
ment in a mongolian imbecile. xA.nn. Eugenics,
London, 17, 37.

Morgan, L. V. 1925. Polyploidy in Drosophila
melanogaster with two attached A' chromo-
somes. Genetics, 10, 148-178.

Morgan, L. V. 1929. Contributions to the ge-
netics of Drosophila simulans and Drosophila
ynelanogaster . VIII. Composites of Drosoph-
ila melanogaster. Publ. Carnegie Inst. Wash-
ington, 399, 223-296.

Morgan, T. H. 1914. Heredity and Sex, 2nd ed.
New York : Columbia University Press.

Morgan, T. H., and Bridges, C. B. 1919. Contri-
butions to the genetics of Drosophila melano-
gaster. I. The origin of gynandromorphs.
Publ. Carnegie Inst. Washington, 278, 1-122.

Morgan, T. H., Bridges, C. B., and Sturtevant,
A. H. 1925. The genetics of Drosophila.
Bibliog. Genet., 2, 1-262.

MoRiWAKi, D., .\ND Kitag.^wa, 0. 1957. Sex-
ratio-J in Drosophila bifasciata. I. A pre-
liminary note. Japan. J. Genet., 32, 208-210.

Moses, M. S., and Metz, C. W. 1928. Evidence

72

BIOLOGIC BASIS OF SEX

that the female is responsible for the sex ratio
in Sciara (Diptera). Proc. Nat. Acad. Sc, 14,
928-930.

MULD.4L, S., AND OcKEY, C. H. 1960. The "double
male," a new chromosome constitution in
Klinefelter syndrome. Lancet, 2, 492 (7148).

MuRDY, W. H., AND Carson, H. L. 1959. Parthe-
nogenesis in Drosophila mangabeirai Malog.
Am. Naturalist, 93, 355-363.

Nelson, L., Ferrari, I., and Bottura, C. 1960.
Chromosomal constitution in a case of Kline-
felter's syndrome. Lancet, 2, 319 (7145).

Neuhaus, M. E. 1939. A cytogenetic study of
the i' chromosome of Drosophila melanogas-
ter. J. Genet., 37, 229-254.

Newby, W. W. 1942. A study of intersexes pro-
duced by a dominant mutation in Drosophila
virilis. Blanco Stock. Univ. Texas Publ., 4228,
113-145.

XiLSsox, I. M., Bergman, S. Reit.\lu, J., and W.^l-
DENSTROM, J. 1959. Haemophilia A in a "girl"
with male sex-chromatin pattern. Lancet, 2,
264-266 (7097)

NoviTSKi, E. 1947. Genetic analysis of an anom-
alous sex ratio condition in Drosophila af-
finis. Genetics, 32, 526-534.

NoviTSKi, E., AND Ki.MBALL, A. W. 1958. Birth
order, parental ages, and sex of offspring. Am.
J. Human Genet., 10, 268-275.

NowAKOwsKi, H., Lenz, W., and Parada, J. 1959.
Diskrepenz zwisclien Chromatinbefund und
genet ischem Gcschlecht l^eim Klinefeltersyn-
drom. Acta Endocrinol., 30, 296-320.

Ono, T. 1930. Further investigations on the cy-
tologv of Rumex VI-VIII. Botan. Mag., Tokyo,
44, 168-176.

Ono, T. 1935. Chromosomen und sexualitat von
Rumex acetosa. Sc. Repts., Tohoku Univ.
Biol., ser. IV, 10, 41-210.

Ono, T. 1940. Vorkommen der triploiden inter-
sexe bei Humidus japnnicns. Japan. J. Botan.
& Zool., 8, 1632-1634.

OvERZiER, Claus. 1955. Hermaphroditismus ve-
rus. Acta Endocrinol., 20, 63-80.

P.AiNTER, T. S. 1921. The i' chromosome in mam-
mals. Science, 53, 503-504.

Painter, T. S. 1923. Studies in mammalian
spermatogenesis. II. The spermatogenesis of
man. J. Exper. Zool., 37, 291-334.

P.\TAU, K. 1960. The identification of individual
chromosomes, especiallv in man. Am. J. Human
r;en(-t., 12, 250-276.

P.'VTAU, K., Smith, D. W., Therman, E., Inhorn, S.
L., .-vno Wagner, H. P. 1960. Multiple con-
genital anomaly caused by an extra autosome.
Lancet, 1,790-793 (7128)."

Patterson, J. T. 1938. Sex differentiation. Al)cr-
rant forms in Drosophila and sex differentia-
tion. Am. Natm-alist, 72, 193-206.

P.\tterson, J. T., Stone, W. S., and Bedechek, S.
1937. Further studies on X chromosome bal-
ance in Drosophila. Genetics, 22, 407-426.

Penrose, L. S., Ellis, J. R., and Delhanty, J. D.
A. 1960. Chromosomal translocations in
mongolism and in normal relatives. Lancet, 2,
409(7147).

Pipkin, S. B. 1940. Multiple sex genes in the X
chromosome of Drosophila melanogaster.
Univ. Texas Publ., 4032, 126-156.

Pipkin, S. B. 1941. Intersex modifying genes in
wild strains of Drosophila melanogaster. Ge-
netics, 26, 164-165.

Pipkin, S. B. 1942. Intersex modifying genes in
wild strains of Drosophila melanogaster. Ge-
netics, 27, 286-298.

Pipkin, S. B. 1947. A search for sex genes in the
second chromosome of Drosophila melanogas-
ter using the triploid method. Genetics, 32,
592-607.

Pipkin, S. B. 1959. Sex balance in Drosophila
melanogaster: Aneuploidy of short regions of
chromosome 3, using the triploid method.
Univ. Texas Publ., 5914, 69-88.

Pipkin, S. B. 1960. Sex balance in Drosophila
melanogaster : Aneuploidy of long regions of
chromosome 3, using the triploid method. Ge-
netics, 45, 1205-1216.

PoLANi, P. E., Bishop, P. M. F., Ferguson-Smith,
M. A., Lennox, B., Stew.\rt, J. S. S., and
Prader, a. 1958. Colour vision studies and
the A^ chromosome constitution of patients
with Klinefelter's syndrome. Natuie, London,
182, 1092-1093.

PoLANi, P. E., Briggs, J. H., Ford, C. E., Clarke,
O. M,, AND Berg, J. M. 1960. A Mongol girl
with 46 chromosomes. Lancet, 1, 721-724
(7127).

PoLANI, P. E., LeSSOF, M. H., AND BiSHOP, P. M. F.

1956. Colour-blindness in "ovarian agenesis"
(gonadal dysplasia). Lancet, 2, 118-120 (6934).

PooLE, H. K., AND Olsen, M. W. 1958. Incidence
of parthenogenetic development in eggs laid
bv 3 strains of dark Cornish chickens. Proc.
Soc. Exper. Biol. & Med., 97, 477-478.

PouLSON, D. F. 1940. The effects of certain X
chromosome deficiencies on the development
of D. melanogaster. J. Exper. Zool., 83, 271-
326.

Price, J. B., Jr., 1949. The location of a sex
limited factor, Ix", in D. virilis. Univ. Texas
Publ., 4920, 22-23.

Puck, T. T. 1958. Genetic studies on somatic
mammalian cells in vitio. J. Cell. Physiol.
Suppl. 1, 52, 287-311.

Puck, T. T., Robinson, A., .and Tjio, J. H. 1960.
Familial primary amenorrhea due to testicular
feminization: a human gene affecting sex dif-
ferentiation. Proc. Soc. Exper. Biol. & Med.,
103, 192-196.

R.\per, J. W. 1960. Control of SOX in fungi. Am. J.
Botan., 47, 794-808.

R.\smussen, I. E. 1957. Tentativi di infezione
sperimentale con il carattere "sex ratio" in
Drosophila bifasciata. Atti III Rium. A. Genet.
Ital. T>a Ric. Sc, Suppl., 27, 71-74.

Ravix, A. \V. 1960. The origin of bacterial species,
genetic recombination and factors limiting it
between bacterial populations. Bact. Rev., 24,
201-221.

Redfield, H. 1926. The maternal inheritance of
a sex-limited lethal effect in Drosophila mel-
anogaster. Genetics, 11, 482-502.

FOUNDATIONS FOR SEX

73

Reynolds, J. P. 1938. Sex determination in a
bisexual strain of Sciara coprophila Lintner.
Genetics, 23, 203-220.

Rick, C. M., and Hanna, G. C. 1943. Deter-
mination of sex in Asparagus officinalis L.
Am. J.Botan., 30, 711-714.

Ris, H., AND Kerr, W. E. 1952. Sex determina-
tion in the honey-bee. Evolution, New York,
6, 444-445.

RoTHENBUHLER, W. C. 1955. Hereditary aspects
of gynandromorph occurrence in honey bees
(.4/j^s mellifica). Iowa State Coll. J. Sc, 29,
487-488.

RoTHENBUHLER, W. C. 1957. Diploid male tissue
as new evidence on sex determination in honey
bees {Apis mellifera). J. Hered., 48, 160-168.

RoTHENBUHLER, W. C. 1958. Progress and prob-
lems in the analysis of gynandromorphic
honey bees. In Proceedings 10th International
Congress of Entomology, vol. 2, pp. 867-874.

RoTHENBUHLER, W. C., GoWEN, J. W., AND PaRK,

O. W. 1952. Androgenesis with zygogenesis
in gynandromorphic honey bees {Apis mel-
lifera L.). Science, 115, 637-638.
Russell, W. L., Russell, L. B., and Gow^er, J. S.

1959. Exceptional inheritance of a sex-linked
gene in the mouse explained on the basis that
X/0 sex-chromosome constitution is female.
Proc. Nat. Acad. Sc, 45, 554-560.

Russell, W. T. 1936. Statistical study of the sex

ratio at birth. J. Hyg., 36, 381-401.
Sandberg, a. a., Crosswhite, L. H. and Gordy, E.

1960. Trisomy of a large chromosome. Associ-
ation with mental retardation. J. A. M. A.,
174, 221-225.

Sandberg, A. A., Koepf, G. F., Crosswhite, L. H.,
AND Hauschka, T. 1960. The chromosome
constitution of human marrow in various de-
velopmental and blood disorders. Am. J. Hu-
man Genet., 12, 231-249.

Sanderson, A. R., and Hall, D. W. 1948. The
cytology of the honey bee, Apis mellifica L.
Nature^ London, 162, 34-35.

Sanderson, A. R., and Hall, D. W. 1951. Sex in
the honey bee. Endeavour, 10, 33-39.

Sasaki, M., and Making, S. 1960. The chromo-
somal constitution of a human hermaphrodite.
Texas Rep. Biol. & Med. 18, 493-500.

S.\tina, S., and Blakeslee, a. F. 1925. Studies on
biochemical differences between (-I-) and ( — )
sexes in mucors. Proc. Nat. Acad. Sc, 11, 528-
534.

S.\to, H. 1942. Cytological studies of the eggs
of silkworm exposed to high temperature. I,
II. Sanshigaku-zassi, 14(1).

Schmieder, R. G., and Whiting, P. W. 1947. Re-
productive economy in the chalcidoid wasp
Melittobia. Genetics, 32, 29-37.

Schrader, F. 1928. Die Geschlechtschrornosomen
{The sex chromosomes). Berlin: Gebriider
Borntraeger.

Schrader, F., and Sturtevant, A. H. 1923. A
note on the theorv of sex determination. Am.
Naturalist, 57, 379-381.

Shen, T. H. 1932. Zytologische Untersuchungen
iiber sterilitat bei Mannchen von Drosophila

mchuiogasler und bei F. Mannchen der Kreu-
zung zwischen D. simulans Werbchen und D.
)uelanogaster Mannchen. Ztschr. Zellforsch.,
15, 547-580.

Shiwago, p. I., AND Andres, A. H. 1932. Die
Geschlechtschromosomen in der spermato-
genese des Menschen. Ztschr. Zellforsch., 16,
413-431.

Smith, B. W. 1955. Sex chromosomes and natu-
ral polyploidy in dioecious Rumex. J. Hered.,
46, 226-232.

Smith, D. W., Patau, K., Therivun, E., and Inhorn,
S. L. 1960. A new autosomal trisomy syn-
drome ; multiple congenital anomalies caused
bv an extra chromosome. J. Pediat., 57, 338-
345.

Smith, S. G. 1955. Cytogenetics of obligatory
parthenogenesis. Canad. EntomoL, 87, 131-
135.

Smith-White, S. 1955. The life history and
genetic system of Leucopogon juniperinus.
Heredity, 9, 79-91.

Sneep, J. 1953. The significance of andromonoecy
for the breeding of Asparagus officinalis L.
Euphytica, 2, 89-95.

Snell, G. D. 1935. The determination of sex in
Habrobracon. Proc Nat. Acad. Sc, 21, 446-
453.

SoNNEBORN, T. M. 1937. Sex, sex inheritance and
sex determination in Paramecium aurelia.
Proc. Nat. Acad. Sc, 23, 378-385.

SoNNEBORN, T. M. 1947. Recent advances in the
genetics of Paramecium and Euplotes. Ad-
vances Genet., 1, 264-358.

SoNNEBORN, T. M. 1949. Ciliated protozoa: Cyto-
genetics, genetics, and evolution. Ann. Rev.
Microbiol., 3, 55-80.

SoNNEBORN, T. M. 1957. Breeding systems, re-
productive methods, and species in protozoa.
In The Species Problem, pp. 163-324. Wash-
ington : American Association for the Advance-
ment of Science.

Sprackling, L. S. 1960. The chromosome com-
plement of the developing eggs produced by
Drosophila parthenogenetica Stalker virgin
females. Genetics, 45, 243-256.

Spurway, H., and Haldane, J. B. S. 1954. Ge-
netics and cytology of Drosophila suhohscura.
IX. An autosomal recessi\-e mutant transform-
ing homogametic zygotes into intersexes. J.
Genet., 52, 208-225. "

Stalker, H. D. 1942. Triploid intersexuality in
D. americana Spencer. Genetics, 27, 504-518.

Stalker, H. D. 1954. Parthenogenesis in Dro-
sophila. Genetics, 39, 4-34.

Stalker, H. D. 1956a. On the evolution of par-
thenogenesis in Lonchoptera (Diptera). Evo-
lution, 10, 34S-359.

Stalker, H. D. 1956b. Selection within a uni-
sexual strain of Drosophila (Abstr.). Rec.
Genet. Soc Am., 25, 662, Genetics, 41, 662.

Stern, C. 1929. Untersuchungen iiber Aberra-
tionen des Y chromosoms \-on Drosophila
melanogaster. Ztschr. indukt. Abstammungs.
Vererb., 51, 253-353.

Stern, C. 1936. Somatic crossing-over and segre-

74

BIOLOGIC BASIS OF SEX

gation in Drosophila melanogaster. Genetics,
21, 625-730.

Stern, C. 1959a. Color-blindness in Klinefelter's
syndrome. Nature, London, 183, 1452-1453.

Stern, C. 1959b. Use of the term "superfemale".
Lancet, 2, 1088 (7111).

Stern, C, and Hadorn, E. 1938. The determina-
tion of sterility in Drosophila males without a
complete Y chromosome. Am. Naturalist, 72,
42-52.

Sternberg, W. H., and Ivloepfer, H. W. 1960.
Genetic and pathologic study of "simulant"
females (testicular feminization syndrome).
Am. Soc. Human Genet., 11-12 (Abstr.).

Stewart, J. S. S. 1959. Testicular feminization
and colour-blindness. Lancet, 2, 592-594
(7103).

Stew.art, J. S. S. 1960a. Gonadal dj'sgenesis.
The genetic significance of unusual variants.
Acta Endocrinol., 33, 89.

Stewart, J. S. S. 1960b. Genetic mechanisms
in Inmian intersexes. Lancet, 1, 825-826
(7128).

Stewart, J. S. S., and Sanderson, A. R. 1960.
Fertility and ohgophrenia in an apparent
triplo-X-female. Lancet, 2, 21-22 (7140).

Stone, W. S. 1942. The Ix" factor and sex de-
termination. Univ. Texas Publ., 4228, 146-
152.

Sturtevant, a. H. 1920a. Intersexes in Drosoph-
ila simulans. Science, 51, 325-327.

Sturtevant, A. H. 1920b. Genetic studies on
Drosophila simulans. I. Introduction. Hy-
brids with Drosophila melanogaster. Genet-
ics, 5, 488-500.

Sturtevant, A. H. 1921. Genetic studies on
Drosophila simulans. III. Autosomal genes;
general discussion. Genetics, 6, 179-207.

Sturtevant, A. H. 1923. Inheritance of direc-
tion of coiling in Limnaea. Science, 58, 269-
270.

Sturtevant, A. H. 1945. A gene in Drosophila
melanogaster that transforms females into
males. Genetics, 30, 297-299.

Sturtevant, A. H. 1946. Intersexes dependent
on a maternal effect in hybrids between
Drosophila repleta and D. neorepleta. Proc.
Nat. Acad. Sc, 32, 84-87.

Sturtevant, A. H. 1949. The Beaded Minute
combination and sex determination in Dro-
sophila. Proc. Nat. Acad. Sc, 35, 311-313.

Sturtevant, A. H. 1956. A highly specific com-
plementary lethal system in Drosophila mel-
anogaster' Genetics, 41, 118-123.

Sturtevant, A. H., and Dobzhanskv, T. 1936.
Geographical distribution and cytology of
"sex ratio" in Drosophila pseudoobscura and
related species. Genetics, 21, 473-490.

Suomal.\inen, E. 1950. Parthenogenesis in ani-
mals. Advance Genet., 3, 193-253.

SuoMALAiNEN, E. 1954. The cytology of the par-
thenogenetic curculionids of Switzerland.
Chromosoma, 6, 627-655.

Syverton, J. T. 1957. Comparative studies of
normal and malignant human cells in continu-
ous culture. In Cellular Biology, Nucleic

Acids and Viruses. New York Acad. Sc,
special publ., 5, 331-340.

Taber, S., 3rd. 1955. Evidence of binucleate eggs
in the honey bee. J. Hered., 46, 156.

Tanaka, Y. 1952. Genetics of the Bombyx mori.
Tokyo : Shokado Company.

T.^naka, Y. 1953. Genetics of the silkworm,
Bombyx mori. Advance Genet., 5, 239-317.

T.\YL0R, J. H. 1957. Tlie time and mode of
duplication of chromosomes. Am. Naturalist,
91,209-221.

T.\ziMA, Y. 1941. Larval striations as a simple
means for the differentiation of sex. J. seri-
cult. Sc. Japan., 12(3).

T.'VziMA, Y. 1952. Sex determination. In Genetics
of the Bombyx mori, Y. Tanaka, Ed., pp. 351-
372. Tokyo : Shokado Company.

Thompson, D. H. 1920. A new type of sex-linked
lethal mutant in Drosophila. Anat. Rec, 20,
215.

Tjio, J. H., AND Levan a. 1956. The chromo-
some number of man. Hereditas, 42, 1-6.

Tjio, J. H., and Puck, T. T. 1958a. Genetics of
somatic mammalian cells. II. Chromosomal
constitution of cells in tissue culture. J. Ex-
per. Med., 108, 259-268.

Tjio, J. H., and Puck, T. T. 1958b. The so-
matic chromosomes of man. Proc. Nat. Acad.
Sc, 44, 1229-1237.

Tjio, J. H., Puck, T. T., and Robinson, A. 1959.
The somatic chromosomal constitution of
some human subjects with genetic defects.
Proc. Nat. Acad. Sc, 45, 1008-1016.

Tokunaga, C. 1958. The Y chromosome in sex
determination of Aphiochaeta xanthina. In
Proceedings 10th International Congress of
Genetics, Montreal, Vol. 2, pp. 295-296 (Abstr.).

Tucker, K. W. 1958. Automictic parthenogene-
sis in the honey bee. Genetics, 43, 299-316.

TuRPiN, R., Lejeune, J., Lafourcade, J., AND Gau-
tier, M. 1959. Aberrations chromosomiques
et maladies humaines. La polydysspondylie a
45 chromosomes. Compt. rend. Acad. Sc, 248,
3636-3638.

Warmke, H. E. 1946. Sex determination and sex
balance in Melandrium. Am. J. Botan., 33,
648-660.

W.ARMKE, H. E., AND Blakeslee, A. F. 1939. Sex
mechanisms in polyploids of Melandrium.
Science, 89, 391-392.

Warmke, H. E., D.widson, H., and LeClerc, G.
1945. Polyploidy investigations. In Annual
Report, Director of Department of Genetics,
Carnegie Institute Year Book, vol. 44, pp. 113-
115.

Weir, J. A. 1958. Sex ratio related to sperm
source in mice. J. Hered., 49, 223-227.

Wei.shons, W. J., AND Russell, L. B. 1959. The
}' chromosome as the bearer of male determin-
ing factors in the mouse. Proc. Nat. Acad. Sc,
45, 560-566.

Westergaard, M. 1940. Studies on cytology and
sex determination in polyploid forms of Me-
landrium album. Dansk Botau. .Vik., 10, 1-
131.

We.stergaard, M. 1946a. Structural changes of

FOUNDATIONS FOR SEX

the }' chromosome in tlie offspring of poly-
ploid Melandrium. Hereditas, 32, 60-64.

Westergaard, M. 194613. Aberrant }' chromo-
somes and sex expression in Melandrium al-
bum. Hereditas, 32, 419^43.

WESTERcaARD, M. 1948. The relation between
chromosome constitution and sex in the off-
spring of triploid Melandrium. Hereditas,
34, 257-279.

Westergaard, M. 1953. Uber den Mechanismus
der Geschlechtsbestimmung bei Melandrium
album. Naturwissenschaften, 40, 253-260.

Westergaard, M. 1958. The mechanism of sex
determination in dioecious flowering plants.
Advance Genet., 9, 217-281.

White, M. J. D. 1954. Animal Cytology and
Evolution, 2nd ed. Cambridge: Cambridge
University Press.

Whiting, P. W. 1933a. Selective fertilization
and sex determination in Hymenoptera. Sci-
ence, 78, 537-538.

Whiting, P. W. 1933b. Sex determination in
Hymenoptera. Biol. Bull., 65, 357-358.

Whiting, P. W. 1935. Sex determination in bees
and wasps. J. Hered., 26, 263-278.

Whiting, P. W. 1943a. Multiple alleles in com-
plementary sex determination of Habrobracon.
Genetics, 28, 365-382.

Whiting, P. W. 1943b. Androgenesis in the
parasitic wasp Habrobracon. J. Hered., 34,
355-366.

Whiting, P. W. 1958. Diploid males and triploid
females in Habrobracon and Mormoniella.
Bull. A. Southeastern Biol., 5, 16.

Whiting, P. W., and Whiting, A. R. 1925.
Diploid males from fertilized eggs in Habro-
bracon. Science, 63, 437-438.

WiESE, L. 1960. Die diplogenotypische Gesch-
lechtsbestimmung. Fortschr. Zool., 12, 295-
335.

Wilson, E. B. 1928. The Cell in Development
and Heredity, 3rd ed. New Yoik: The
Macmillan Company.

WiNGE, O. 1922. A pecuhar mode of inheritance
and its cytological explanation. Compt. rend,
tr^.v. lab.Carlsberg, 17, 1-9.

WiNGE, 0. 1934. The experimental alteration of
sex chromosomes into autosomes and vice

versa, as illustrated by Lebistes. Compt. rend,
trav. lab. Carlsberg, ser. physioL, 21, 1-49.

WiNGE, 0., AND DiTLEVSEN, E. 1948. Colour in-
heritance and sex determination in Lebistes.
Compt. rend. trav. lab. Carlsberg, ser. phvsiol.,
24, 227-248.

WiTSCHi, E., Nelson, W. O., and Segal, 8. J. 1957.
Genetic developmental and hormonal aspects
of gonadal dysgenesis and sex inversion in man.
J. Clin. Endocrinol., 17, 737-753.

Wolf, E. B. 1960. Eine Nachuntersuchung zur
Cytologie der Honigbiene {Apis m,ellifica L.).
Zool. Beitr., neue folge, 5, 373-391.

Yamamoto, T.-O. 1953. Artificially induced sex-
reversal in genotypic males of the medaka
{Oryzim latipes). J. Exper. Zool., 123, 571-
594.

Yamamoto, T.-O. 1957. Esterone-induced inter-
sex of genetic male in the medaka, Oryzias
latipes. J. Fac. Sc. Hokkaido Univ., ser. VI,
Zool., 13, 440-444.

Yaalamoto, T.-O. 1959a. A further study on in-
duction of functional sex-re\'ersal in geno-
typic males of the medaka {Oryzias latipes)
and progenies of sex-reversals. Genetics, 44,
739-757.

Yam.amoto, T.-O. 1959b. The effect of estrone
dosage level upon the percentage of sex-
reversals in genetic male {XY) of the
medaka (Oryzias latipes). J. Exper. Zool., 141,
133-153.

Yamamoto, Y. 1938. Karyogenetische I'ntersu-
chungen bei der Gattung Rumex. VI. Ge-
sclilochtsbestimmung bei eu- und aneuploiden
Pflanzen von Rumex acetosa L. Mem. Coll.
Agric. Kyoto Imp. Univ., 43, 1-59.

Yao, T., and Olsen, M. W. 1955. Microscopic
observations of parthenogenetic embryonic
tissues from virgin turkeys. J. Hered., 46,
133-134.

YoKOYAMA, T. 1959. Silkworm Genetics Illus-
trated. Tokyo: Japan Society for Promotion
of Science.

Zoschke, U. 1956. Untersuchungen iiber die
Bestimmung des Geschlechtes beim Spinat
unter besonderer Beriicksichtigung der Zuch-
tung eines monozischen oder gleichzeitig
schossenden Spinats. Ztschr. Pflanzenziicht.,
35, 257-296.

AL

ROLE OF HORMONES IN THE
DIFFERENTIATION OF SEX

R. K. Burns, Ph.D., D.Sc. ihon.)

CA.RNEGIE INSTITUTION OF WASHINGTON, DEPARTMENT OF EMBRYOLOGY,
THE JOHNS HOPKINS UNIVERSITY, BALTIMORE, MARY'LAND

I. The Hormone Theory of Sex Dif-
ferentiation 76

II. Methods of Experimental Analy-
sis 78

A. Grafting of Gonads or Gonad Tis-

sues in Bird Embryos 78

B. Grafting Experiments in Amphib-

ian Embrj'os 79

C. Use of Pure Hormones as Sex Dif-

ferentiating Agents 82

D. Sex Differentiation in the Absence

of Hormones 82

III. The Bisexual Organization of the

IiIarly Embryo as the Basis of Sex
Reversal 83

IV. Experimental Reversal of Sex Dif-

ferentiation IN the Gonads 83

A. Bisexual Organization of the Gonad

and the Physiologic Mechanism

of Sex Differentiation 83

B. Sex Reversal in Amphibian Gonads 86

1. Constitutional differences and

the character of the reversal
process 80

2. Parabiosis and grafting of the

gonad or the gonad primordium 87

3. Administration of steroid hor-

mones 91

C. Sex Rkv'ersal in Avian (Ionads. 95

1. Organization of avian gonads. . . 95

2. Effects of administering pure

hormones 9(i

3. Effects of grafting gonads into

the coelomic cavity 99

4. Sex reversal in vitro 100

D. The Problem of Sex Reversal in

Mammalian Gonads 100

1. Bisexual potentialities in the

embryonic gonads of mammals 100

2. Bisexual potentiality in the em-

bryonic ovary of the rat 103

3. Experimental transformation

of the testis in the opossum. . . 105
V. The Role of Hormones in the De-
velopment OF the Accessory Sex
Structures 110

A. Differentiation of the Embryonic

Gonaducts Ill

1. The Miillerian ducts 112

2. The male duct system 120

B. Derivatives of the Cloaca and

Urinogenital Sinus 121

C. External Genital Structures 127

D. Differentiation of Other Types of

Sex Character 129

VI. The Pituitary and the Differen-
tiation OF Sex 132

VII. Group Differences in the Rela-
tions OF Hormones to Sex 134

VIII. The Organization of the Sex Pri-
mordium AND Its Role in the Dif-
ferentiation OF Sex 137

A. Constitution and the Morphologic

Representation of Sex Primordia 137

B. Constitutional Factors and Physi-

ologic Differences in the Organi-
zation of Sex Primordia 138

C. Influence of Sex Genotype on the

Reactions of Sex Primordia 139

IX. The Time F'.^ctor in the Responses
OF Sex Primordia: Receptivity

and "Critical Periods" 140

X. Specificity of Hormone Action and
the Significance of P.\radoxical

Effects 141

XI. Time of Origin and the Source of

(ioNAD HoR.MONES 143

XII. A Comparison of the Effects of Em-
BRYt)Nic and Adult Hormones in
Skx Differentiation 145

.\1I1. IvMBRYONic Hormones and Inductor

Substances 148

XIV. References 151

I. The Hormone Theory of
Sex Diflfereiitiation

The modern era in the study of the pliys-
iology of sex differentiation is iisually dated
from the sohition of the freemartin iiroblem
through the simultaneous but entirely inde-

76

HORMONES IN DIFFERENTIATION OF SEX

pendent studies of Lillie (1916, 1917) and
Keller and Tandler (1916). The theoretic
explanation of the anomaly proposed by
these authors was generally accepted and
had two far-reaching effects ; it at once pro-
vided a simple, functional concept of the na-
ture of embryonic sex differentiation which
was readily susceptible of experimental test,
and it directly stimulated the pioneer ex-
periments in the field — the attempt to con-
trol sex differentiation in chick embryos by
grafting gonad tissue to the chorioallantoic
membrane (Minoura, 1921) and the appli-
cation of the technique of parabiosis to the
problem in amphibian embryos (Burns,
1924, 1925). However, as is usually the case
with epoch-making theories, the concept of
hormonal control of embryonic sex differen-
tiation had roots going far back into the
past.

The effects of castration in domestic ani-
mals and in humans had been familiar since
earliest times, and it was long appreciated
that in the vertebrates the gonads are neces-
sary for maintaining the structural integrity
of the accessory organs of reproduction and
for the regulation of their functional acti^vi-
ties. It was first clearly demonstrated by
Berthold (1849) that this control is exer-
cised through the agency of a substance (of
a nature as yet unknown) produced in the
gonads and carried throughout the body in
the circulating blood (for the historical
background of this experiment see Forbes,
1 949 ) . Thus the conception of a blood-borne
agent capable of controlling the growth and
activities of distant structures, was estab-
lished long before the name hormone was
given to such substances. The theory that
the differentiation of the genital structures
in the embryo is controlled by a hormone,
or hormones, produced by the embryonic
gonads was a natural outgrowth of this
knowledge. This view was first proposed as
an hypothesis by Bouin and Ancel in 1903,
suggested by the observation of an unusu-
ally rich interstitium in the testes of pig
embryos during the period of sex differentia-
tion. No direct evidence in support of the
hypothesis was forthcoming, however, until
the hormone theory, in virtually its present
outlines, was formulated by Lillie and by
Keller and Tandler as an explanation of the
freemartin.

The freemartin^ had been familiar to
breeders of cattle for centuries as a sexually
abnormal calf, born as twin to a normal
male, and its anatomy had been accurately
described by John Hunter in the eighteenth
century (for references see Lillie, 1917).
The external genitalia and mammary glands
are typically female in character and the
animal was usually regarded as a female,
but in rare cases the clitoris may be greatly
enlarged and peniform (Numan, 1843, see
Lillie, 1917, Fig. 29; Buyse, 1936). Inter-
nally, however, elements of the genital tracts
of both sexes are present and frequently
well developed, and later investigators were
often in disagreement as to the primary sex
of the creature. The gonads of the freemartin
are rudimentary in form but usually show
the histologic structure of an abnormal testis
which is almost invariably sterile (for an
exception see Hay, 1950, and also Hart,
cited by Lillie, 1917, page 417). In many
cases, however, the gonads are intersexual,
showing varying degrees of agenesis of the
ovarian cortex associated with rudimentary
tubular structures in the medullary or hilar
regions (Chapin, 1917; Willier, 1921). Com-
monly a well developed male duct system
is present, but the development of the fe-
male genital tract is variable and in the
more modified cases it may be virtually ab-
sent.

It is unnecessary to go into the various
lines of evidence which were used to es-
tablish the fact that the freemartin is zy-
gotically a female (Lillie, 1917, 1923) ; the
point has recently been demonstrated cyto-
logically by the Barr method (Moore,
Graham and Barr, 1957). It will be useful,
however, to review the circumstances which
pointed to the cause of the anomaly. A free-
martin is always associated at birth with a
male twin (which is normal) and never with
another female; in addition, the dizygotic
origin of the pair was demonstrated by
Lillie in many cases. Furthermore, for the
birth of a freemartin it is necessary that the
placentas of the twins be united, with the
presence of vascular anastomoses (Fig. 2.1).
In the absence of such connections the fe-
male twin is always normal. There is a cor-
relation between the degree of abnormality

^ For a discussion of the origin of this name see
Forbes (1946).

78

BIOLOGIC BASIS OF SEX

Fig. 2.1. Twin calves removed fiom the uterus, showing chorionic fusion and anastomosis
between major vessels; male twin, left; freemartin, right. (After F. R. Lillie, J. Exper. Zool.,
23, 371-452, 1917).

observed and the extent of the vascular
union, and also with the stage of develop-
ment at which the anastomosis was pre-
sumably established. The constancy with
which these conditions appear pointed in-
evitably to the conclusion that the abnor-
mal development of the female twin is
caused by the transfusion, from an early
stage of development, of a hormone pro-
duced by the gonads of the male twin. The
invariable dominance of the male member
of the pair was explained provisionally on
the basis of histologic studies (Lillie and
Bascom, 1922; Bascom, 1923) which indi-
cated that the testis is active endocrinologi-
cally long before the ovary (rf. Bouin and
Ancel, 1903). This conclusion has been sup-
ported in recent years by the results of
castration in mamnialian embryos (q.i\).-

■Reccnll_\' eNidciice lias come from (|uitc (hl-
ferent sources that in most twins or muUiplc biilhs
in cattle, placental anastomoses are established at
an early stage. Calves, even when of different sex
and with other characteristics indicating dizygotic
or polyzygotic origin, possess identical comple-
ments of l)lood factors (red cell agglutination
types), which can only l)e explained on the basis
of an early exchange of blood (Owen, Davis and
Morgan, 1946). Since erythrocytes are compara-
tively short-lived cells, it is indicated in these cases
that in-imitive erythroblasts must have been ex-
changed early in development and colonized the
hemopoietic tissues of the recipients. It has been
shown also (Anderson, Billingham, Lampson and
Medawar, 1951) that diz3^gotic twin calves of dif-

II. Methods of Experimental Analysis

The demonstration in tlie case of the free-
martin of the probable nature of the trans-
forming agent and its mode of transmission
at once suggested means of attacking the
problem of embryonic sex differentiation ex-
perimentally. At first grafting methods were
mainly employed, using the embryos of
birds and amjihibians.

A. CRAFTING OF GONADS OR GONAD TISSUES
IX BIRD EMBRYOS

Historically, the first experiments were
those of JNIinoura (1921) who trans})lanted
pieces of adult testis or ovary to the chorio-
allantois of chick embryos during the period
of sex differentiation. Such grafts become
vascularized and are then in communication
witli the host embryo by way of the umbili-
cal circulation. Various modifications of

ferent sex are, with few exceptions, tolerant to
grafts of each other's skin. Tliis is true of skin ex-
changes in monozygotic twins (as would be ex-
jiected) but is never found in other degrees of
r(>hitionship. As in tlie preceding case, the ex-
])lanalion is found in an early transfusion of blood
l)etwe(ni the twins. The exceptional cases are
doubtless to be explained (as in the freemartin
study) on the occasional failure of placental anas-
tomosis to occur. This evidence is cited for its
bearing on tlic point of early exchange of blood;
tlie fact that blood cells and other elements are
exchanged does not seem to be of significance for
the sex hormone theory.

HORMONES IX DIFFERENTIATION OF SEX

f9

the embryonic genital structures of the host
(especially of the Miillerian ducts) were
noted and attributed to the influence of the
grafted tissues. But later investigators failed
to confirm these findings, and it was shown
eventually that the anomalies observed by
]\Iinoura were unspecific in character and
bore no constant relation to the sex of the
grafted tissue (for a review and discussion
see AVillier, 1939). Similar modifications
were also found after transplantation of
various nongonadal tissues, and are appar-
ently induced by changes in the physical
environment incidental to operation, such
as lowering of the temperature and the hu-
midity (Willier and Yuh, 1928). Evidently
the original experiments had not been ade-
quately controlled with respect to such fac-
tors.

Obviously this type of experiment does
not correspond exactly with conditions in
the freemartin. The grafted tissue came
from adult gonads, and there was no way
of determining the hormone output of such
grafts or whether, indeed, they produced
hormones at all. Furthermore, sexual dif-
ferentiation in the host embryos has gen-
erally begun before transplantation to the
chorioallantois is practicable. This objection
was avoided, however, by modifying the ex-
l)eriment. Sexually undifferentiated embry-
onic gonads were transplanted to the chorio-
allantoic membrane of a host embryo
already well advanced in sex differentiation.
In this case changes might be anticipated
in the grafts. However, this experiment, as
well as transplantation of the gonad-form-
ing region of the blastoderm (Willier, 1927,
1933), also gave negative results. The
grafted gonads differentiated to a variable
degree, depending on the state of develop-
ment of the primordia at the time of trans-
plantation, but when sexual differentiation
occurred it showed no constant relation to
the sex of the host.

These failures to obtain evidence that
gonad tissues growing on the chorioallantois
influence the differentiation of host struc-
tures, or are themselves modified by the
hormones of the host, raised serious doubt
as to the role of hormones in embryonic sex
differentiation in birds ; and this feeling was
not entirely removed by the demonstration

some years later that the sexual differentia-
tion of the chick can be readily modified by
treatment with pure hormone preparations.
It was not until embryonic gonads were
transplanted directly into the body cavity
of another embryo (Bradley, 1941 ; Wolff,
1946) that unmistakable evidence was ob-
tained. The conditions under which this re-
sult was achieved — close proximity of the
graft to the developing host structures —
suggested that the failure to obtain positive
results by chorioallantoic grafting was per-
haps largely a matter of the quantity or
concentration of the hormone. Recently,
however, a true "freemartin effect" has been
reported in twin chicks of different sex,
which developed from an egg with two yolks
(Lutz and Lutz-Ostertag, 1958). There was
local development of cortex on the left
testis of the male twin, and a marked inhibi-
tion of the ]Miillerian ducts of the female,
effects paralleling those produced by in-
tracoelomic gonad grafts. This is the only
recorded case of a natural freemartin in
birds.

B. GR.\FTIXG EXPERIMENTS IX AMPHIBIAN
EMBRYOS

The principal experimental procedures de-
veloped for amphibian embryos are illus-
trated diagrammatically in Figure 2.2,
which shows the different modes of grafting,
and the resulting vascular relationships, as
compared with the freemartin (Fig. 2.1).
The first experiments actually undertook to
reproduce as nearly as possible the situation
which arises by chance in the freemartin.
The method devised was parabiosis— the
grafting together of two embryos in the
manner of "Siamese twins" (Burns, 1925)
so that in later development there is a
common circulation (Figs. 2.2A and 2.9).
When members of such a pair happen to be
genetically of the same sex, normal sexual
differentiation would be expected to follow;
but in pairs of different sex opportunity for
cross-circulation of sex hormones is pro-
vided. Circulatory anastomosis is estab-
lished in such pairs long before the begin-
ning of sex differentiation in the gonads,
and so a favorable situation is provided for
testing the possibility of hormone action.
The results obtained by this method vary

80

BIOLOGIC BASIS OF SEX

A B C

Fig. 2.2. Diagram illustrating different modes of grafting in amphibians in order to bring
about vascular continuity between individuals, and association of gonads of different sex. A.
Homoplastic twins in salamanders. The body cavities are largely separated and vascular com-
munications between the gonads are remote. For combination of dissimilar species see Fig.
2.9B. B. Anuran twins, showing side-to-side or head-to-tail union; reversal changes appear
only under the first condition, when the gonads are in close proximity. (After E. Witschi, in
Sex mid Internal Secretions, The Williams & Wilkins Company, 1932). C. Orthotopic trans-
plantation of the gonad primordium by the method illustrated in Fig. 2.4, resulting in two
gonads of opposite sex resident in a single individual (Humphrej''s method).

greatly, depending on the species under
study and on various experimental condi-
tions, as will appear later.

The grafting of gonads alone (as opposed
to the union of whole organisms) can be car-
ried out in embryonic stages of development
or in early larval life. The latter method
was tried first. The gonads, attached to a
segment of the mesonephric bodies, were re-
moved from young larvae at or soon after
the onset of sex differentiation and inserted
into the body cavity of older larvae (Burns,
1928) . The development and activity of such
grafts depends on the extent to which they
become attached and vascularized. When
graft and host are of different sex the grafts
typically become intersexual, developing the
structure of ovotestes; and when a large and
well differentiated graft is in close proximity
to the gonads of the host the latter may be
similarly modified (Fig. 2.3). This method
of grafting has the disadvantage, however,
that reversal is usually incomplete, and
when graft and host gonads show reciprocal
modification it is sometimes difficult to de-
termine the primary sex of either.

The method described above was soon
greatly improved upon by the development
of a technique for transplanting, at an ear-
lier stage, the prospective gonad-forming
tissue from one embryo to another (Hum-
phrey 1928a, b). At first such grafts were
placed in ectopic locations, but later it was
found advantageous to place them in the
normal (orthotopic) position in an embryo
from which the corresponding gonad pri-
mordium had been excised (Fig. 2.4). After
such an operation the host embryo bears on
one side its own gonad and on the other a
gonad which, in approximately half of all
cases, has come from an embryo of the
other sex (Fig. 2.2C). This method has im-
portant advantages over those previously
described. A single embryo bearing an or-
thotopic graft survives better and is more
easily reared tlian are parabiotic pairs, and
gonads grafted in tlio orthotopic position
usually develoj) better than in foreign sur-
roundings. Most important of all, the donor
embryo may be reared, thus establishing
with certainty the original sex of the grafted
gonad. This precise method has yielded un-

HORMONES IN DIFFERENTIATION OF SEX

81

Fig. 2.3. Transplantation of the salamander gonad in early larval life (Burns, 1928):
previously unpublished photographs. A. A large, but somewhat degenerate, grafted ovary
lies just anterior to (above) the gonads of the host, which show changes in external form in
the vicinity of the graft. B. Cross-section of host's right testis at the level of the white line,
showing normal medullary development with well differentiated testis lobules, and periph-
erally a strongly developed cortex.

A B

Fig. 2.4. Diagrams illustrating Humphrey's orthotopic transplantation method. A. Posi-
tion of the gonad- and mesonephros-forming area of the embryo (stippled) which is excised
and reimplanted in the corresponding position in a host embryo from which the primordium
has just been removed. B. Cross-section of host at later stage showing position of the im-
planted material (between heavy lines) at the left. In the mesodermal layer a part of the
lateral mesoderm lies below, the gonad- and mesonephros-forming material above, with
the Wolffian duct at the top. Medial to the Wolffian duct lies the mass of primordial germ
cells (more densely stippled).

82

BIOLOGIC BASIS OF SEX

equivocal results which have in general con-
firmed and extended those obtained by para-
biosis.

C. USE OF PURE HORMONES AS SEX
DIFFERENTIATING AGENTS

Early attempts to influence embryonic
sex differentiation by the use of crude hor-
mone preparations were almost entirely un-
successful because of lack of potency, or
the toxicity of the extracts. However, the
isolation and eventual synthesis of steroid
hormones made available a variety of ac-
tive and nontoxic substances, and the use of
pure hormones largely superseded grafting
techniques. Direct administration of stand-
ard hormone preparations has the great ad-
vantage that dosages can be exactly known
and regulated; also the timing of treat-
ments is readily controlled and varied.

The first successful experiments using
pure hormones were carried out on chick
embryos. Similar results were obtained at
almost the same time by several groups of
investigators (Kozelka and Gallagher, 1934;
Wolff and Ginglinger, 1935; Dantchakoff,
1935, 1936; Willier, Gallagher and Koch,
1935, 1937j who introduced the hormones,
in oily or in aqueous solution, into the in-
cubating egg. Striking transformations were
produced, involving both the structure of
the gonads and the accessory organs of sex.
In the best cases reversal of the gonads was
histologically almost complete.

The effects of crystalline sex hormones
have also been investigated in many species
of amphibians. Two methods have been
utilized. Larvae may be treated individually
by repeated injections, or in groups by con-
tinuous immersion, the hormone being dis-
solved in the water in which the larvae are
reared. The latter method is particularly
convenient for anuran tadpoles. Treatment
can be started very early, the concentration
is readily varied, and in many cases com-
plete transformations have been obtained
with the use of extremely low concentra-
tions.

In mammalian embryos experimental
study of sex differentiation was long delayed
by the lack of operative techniques ade-
quate for dealing with embryos in utero.
The advent of pure hormones made possible
the first successful experiments in this field.

In i^lacental forms, in spite of a very high
mortality, large doses of crystalline hor-
mones can be administered to the mother
during early stages of pregnane}" with pro-
nounced effects on the genital systems of the
embryos (for a review of the earlier experi-
ments see Greene, 1942, and for a recent
summary Jost, 1955) . About the same time,
experiments were begun using the pouch
young of a marsupial, the North American
opossum (Burns, 1939a, b; Moore, 1941).
So undeveloped are young marsupials at
birth that virtually the entire course of
morphologic sex differentiation takes place
postnatally, and the embryos in the pouch
are directly accessible for experimentation.
Hormones were administered by injection
(Burns) or by inunction, the application of
an ointment containing the hormone to the
skin. Except for minor differences attributa-
ble to dosage or other experimental factors,
the results were similar, and in general
agreement with those obtained in placental
mammals by treatment during pregnancy.

D. SEX DIFFERENTIATION IN THE
ABSENCE OF HORMONES

Although the evidence obtained by graft-
ing techniques and by administration of
hormones shows that the differentiation of
embryonic genital structures may be pro-
foundly modified or even completely re-
versed, such evidence is not in itself con-
clusive with respect to the central problem,
the role of hormones in the normal differ-
entiation of sex. The transmissible sub-
stances responsible for sex reversal in graft-
ing experiments have not been isolated or
identified and it may be argued that experi-
ments with pure hormones show merely that
the differentiating embryonic sex primordia
are capable of reacting when hormones are
introduced experimentally. Such evidence
docs not prove, however, that embryonic
gonads actually produce such hormones. For
this question the crucial test is the capacity
of the embryonic genital structure to de-
velop in the absence of gonads or removed
from all hormonal influence.

Evidence on this point has been forth-
coming in recent years and gives strong
supi)ort to tiie hormone theory. Two differ-
ent experimental approaches have been de-
veloped. Early castration of the embryo has

HORMONES IN DIFFERENTIATION OF SEX

83

now been achieved in both mammals and
birds, and improved methods of culturing
embryonic organs in vitro have made is pos-
sible to observe for a sufficient time the de-
velopment of sex primordia in complete
physiologic isolation.

Since 1947 castration has been success-
fully performed in amniote embryos by two
techniques, surgical castration and irradia-
tion of the gonad region (for summaries see
Jost, 1950; Wells, 1950; Raynaud, 1950;
Wolff, 1950; Huijbers, 1951). In all cases
serious failures of sexual differentiation fol-
low removal or destruction of the gonads.
Finally, the cultivation in vitro of indi-
vidual sex primordia in virtual absence of
hormonal influences has yielded results simi-
lar in all respects to those of castration. The
results of these experiments will be taken
up in detail as they relate to the develop-
ment of particular structures.

in. The Bisexual Organization of the

Early Enihryo as the Basis of

Sex Reversal

The capacity of vertebrate embryos to
undergo a reversal of sex, either spontane-
ously, as in various developmental anoma-
lies of undetermined etiology ("intersexu-
ality," "hermaphroditism"), or as a result
of experiment, is based on the fact that every
individual, regardless of genie sex constitu-
tion, passes in early development through a
sexually undifferentiated or ''indifferent"
phase. During this period virtually all of
the embryonic structures necessary for the
development of either sex are laid down
morphologically and are present for a cer-
tain time as discrete 'primordia. The extent
to which the primordia of the genetically
recessive sex are developed and the length
of time during which they are present
vary in different groups and species. This
fact is of great importance in the experi-
mental transformation of sex. In species in
which the structures of the recessive sex are
imperfectly represented, or are present for
only a brief period in early development,
opportunity for sex reversal is correspond-
ingly limited; but in other cases the rudi-
mentary structures of the recessive sex (as
for example Miillerian ducts in males or
vestigial prostatic glands in females) per-
sist indefinitely and mav even survive in the

adults of some species. The existence of a
considerable degree of embryonic bisexual-
ity in most groups (see Fig. 2.22) provides
a definite morphologic basis for experi-
mental transformation of sex and for the
sporadic occurrence of sex anomalies as well.
The derivation of the various embryonic
primordia which give rise to the male and
female genital systems, and their history
in the normal differentiation of sex, have
been extensively reviewed by AVillier (1939)
and will not be taken up again in detail.
The main features of normal development
will be outlined briefly when dealing with
the experimental behavior of individual
structures. It must be remembered, however,
that the individual parts of the genital sys-
tem have widely different embryonic origins,
and are morphologically and physiologically
very dissimilar, and at any particular stage
of development may vary greatly in their
relative maturity and so in their reactivity
to hormones. Many of the basic structures
(e.g., the embryonic sex ducts, the urino-
genital sinus) are taken over bodily from
other systems and only secondarily acquire
a sexual status. It cannot be expected, there-
fore, that all parts of the developing sex
complex will be capable at all times of re-
sponding harmoniously to experimental con-
ditions which are often of necessity rigid
and artificial or improperly timed. The rec-
ognition of such differences aids in under-
standing the variability so frequently en-
countered in the reactions of sex structures
to hormones, and the importance of such
experimental factors as the timing of treat-
ment and dosage.

IV. Experimental Reversal of Sex

Differentiation in

the Gonads

A. BISEXUAL ORGANIZATION OF THE GONAD

AND THE PHYSIOLOGIC MECHANISM

OF SEX DIFFERENTI.ATION

The sexually undifferentiated gonads of
most amphibians exhibit bisexual organiza-
tion in a primitive and simple form. In
early larval life the gonad, irrespective of
its future sex, contains two histologically
distinct components in which male and fe-
male potentialities are segregated. Inter-
nally, there is a hilar or centrally placed

84

BIOLOGIC BASIS OF SEX

NDIFFERENT GONAD

OVARY TtSTIS

Fig. 2.5. Diagrammatic representation of the
male and female components of the sexually un-
differentiated amphibian gonad and their roles in
sex differentiation: the medulla is stippled, the
cortex is plain. The broken arrows indicate the
mutually antagonistic or inhibitory actions exerted
between the two components in the course of sex-
ual differentiation (Witschi).

mass, the medulla, which has the develop-
mental potentiality of a testis. Surrounding
the medulla is a peripheral zone, the cortex,
which is specifically female in potency. The
topographic relationships of medulla and
cortex are illustrated diagrammatically in
Figure 2.5, and as they appear histologically
in male salamanders in Figure 2.QA and B.
Male and female potentialities are evidently
pre-established in the medullary and corti-

cal components at an early stage since final
differentiation as a testis or an ovary does
not involve the transformation of one sex
comj^onent into the other but rather the
gradual predominance of one element and
the recession of the other.

Not only are the two components dis-
tinctly segregated in the indifferent gonad,
they have separate origins. It has long been
recognized that the medulla in both sexes
is derived from the mesonephric blastema
in the form of a series of cellular strands,
the medullary cords or rete cords, which
grow into the genital ridge at an early stage.
At first similar in appearance in the two
sexes, their later differentiation follows very
different patterns. In the ovary the cords
expand distally, forming a series of saccular
cavities, the ovarian sacs. Most of the germ
cells are excluded from these sacs and come
to lie in a peripheral layer beneath the peri-
toneal epithelium covering the gonad. This
zone becomes the cortex. In prospective
testes the cords branch and proliferate rap-
idly, enveloping and incorporating the ma-
jority of the germ cells in a compact central
mass, the medulla ( Fig. 2.6 ; for a fuller de-
scription see Willier, 1939). Thus the rela-
tive proportions of cortex and medulla in
the sexes depend on the pattern of differen-

FiG. 2.6. Sections showing the histologic appearance of cortex and medulla in early larval
stages under various conditions. A. Normal testis of an Amby stoma tigrinum larva; note the
thin cortical zone, comparable to a germinal epithelium, which covers most of the surface.
B. The cortex of an intersexual testis dissected free from the medullary core, from the punc-
tatum member of an Ambtjstoma tigrinum, 9 ; A. Punctatum, $ pair (c/. Burns, 1935, plate
4). C. Intersexual testis of the male member of a tigrinum-tigrinum pair, showing tl)e rela-
tive development of the medullary and cortical components (see Burns, 1930).

HORMONES IN DIFFERENTIATION OF SEX

85

tiation of the medullary cord, which results
in a very unequal allocation of the germ
cells between the two components. Prob-
ably the sex genotype acts primarily by
determining the developmental pattern of
the medullary cord. The role of the germ
cells in the formation of the gonad appears
to be a purely passive one since the non-
germinal tissues are capable of producing
the typical structure of a testis or an ovary
in the absence of all germinal elements (for
a review of this subject see Burns, 1955b).

The origin of the medullary component
of the gonad from the mesonephric blastema
and its dominant role in gonad formation
has been demonstrated in a striking experi-
ment by Houillon (1956). Formation of the
gonad is dependent on the normal develop-
ment of the mesonephros, and this can be
repressed or entirely prevented by blocking
the development of the primary nephric
duct (pronephric ductj at an early stage.
In the absence of the nephric duct the meso-
nephric blastema is reduced in quantity and
delayed in its appearance; typically only a
few mesonephric tubules develop and these
are poorly differentiated. In consequence of
the suppression of the mesonephric blas-
tema, medullary cords are lacking or poorly
developed, and the result is a vestigial gonad
consisting chiefly of a rudimentary cortex.
The essential role of the medullary cords in
gonadogenesis is also demonstrated in the
gonads of toads (Witschi, 1933) in which
the so-called "organ of Bidder" corresponds
to an anterior segment of the genital ridge
in which medullary cords are absent.

The proportion of cortex to medulla as
laid down in embryonic gonads depends
primarily on genie constitution, but the
stage of development is a factor in the rep-
resentation of the two elements at any par-
ticular time, since during the progress of
sexual differentiation one component (cor-
responding to the genetically determined
sex) shows an increasing predominance from
stage to stage. Furthermore, the morpho-
logic representation of the two sex compo-
nents in the indifferent gonad is subject to
variation in different groups, species, or
races. In some species the recessive com-
ponent is weakly represented or virtually
absent, even in early development, or when
present its existence may be of brief dura-

tion. In such cases capacity for sex reversal
is reduced or lacking. In other species, in
which the recessive sex component is well
developed, or when it persists over a con-
siderable period of time, capacity for ex-
perimental reversal is correspondingly in-
creased.

The alternative behavior of the cortical
and medullary components of the gonad in
normal differentiation, as well as their be-
havior under experimental conditions, long
ago suggested that the physiologic mecha-
nism of sex differentiation consists essen-
tially of an antagonistic interaction between
the two elements, in which the genetically
dominant component gradually inhibits its
antagonist (Fig. 2.5). This concept of "cor-
ticomedullary antagonism" (Witschi, 1932)
has been generally accepted as the basic
mechanism in the histologic differentiation
of the gonad and forms the starting point
for the inductor theory of sex differentia-
tion.^ A number of seemingly unrelated ex-
perimental procedures which are capable of
inducing sex reversal all appear to have a
common base of action by influencing or
controlling this simple mechanism. For ex-
ample, sex reversal is in some cases readily
induced by external or environmental influ-
ences of an unspecific character, which ap-
parently produce their effects by depressing
or destroying the dominant gonad compo-
nent.

Classical experiments of this type are the

^ As originally formulated, this theory postulated
simply that each gonad component produces a
substance which specifically inhibits the differen-
tiation of the other. These substances, called med-
ullarin and corticin, were considered to be similar
in character and to behave like the embryonic in-
ductors of earlier development, being transmitted
by diffusion and having strictly localized effects.
Subsequently the theory was elaborated to allow
for stimulatory as well as inhibitory action, with
each inductor system assigned a dual role; ac-
cordingly, positive and negative factors were as-
sumed and so designated, e.g., medullarin* and
medullarin'. More recently it has been proposed
that interactions between the sex inductors are of
the nature of an immunologic reaction (Chang
and Witschi, 1956), the positive factor appearing
first in the role of an antigen which stimulates the
other system to produce an antibody, the inhibi-
tory factor (for the development of the inductor
theory see Witschi, 1934, 1939, 1942, 1950, 1957). To
account for the great taxonomic variability in the
action of the sex inductors thej' are assumed to
be proteins.

BIOLOGIC BASIS OF SEX

use of extreme^; of temperature to induce
reversal of differentiation in the gonads of
anuran larvae (Witschi, 1929; Piquet, 1930;
Uchida, 1937 1 . The reversal is due primarily
to an unfavorable effect on the dominant
component, high temperatures causing cor-
tical degeneration in females and low tem-
peratures inhibiting medullary development
in males. Also, simple surgical interventions
or even pathologic injury may have the
same effect. Castration in certain cases re-
sults in complete reversal of sex through the
reactivation and renewed development of a
recessive gonad component left behind at
operation. In adult male toads removal of
the testes permits the organs of Bidder to
develop into ovaries, which may become
fully functional (Ponse, 1924). A compara-
ble case is found in the reversal of sex which
takes place in female chicks castrated soon
after hatching. Removal of the dominant
left ovary is followed by development of the
rudimentary right gonad, composed largely
or entirely of medullary tissue, into a small
testis. Finally, rare cases of partial or com-
plete sex reversal in adult hens, which oc-
cur as a result of pathologic destruction of
the functional ovary, appear to have the
same morphologic basis (Crew, 1923; for a
discussion see Domm, 1939 1 .

But although sex differentiation in most
vertebrates ends in the complete dominance
of one sex component, remarkable devia-
tions from this plan are known in certain
groups and species. An extreme is found in
the prevalence of hermaphroditism, in vary-
ing degree, in many teleost fishes and in
cyclostomes (see ch. 17) which may be of
the juvenile type and temporary, or may
persist in adults. In toads the curious struc-
ture known as Bidder's organ is present in
adults of both sexes; it represents a local
region of the genital ridge in wliich medul-
lary cords are never formed and furtlier dif-
ferentiation does not occur. Since it corre-
sponds morphologically to the cortical
component of the gonad it retains through-
out life the potentiality of an ovary. This
condition is apparently possible in toads
because of the very low level of antagonism
in this genus. Stranger still is the situation
found in the female of certain insectivores
(the mole, Godet, 1949, 1950; the desman,

Peyre, 1952, 1955j in which the medulla
of the adult ovary is testis-like and devel-
oped to a remarkable degree. Except during
the reproductive period it greatly exceeds
the cortex in bulk. Its cords are tubular in
form, resembling testis tubules, and a well
developed interstitium indicates an endo-
crine activity which is reflected in the strong
masculinization of the genital tract. The
clitoris is large and penis-like, and male ac-
cessory glands, absent or rudimentary in the
females of most mammals, are well devel-
oped. Another species in which the ovarian
medulla is highly developed, at least
throughout fetal life, is the horse (Cole,
Hart, Lyons and Catchpole, 1933). Thus
many patterns are found with respect to the
persistence of the heterotypic sex compo-
nent of the ovary and its final fate.

B. SEX REVERSAL IN AMPHIBIAN GONADS

1. Constitutional Differences and the Char-
acter of the Reversal Process

Modern amphibians, far from being a
homogeneous group, are extremely diversi-
fied in structure and function and are often
highly specialized. Such diversification ob-
viously has had a long evolutionary history.
Correspondingly, the processes of sex re-
versal as evoked experimentally in amphib-
ian gonads, often follow very different histo-
logic and physiologic patterns in different
groups, species, or races. These differences
must rest ultimately on genetic constitution ;
more immediately they are predetermined,
in labile fashion at least, in the structural
and physiologic organization of the gonad
primordium which is in itself a complex sys-
tem.

The organization of the early gonad may
vary greatly (according to species and ac-
cording to sex) with respect to the cortical
and medullary elements as laid down histo-
logically in the primordium — is the hetero-
tyi:)ic sex component of the primordium well
represented or is it quantitatively deficient
fiom the beginning? The subsequent be-
liavior of the heterotypic component is also
important— does it jiersist and regress slowly
over a considerable period of time or is its
existence transient? Furtliermore. how does
it react when the normal balance of the dif-

HORMONES IN DIFFERENTIATION OF SEX

87

ferentiation process is experimentally dis-
turbed — does it respond readily by growth
and differentiation or is it relatively inert
and refractory? In particular cases the ca-
pacity of a gonad for reversal under experi-
mental conditions obviously depends on
which of the various situations prevails. An-
other variable concerns the humoral activity
of the gonad, as regards the time of onset
and the factors of quantity or rate of pro-
duction. Species differ widely in this respect
and marked sex differences are also found.
Presumably all such characteristics exist as
predispositions within the gonad primor-
dium, but they are not as a rule irreversibly
determined.

In addition the process of reversal, as seen
histologically, may be influenced by experi-
mental conditions such as the procedure em-
ployed, the stage of development at which
reversal is initiated and the duration of the
experiment. If conditions are favorable at
the beginning of sex differentiation, reversal
may take place directly without leaving ob-
vious histologic traces, i.e., an individual of
one sex may adopt the developmental pat-
tern of the other virtually from the start.
If, on the other hand, transformation is not
initiated until sex differentiation is well ad-
vanced, various stages of intersexuality will
appear in the transforming gonads, until
one sex component finally establishes com-
plete dominance and the other disappears.
The first situation commonly occurs when
larvae are reared in optimal concentrations
of hormone dissolved in the aquarium wa-
ter; exposure to the hormone is continuous
from a stage long antedating the appearance
of morphologic sex differentiation, and all
embryos regardless of sex genotype develop
as one sex. However, the same result may
occur also in grafting experiments (parabio-
sis or gonad transplantation) in which het-
eroplastic combinations of different species
assure decisive predominance of one sex
from the beginning by virtue of great in-
equality in size and in rate of development.
The second situation is encountered when
the conditions of the experiment do not lead
to establishment of dominance at an early
stage. Reversal sets in late, intersexual
stages may be prolonged, and complete
transformation mav never occur.

2. Parabiosis and Grajting of the Gonad or
the Gonad Primordium

Experiments of this kind involve the in-
teraction of embryonic or larval gonads
through the agency of substances of a hu-
moral nature but of unknown chemical con-
stitution. In some species, or under certain
experimental conditions, the effects may be
limited and highly localized, appearing only
when the interacting gonads are in contact
or in close proximity. Transport of the hu-
moral agent takes place apparently by dif-
fusion through the intervening tissues (Figs.
2.25 and 2.3 j . In other cases the effects are
exerted over great distances and the sub-
stances must of necessity be carried in the
blood. This does not necessarily mean, how-
ever, that different substances are involved
in the two cases. As will be shown, hor-
mones in low concentrations may have only
local effects and, given a sufficient concen-
tration in the blood stream, there is no rea-
son to suppose that the so-called "inductor
substances" could not act at a distance. The
mode of transport does not seem to be cru-
cial for the definition of these substances
(for discussions see Willier, 1939, page 134,
and Burns 1949, 1955b).

In most species of amphibians which have
been investigated the male is the dominant
sex. In grafting experiments, whether the
method is parabiosis or transplantation of
embryonic gonads, testes as a rule induce sex
reversal in ovaries without being greatly
modified themselves. In some cases domi-
nance of the testis is so extreme that no real
reversal of the ovary occurs, only an almost
complete suppression and sterility. This
type of response is seen, for example, in
parabiotic pairs of the wood frog (Witschi,
1927) and in the newt Triturus (Witschi
and McCurdy, 1929), and is probably cor-
related with a constitutional inadequacy of
the medullary component in the embryonic
ovaries of the species in question. In other
species, on the other hand, a severe initial
repression of the ovary is followed by a de-
layed reversal, which may have a prolonged
course but which eventuall}^ may be quite
complete. In parabiotic pairs of certain spe-
cies of Ambystoma (Fig. 2.7) there is a se-
vere inhibition of the ovary before active
transformation is initiated, and when rcver-

88

BIOLOGIC BASIS OF SFA'

Fig. 2.7. An extreme degree of inhibition and reduction in certain regions of an ovary
under the dominance of a well developed testis (Humphrey, 1942). A. Level showing sterile
medulla above, with degenerate cortical zone below. B. Medulla with rete cord and a single
germ cell above, small cortical remnant below. C. Region showing complete atrophy. (From
Biological Symposia, Vol. IX, Jacques Cattell Press, Lancaster, Pa.)

*^-V:

•4

Fig. 2.8. Sections Uuoiifiii a tian.sforniing o\;ny in an older case. Tlir cortex is extremely
reduced and the medidlary area is well differentiated as a testis. In B, except for the cortical
renmant, the histologic picture is that of a normal testis of intermediate development,
with well defined lobules (Humphrey, 1942; cf. Fig. 2.7).

sal begins it may be confined to local re-
gions of the ovary. Transformation may set
in independently at several sites, resulting
in localized masses of testicular tissue which
are, however, histologically normal (Fig.

2.8). Ultimately all renmants of cortex dis-
ai)i)ear and transformation is complete. Such
individuals are capable of breeding as males
(this depends on the new testis establishing
proper connections with the duct system)

HORMONES IX DIFFERENTIATION OF SEX

89

notwithstanding they have the genotype of
the opposite sex (for a discussion see Hum-
phrey, 1942).

To insure invariable predominance of the
ovary in sex reversal it is usually necessary
to provide a marked advantage in size and
rate of development in favor of the female.
This can be done experimentally by resort-
ing to heteroplastic combinations (Fig. 2.9).
In parabiotic pairs composed of two species
of very different size {e.g., Amhystoma ti-
grinum-Amhy stoma maculatum) and with
a corresponding difference in growth rate,
when the members are of different sex the
larger species is almost invariably domi-
nant (Fig. 2.11; Burns, 1935). When the
large partner is a female the ovaries are
enormously larger than the testes of the
male and are always normal. The testes in
some cases undergo reversal almost from the
beginning of differentiation, and toward
metamorphosis are represented by very
small ovaries which contain a few well de-
veloped ovocytes. However, in most indi-

viduals transformation sets in after consid-
erable differentiation has occurred, the testis
cords becoming hollowed out to form ovarial
sacs (Fig. 2.11) while the cortex persists and
grows rapidly. At metamorphosis males are
either completely transformed or the process
is far advanced. Complete transformation
of this type has also been reported by Wits-
chi (1937) in A. tigrinum-A. jefjersonianum
pairs.

A similar result is obtained when single
gonad primordia are transplanted hetero-
plastically by Humphrey's method (Fig.
2.4) . In individuals bearing gonads of differ-
ent sex, when the ovary is of the larger spe-
cies it is dominant regardless of whether it
belongs to the host organism or was derived
from the graft. Histologically the reversal
process is the same as in parabiotic pairs
(Humphrey, 1935a, b). For fuller discus-
sions of species and racial differences as
they affect physiologic sex dominance and
the capacity of the gonads to undergo re-
versal in different species see Witschi (1934,

B

Fig. 2.9. Heteroplastic combinations uniting diiierent species of salamander {Amhystoma
tigrinum and A. punctatum) which differ greatly in eventual size and rate of growth. A. Ven-
tral view of paired embryos just after operation, showing fusion in the cervical region (punc-
tatum member at left). B. A pair after metamorphosis showing the great difference in size;
the larger animal is the tigrimnii member.

90

BIOLOGIC BASIS OF SEX

1957). Humphrey (1942), and Gallien
(1955).

The progress of sex reversal, and the
mechanism by which the transformation is
effected, can be analyzed histologically only
when it takes place as a secondary process,
after a certain amount of sex differentiation
has previously occurred. In this case both
histologic components are distinctly repre-
sented but the normally recessive compo-
nent seeks to become dominant; ovaries
become testes by regression of the differenti-
ated cortex accompanied by growth and dif-
ferentiation of the medullary element (Figs.
2.8 and 2.10), and testes are converted into
ovaries by the reverse process (Figs. 2.6C
and 2.11A-D). The mechanism is flexible,
however, and there is much variability, even
among individuals in the same experiment,
with respect to the stage at which reversal
sets in, its progress, and its final outcome.
In some cases removal of the dominant go-
nad after reversal is far advanced may be
followed by a second reversal toward the
original sex (Humphrey, 1942).

However, transformation is not always a

secondary process, set in action only after
a certain amount of differentiation has al-
ready occurred; as noted above, when the
ciuantitative disparity between the interact-
ing gonads is sufficiently marked reversal
may proceed from the earliest stages of dif-
ferentiation. In this case the term "reversal"
is less apt since there were no previous his-
tologic steps to be retraced. Transformation
is indicated chiefly by the unbalanced sex
ratios at the end of the experiment; but in
certain cases it is confirmed by character-
istic histologic peculiarities (Burns, 1935,
Fig. 28; Witschi, 1937, Fig. 39). In the het-
eroplastic grafting experiments of Hum-
phrey, on the other hand, direct proof is
available since in many cases the sex of the
grafted gonad, although conforming with
that of the host, differs from the sex of the
donor animal which is reared to provide a
direct control.

Although primary reversal of sex differ-
entiation, as described above, occurs more
readily when a marked disparity in size
leads to early dominance, it may also occur
under conditions which greatly retard de-

H^iH^

Fig. 2.10. Two views of an ovary of Atnbystoma tigrinuin uudcigoing reversal under the
influence of the testes of a male partner. The cortical zone, witli characteristic early ovo-
cytes, is still prominent; however, medullary development is i)roceeding and in tlie region
rei^resonted in B, testis lobules are forming. (From l^ K. Buiiis, J. Exper. Zool., 55, 123-129,
1930; 60,339-387, 1931.)

HORMONES IN DIFFERENTIATION OF SEX

91

^*V^ .^v- _^' A

S-.-' / V

"-^~ .\:!^>

Fig. 2.11. Stages in the transformation of testis to ovary in the male (punci.ii luii ) imni-
ber of an Ambystorna tigrinum-A. punctatum pair, joined in heteroplastic parabiosis (Burns,
1935). yl to D show, at successive levels in the same gonad, the degeneration of the medulla
by vacuolation of the rete canals and lobules, accompanied by persistence and growth of the
cortex. E shows a section through one of the dominant and entirely normal tigrinum ovaries;
a gross picture of these ovaries is seen in F. (From R. K. Burns, Anat. Rec, 63, 101-129, 1935.)

velopinent and delay the beginning of sex
differentiation. This appears to be the ease
in the first parabiosis experiments (Burns,
1925) in which the method of joining inter-
fered with feeding and resulted in a severe
retardation of growth. Such pairs developed
so slowly that sex differentiation was de-
layed for weeks and in many cases for
months. When it eventually took place
members of a pair were almost invariably
of the same sex (c/. Humphrey, 1932) al-
though there was great variation in the size
and stage of differentiation of the gonads.
In this experiment the usual physiologic
dominance of the male was also disturbed,
male-male and female- female pairs appear-
ing in nearly equal numbers.^ The manner

^ The author's interpretation of the results in
this experiment has often been questioned by
Witschi. However, all but a small part of this ma-
terial was subsequently re-examined by Humphrey
(1932) whose study confirmed the original conclu-
sions except for minor details in\-olving at the
most only eight pairs. In Humphrey's opinion
there were certain histologic indications that these

in which extremes of temperatures induce
sex reversal in tadpoles through a differen-
tial inhibitory effect on the medullary or the
cortical component of the embryonic gonad
has been referred to earlier. The above re-
sult may have a similar physiologic basis
if, under prolonged repression, the usually
dominant male gonad should prove to be
more susceptible to unfavorable conditions
than the female.

3. Administration of Steroid Hormones

Since sex hormones of adult type became
available in pure form their effects on the
differentiation of sex have been tested in
many species of amphibia. They are readil}'
administered in two ways, by injecting di-
rectly into the body cavity or, in aqueous so-
lution, by adding them to the water in which
the larvae are reared. The results on the
whole are striking; in certain species there

pairs were originally heterosexual, although trans-
formation had proceeded to the point where a com-
plete reversal was imminent.

92

BIOLOGIC BASIS OF SEX

is complete transformation of ovary to testis
or of testis to ovary, and in some cases the
transformed individuals have been proved
functional and capable of breeding. In other
si)ecies, however, negative, equivocal, and
in many cases paradoxical results have been
obtained by the use of the same substances.
A hormone that completely transforms all
individuals of the opposite sex in one species
may have only a weak or impermanent ef-
fect in another, or no effect at all in a third.
Obviously the gonads of different species
differ greatly in their responses to steroid
hormones. There is also a correlation with
sex. In some species the gonad of one sex
undergoes reversal with relative ease
whereas that of the other is difficult or im-
possible to transform, although increased
dosages are sometimes effective. In certain
species in which the sex chromosome com-
plex is known it is the homogametic sex
that is readily reversible (Gallien, 1955). It
is also clear that experimental conditions
strongly influence the result. The time fac-
tor, that is to say the stage of differentiation
at which treatment is initiated, is obviously
important; in general, the most complete
transformations are obtained when the hor-
mone has been present from the beginning
of the differentiation process. Dosage is like-
wise of great importance and the optimal
dosage varies greatly from one species to
another. A negative response at one dosage
may become positive when the dosage is in-
creased; on the other hand, strong "para-
doxical effects" (stimulation of the charac-
ters of one sex by the hormone of the other)
are often encountered with high dosages
which are absent at lower levels. At present
it is not possible to give consistent explana-
tions for all such contradictory results ; how-
ever, better understanding is gained when
they are classified into convenient cate-
gories.

The effects of male hormones on sex dif-
ferentiation in frogs. On the whole, the most
successful reversals obtained by the use of
steroid hormones have been in frogs of the
family Ranidae, of which some six species
have now been studied with similar results.
Both male and female hormones induce re-
versal of sex in young tadpoles, but male
hormones are more effective and far more
consistent in their effects. In Rana tempo-

raria treatment with testosterone propionate
transforms all genie females into males. The
transformation is complete and permanent;
moreover, transformed individuals are capa-
ble of functioning in their new capacity
(Gallien, 1944). A similar transformation
has been obtained in Rana sylvatica, a phe-
nomenally low concentration of the hor-
mone (1/500,000,000 parts dissolved in the
aquarium water) inducing complete histo-
logic transformation (Mintz, 1948). Com-
parable results have been reported after use
of the male hormone in several other species
of this genus (Table 2.1). In the case of
Rana catesbiana it was found necessary to
"prime" the gonads by simultaneous treat-
ment with gonadotrophin, otherwise they
were unresponsive; when so treated, how-
ever, complete transformation is obtained.
The gonadotrophic substance alone initiates
a precocious differentiation of sex but with-
out any tendency to transformation, serving
only to precipitate the normal differentia-
tion process. Complete and permanent mas-
culinization of females by testosterone pro-
pionate has also been reported recently in a
tree frog, Pseudacris (Witschi, Foote and
Chang, 1958), and a virtual transformation
at the age of metamorphosis in Rhacophonts
(Iwasawa, 1958), indicating that other anu-
ran families may resemble the Ranidae in
their reactions to the male hormone. In
marked contrast, male hormone is without
effect on gonad differentiation in the toad
(Chang, 1955).

The effects of female hormones in the
Ranidae. These are variable and less con-
clusive. The results frequently depend on
dosage and the effective dose may vary
greatly in different species. Estradiol benzo-
ate in weak doses has a slightly feminizing
action in male tadpoles of Raiia temporaria
(Gallien, 1941), but stronger doses produce
complete feminization in an "undifferenti-
ated race" (one in which sex differentiation
of males occurs relatively late) of the same
species, all tadpoles at metamorphosis be-
ing females (Gallien, 1940, 1955) . It is note-
worthy that in this case both genie constitu-
tion (race) and the dosage of the hormone
may be factors in the result. In other ranid
species estradiol, administered in low dos-
ages during the period of sex differentiation,
has a completely feminizing effect ; at meta-

HORMONES IN DIFFERENTIATION OF SEX

93

TABLE 2.1

The ejects of synthetic male and female sex hormones on the differentiation of
the gonads in various species of amphibians.
The cases listed are those in which a complete, or near complete, and histologically norma
transformation was achieved by the age of metamorphosis. In some species the reversal wa
permanent and functional. For details see text.

ACTION OF MALE HORMONE ON FEMALES

SPECIES

INVESTIGATORS

RESULT

COMMENT

RANA TEMPORARIA

GALLIEN 1938, 1944

COMPLETE TRANSFORMATION

PERMANENT AND FUNCTIONAL

RANA SYLVATICA

MINTZ 1948

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

DOSAGE 1/500 000 000

IN ACQUARIUM WATER

RANA PIPIENS

FOOTE 1938

COMPLETE TRANSFORMATION-
TREATMENT FOR 65 DAYS

RANA CATESBIANA

PUCKETT 1939, 1940

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

ADMINISTERED WITH
GONADOTROPIN

RANA CLAMITANS

MINTZ, FOOTE 8 WITSCHI
1945

COMPLETE TRANSFORMATION-
TREATMENT FOR 95 DAYS

SOME PRODUCED SPERM

RANA AGILIS (DALMATINA)

VANNINI 1941, PADOA 1947

COMPLETE TRANSFORMATION

PSEUDACRIS NIGRITA

WITSCHI, FOOTE 8 CHANG
1958

COMPLETE TRANSFORMATION

EFFECT PERMANENT

RHACOPHORUS SCHLEGELII

IWASAWA 1958

TRANSFORMATION ALMOST
COMPLETE

AT THE AGE OF

METAMORPHOSIS

ACTION OF FEMALE HORMONE ON MALES

PLEURODELES WALTLII

GALLIEN 1954

COMPLETE TRANSFORMATION

PERMANENT AND FUNCTIONAL

XENOPUS LAEVIS

GALLIEN 1953

COMPLETE TRANSFORMATION

PERMANENT AND FUNCTIONAL

RANA TEMPORARIA

GALLIEN 1941, 1944

COMPLETE TRANSFORMATION

EFFECT NOT PERMANENT

RANA ESCULENTA

PADOA 1938, 1942

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

AT LOW DOSAGES ONLY

RANA SYLVATICA

WITSCHI 1952, 1953

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

AT LOW DOSAGES ONLY

RANA CATESBIANA

PUCKETT 1939, 1940

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

ADMINISTERED WITH
GONADOTROPIN

BUFO AMERICANUS

CHANG 1955

COMPLETE TRANSFORMATION
AT METAMORPHOSIS

EFFECT NOT PERMANENT j

morphosis all tadpoles are females (Table
2.1). This result has been reported in R.
esculenta (Padoa, 1942), in R. sylvatica
(Witschi, 1951, 1952, 1953), and Puckett
(1939, 1940) obtained the same effect in R.
catesbiana when a gonadotrophin was given
simultaneously with the estrogen. However,
such transformations, although histologi-
cally complete, are not in all cases perma-
nent (Gallien, 1955) ; moreover, the effects
of high dosages of the same hormone may
be quite different, as wdll be shown. For
other species only partial transformations
have been found, as in R. clamitans (jMintz,
Foote and Witschi, 1945) and in R. pipiens
(Foote, 1938). A similar type of incomplete
transformation by the female hormone has
recently been reported in the tree frog,
Pseudacris (Witschi, Foote and Chang,
1958) , again suggesting that in their pattern
of response to sex hormones the Hylidae re-
semble the Ranidae.

Various other anuran species have yielded
divergent results. In the primitive frog,

Xenopus laevis, complete and functional
transformation of males is effected by an
aqueous solution of estradiol benzoate (Gal-
lien, 1953), and in the toad {Bufo ameri-
canus, Chang, 1955) low doses of estradiol
completely transform testes into ovaries
although high doses have little effect. Fi-
nally, in Discoglossus the effect of estradiol
is purely feminizing but the transformation
is incomplete, the males exhibiting all de-
grees of intersexuality without obvious re-
lation to dosage (Gallien, 1955).

Paradoxical ejfects. Thus far emphasis
has been placed chiefly on cases in which
sex hormones have acted in a sex-specific
manner, each type of hormone directly or
indirectly promoting the development of
structures of the appropriate sex, while in-
hibiting or behaving in neutral fashion to-
ward those of the other. Reference has been
made more than once, however, to the fact
that in other cases the effects are just the
reverse of theoretical expectation and op-
posed to the concept of hormones as specific

94

BIOLOGIC BASIS OF SEX

sex-differentiating agents. Anomalous or
paradoxical results of this kind have ap-
peared in experiments with both types of
hormone, and involve not only the gonads
but other sex structures as well.

Such a result was first reported by Padoa
(1936) who found that a crystalline form of
female hormone (Crystallovar) had a strong
masculinizing effect on the sexual differen-
tiation of tadpoles of Bana esculenta, all
gonads developing as testes. This unex-
pected result (the so-called "paradoxical
effect") was confirmed by others and has
been found to be usually associated with
the use of high dosages. In the course of time
it was shown in this and in two other spe-
cies [Rana temporaria, Rana sijlvatica)
that the same hormone (estradiol) may
have diametrically opposite effects when
administered in different dosages. As was
emphasized earlier, low doses have a proper
feminizing action, producing all female in-
dividuals according to theoretical expecta-
tion; with high dosages, on the contrary,
only males are obtained, and at intermedi-
ate levels all individuals become intersexual
(Padoa, 1938, 1942; Gallien, 1941, 1955;
Witschi, 1952, 1953). Indeed, identical
amounts of the same substance may have
opposite effects when different solvents are
employed. Administered in oil the effect on
the gonads is feminizing but in aqueous so-
lution complete masculinization occurs
(Gallien, 1941). This also would appear to
be a dosage effect since in aqueous solution
the rate of uptake is presumably much
faster than in oil. There are no histologic
indications in the above experiments as to
how the paradoxical effect is mediated. But,
although such effects are found in various
ranid species, they do not occur in Dis-
coglossus regardless of dosage, thus empha-
sizing the importance of species differences
in the phenomenon (Gallien, 1955).

Male hormones also produce paradoxical
effects on the gonads and, although the
doses employed have generally been high,
it again appears that the result often de-
pends on the species tested. The same dose
of the same substance may have opposite
effects in different species. Testosterone or
cthinyl-testostcrone in large doses have
strong feminizing effects on the testes of the
salamander Pleurodeles, whereas the devel-

o])ment of the ovaries is retarded but other-
wise unaffected, a typical paradoxical effect
(Gallien, 1950, 1955). Ethinyl-testosterone
has the same effect in Discoglossus, but in
Rana temporaria this hormone has only the
expected masculinizing action. Such differ-
ences in response may arise from differences
in sensitivity on the part of the gonads or
gonad components; a dose which is rela-
tively large for one species may not be so in
the case of another.

The effects of sex hormones in urodele am-
phibians. In urodele amphibians the effects
of sex hormones on the differentiation of the
gonads are perhaps even more variable;
however, as opposed to the situation in the
Anura, it is the female hormones which are
more effective in producing reversal than
the male (Gallien, 1955). In only one in-
stance, the newt Pleurodeles, has a func-
tional transformation of sex been achieved
(Gallien, 1954). In this species prolonged
treatment with estradiol benzoate com-
pletely reverses the differentiation of all
males, some of which become capable of
laying eggs. Varying degrees of transforma-
tion have been reported in other urodele
genera after shorter periods of treatment, in
Amby stoma (Burns, 1938a; Ackart and
Leavy, 1939; Foote, 1941) and in Hynobius
(Hanaoka 1941a). There was great varia-
tion in the timing of treatment and in the
dosages employed in these experiments; the
incomplete character of the reversal may be
due in part to such factors, but the role of
species variability must also be great.

On the other hand the male hormone, in
marked contrast to its dominating role in
the Anura, has but a limited transforming
action in Urodeles. Indeed, it frequently, but
not always, produces paradoxical effects
icf. Burns, 1939c; Foote, 1941; Bruner,
1952). In Pleurodeles, in which the males
arc completely transformed by estradiol, the
effect of testosterone is limited to a severe
inhibition, which affects the gonads of both
sexes but is more extreme in males (Gallien,
1955). Medullary development is almost
completely suppressed, and after an interval
of recovery the vestigial gonads give rise
almost exclusively to rudimentary ovaries,
a result mentioned pi'e\-i()usly in discussing
paradoxical effects. In tliis case, however,
there is clear histologic evidence as to how

HORMONES IN DIFFERENTIATION OF SEX

95

the effect is mediated. Reversal is caused in-
directly by a severe inhibition of mesoneph-
ric development. Since the sex cords which
give rise to the medulla of the testis are de-
rived from the mesonephric blastema, in-
hibition of this tissue prevents their forma-
tion. In the absence of proper medullary
development, the cortical rudiment of the
testis eventually becomes active to produce
an ovary. This, apparently, is another ex-
ample of spontaneous differentiation of the
heterotypic gonad component when released
from domination.

A summary of the effects of steroid hor-
mones in am'phihians. Sex hormones of adult
type, such as testosterone and estradiol,
have effects which vary greatly in different
taxonomic groups of amphibians and also
according to experimental conditions, such
as dosage, timing, and duration of treat-
ment. In many species their effects are spe-
cific to a degree, closely simulating the
effects expected of natural hormones. Histo-
logically complete and in some cases func-
tional transformations of the gonads have
been produced in a number of species, in-
volving two orders and several families
(Table 2.1). Nevertheless, in other species
only partial or temporary reversals are ob-
tained, and negative or even paradoxical re-
sults have come from use of the same hor-
mones. Constitutional differences between
taxonomic units obviously underlie some of
the conflicting results. Sex genotype is also
involved, because in a particular species
reversal may proceed easily in one direction
whereas in the other it is difficult or im-
possible to produce. Following in part Gal-
lien (1955), the results may be tentatively
grouped as follows:

a. In the higher anurans of the family
Ranidae, and perhaps also in the Hylidae,
the male hormone induces complete, and in
many species a permanent reversal of sex.
The action of female hormones on the con-
trary is highly variable; transformation
may be incomplete or unstable, and with
high dosages paradoxical or masculinizing
effects often appear. On the other hand, loiv
doses of the same hormone have in many
cases proper feminizing effects in the same
species. In one species (Rana catesbiana)
complete reversal of sex in both directions
has been obtained.

b. In certain urodeles and lower anurans,
female hormones induce a transformation of
male gonads which may be complete and
stable, as in Pleurodeles and Xenopus, or
partial, as in Discoglossus and various spe-
cies of Ambystoma. The extent to which
partial reversals are attributable to particu-
lar experimental conditions is uncertain.
Male hormones in general are much less ef-
fective, and are prone to induce a paradoxi-
cal inhibition or a feminization of the testis.
These effects have in some cases been shown
to be mediated indirectly, through an in-
hibition of nephrogenesis which suppresses
the differentiation of the medullary sex
cords. The paradoxical effects of female hor-
mones, on the other hand, are in many cases
a matter of high dosages; how such effects
are exerted is unknown but the action is
probably indirect. This point will be dis-
cussed elsewhere in connection with the
problem of paradoxical effects on other sex
structures.

C. SEX REVERSAL IN AVIAN GONADS

1. Orgariization of Avian Gonads

In avian as in amphibian gonads a spe-
cific morphologic basis for sex reversal ex-
ists during early development in the form
of medullary and cortical components which
have the usual potentialities. In birds, how-
ever, the situation is complicated by the
peculiar lateral asymmetry which affects in
some degree the entire genital system and
which is especially pronounced in the fe-
male of most species (Fig. 2.12). The sum-
mary which follows is based primarily on
the chick (for a fuller account see Willier,
1939). In the left embryonic ovary the pre-
ponderance of the cortex is great, even in
the early stages, whereas in the rudimentary
right ovary the cortex is essentially absent,
being briefly represented by a transient
germinal epithelium which disappears even
earlier than that of the testis (Wolff, 1948) .
In fact, the right ovary virtually ceases to
develop at a stage when only medullary
tissue (primary sex cords) has been laid
down. In the male the asymmetry is mor-
phologically less marked but it is expressed
nevertheless in the better development and
longer survival of the germinal epithelium
(potential cortex) on the left testis. These

96

BIOLOGIC BASIS OF SEX

INDIFFERENT STAGE

mwim hwwvm

RIGHT GONAD LEFT GONAD

EFFECTS OF ? HORMONE ON MALE

EFFECTS OFd* HORMONE ON FEMALE

i005mg

Mlko.M

Omg \l'^\i\i\i\llll/l/yf\
LEFT TESTIS

IW<Wl!rrV\ 20mg, A\»Mff//>A
RIGHT OVARY LEFT OVARY

Fig. 2.12. Diagrams sliowing lateral differences in gonad organization in the chick embryo
with respect to the representation of cortical and medullary elements; and differences in
reaction to sex hormones based on these differences. (After N. T. Spratt, Jr., and B. H.
Willier, Tabulae Biologicae, 17, 1-23, 1939). Note the stronger representation of the cortical
component in the left gonad in both sexes, and its influence on the responses of gonads to
hormones in relation to dosage. For details see text.

structural differences are correlated with
different capacities for sex reversal under
experimental conditions, as will appear.^

The experimental study of sex differentia-
tion in birds has been limited largely to two
domestic species, the chick and the duck.
Histologic sex differences first appear in the
gonads of these species around the seventh
and the ninth days of incubation, respec-
tively, but the future pattern of develop-
ment is essentially determined much earlier.
This is shown by the fact that when the
sexually indifferent gonads of chicks are
transplanted at the genital ridge stage to the
choi'ioallantoic membrane of another em-
bryo, they continue in most cases to develop
independently, in accordance with genotype,
giving rise to typical testes or ovaries, and
in the case of gonads of female constitution
to characteristic right or left ovaries as well

° The spontaneous reversal of sex that frequently
follows removal of the dominant left ovary in the
young female chick is an example, and the basis
for this phenomenon lies in the predominantly
medullary ciharacter of the rudimentary right ovary
as described earlier.

(Willier, 1933, 1939j. The same capacity
for self-differentiation has been demon-
strated under the more radical conditions of
isolation. Histologically undifferentiated
gonads of either species when cultured in
vitro differentiate into testes, or into right
and left ovaries of characteristic structure
(Wolff and Haffen, 1952a). In some cases
there is injury to the germinal tissue under
culture conditions; the gonads may show a
reduction in the number of germ cells, or
in some cases complete sterility, but other-
wise the structure is normal (Fig. 2.13).
Duck gonads seem to be more hardy under
conditions of culture than those of the chick,
and in general show better growth and liisto-
logic differentiation.

2. Effects of Administering Pure Hormones

As noted earlier, sex reversal in the gon-
ads of birds was not demonstrated experi-
mentally until \)uve hormones became avail-
able. The first successful experiments were
those of Kozelka and Gallagher (1934) ;
Wolff and Ginglingcr (1935) ; Willier, Gal-

HORMONES IN DIFFERENTIATION OF SEX

97

Fig. 2.13. Histologic difYerentiation in the gonads of duck embryos developing in vitro,
after isolation just at the beginning of sexual differentiation (Wolff and Haffen, 1952a). A.
Normal form and histologic differentiation of the testis in comparison with an ovary develop-
ing under the same conditions (B). C and D show, respectively, the structure of these gonads
under higher magnification. (From Et. Wolff and K. Haffen, J. Exper. Zool., 119, 381-404,
1952.)

lagher and Koch (1935, 1937) ; and Dant-
chakoff (1935, 1936) who introduced steroid
hormones into incubating eggs before the
beginning of sex differentiation. The results
vary in detail but are consistent in the main
outlines; they may be stated briefly, follow-
ing chiefly the reports of Willier, Gallagher
and Koch and of Wolff and Ginglinger.

Female hormones (estrone or estriol) do
not significantly affect the differentiation of
embryonic ovaries but testes are highly
transformed. Because of the better develop-
ment and longer survival of the germinal
epithelium the left testis is more amenable
to reversal than the right. Relatively low
doses convert it into an ovotestis. The distal
ends of the medullary cords become hol-
lowed out into tubular structures like the
medullary cords of the ovary (Fig. 2.14.4) ;
at the same time a zone of cortex develops
peripherally, arising as a proliferation of the
germinal epithelium. A small, unchanged
medullary mass usually persists at the hilus.
But with larger doses even this may dis-
appear and the cortex becomes much thicker.

Such cases are practically indistinguishable
from ovaries. The right testis, however, is
more difficult to transform. In the above ex-
periments it was not greatly modified at
lower dosages ; even when the left testis was
almost completely transformed into an
ovary the right never entirely lost its tes-
ticular character. Because of the poor de-
velopment of the germinal epithelium its
capacity to produce cortex is limited. How-
ever, Wolff (1948) made a special study of
the right gonad in both sexes, assuming that
stimulation of the gonad at an earlier stage,
before regression of the germinal epithelium
can be detected in either sex, might reveal
a greater capacity for cortical differentia-
tion. To insure rapid action a water soluble
form of the hormone was used. In this way
a considerable differentiation of cortex was
obtained. The importance of a persistent
search for the proper experimental condi-
tions is again demonstrated.

Male hormones, on the other hand, are
less effective in transforming the embryonic
ovaries of birds. Again lateral differences in

98

BIOLOGIC BASIS OF SEX

B^\^

Fig. 2.14. Hi.Ntolojiic m'X rraiisl'ormation in the te.sles uf cluck eiiil)iyu« lrcate(i with fe-
male sex hormones. A. Ovotestis produced from a left testis by treatment with a dose of
0.2 mg. of estriol. Note the presence of a thick cortex peripherally, and the reduction of the
testicular tissue to a hilar mass of medullary cords. The intervening highly vacuolated tissue
is characteristic of ovarian medullary cords. B. Section through the cortex and medullary
region of a left tcslis coinplolcl}- transfoinied into an ovary by a dose of 2.0 mg. of estrone.
Note the thick cdilrx (aboxc) coxcicd \>y a liciniinal epithelium, and the loss of structure
in the medullary ii'gioii Ixlow. (From B. H. W'illior, in Sex and Internal Secretio7is, 2nd ed.,
The Williams & Wilkins Co., 1939.)

reaction are found due to the different liisto-
logic constitution of the right and left pri-
mordia, and the results also differ when
different forms of the male hormone are em-
ployed. After treatment with testosterone
the cortex of the left ovary is reduced in
thickness and shows degenerative changes;
however, it does not entirely disappear. At
the same time there is hypertrophy of the
medulla and some of its cords acquire the
solid structure of testis cords. The result is
an ovotestis. Because of its predominantly
medullary constitution the rudimentary
right ovary is converted superficially into
a testis-like gonad. The medullary mass
hypertrophies and some of the cords arc
transformed into testis cords. In general,
larger amounts of hormone are required to
transform ovaries than for the conversion
of testes (for summaries see Willier, 1939;

Wolff, 1950). After hatching there is a tend-
ency for experimentally modified gonads to
revert toward the original sex (Wolff, 1938) .
Similar conditions of reversal have been
produced in the embryonic gonads of ducks
by hormone treatment (Lewis, 1946) .

The problem of the jiaradoxical action of
hormones presents itself again in the case of
a^•ian gonads. Certain male hormones of
ui'inary origin (androsterone, dehydro-an-
drostcrone) have a marked feminizing ef-
fect, like that produced by t'cnuile hormones.
In relatively large doses both substances
induce cortical differentiation in testes, es-
pecially the left, which may be transformed
into an" ovotestis (Willier, 1939; Wolff 1938).
Other androgenic substances have like ef-
fects, but again it has been shown that they
ai-e not jiroduced by low dosages. However,
as the concentration of the hormone is raised

HORMONES IX DIFFERENTIATION OF SEX

99

the degree of intersexiiality and the number
of intersexual gonads steadily increase
(Wolff, Strudel and Wolff, 1948). Since
various accessory sex structures also show
paradoxical reactions the problem will ap-
pear again.

3. Effects of Grafting Gonads into the Coe-
lomic Cavity

The demonstration that steroid sex hor-
mones are capable of inducing sex trans-
formation in avian gonads led to a reinvesti-
gation of earlier failures to obtain reversal
by means of chorioallantoic grafting. Even-
tually it was shown that the difficulty was
largely a matter of the method. When gonad
primordia are transplanted directly into the
coelomic cavity of a host embryo of different
sex, varying degrees of transformation, or
even a virtual reversal, are obtained.

The first experiments of this type were
only a partial success (Bradley, 1941). Em-
bryonic gonads of the chick and the duck,
isolated at 96 to 120 hours of incubation,
were inserted into the body cavity of host
embryos through a small slit in the somato-
pleure, using both homoplastic and hetero-
plastic host-graft coml)inations. In the case
of the chick, host embryos were always con-
siderably younger than donors. In all cases
the grafts underwent primary sex differenti-
ation in accordance with genotype, and only
a small minority showed specific modifica-
tions. The results were rather inconclusive
because in no case were the changes of a
conspicuous character, and there was great
variability in the growth and differentiation
of the grafts, making it difficult to assess
the significance of the modifications. In
some cases changes of the same type ap-
peared in the gonads of the host embryo.

The modifications noted by Bradley fall
into three main classes. (1) Vacuolation of
the medullary cords of testes growing in fe-
male hosts (in a few cases) caused them to
resemble the hollow medullary cords of
ovaries. (2) In some cases ovaries growing
in male hosts developed solid medullary
cords of male type. (3) Rudimentary right
ovaries (always testis-like in character) had
a tendency to become enlarged when grow-
ing in male hosts. A similar effect was some-
times seen in the right ovaries of hosts bear-

ing testis grafts. No ready explanation was
available for the inconstant occurrence of
these effects or for their quantitative varia-
bility, because no clear correlation was
found between the degree of modification
and the relative proximity of the interacting
gonads. Finally, similar changes appeared in
a few cases when the host-graft combina-
tions involved the same sex; consequently
the specific character of the modifications
was left in doubt.

The matter was clarified by the experi-
ments of Wolff (1946) who used a modifica-
tion of the method with better results. Grafts
taken from older embryos (6 to 11 days)
were implanted into hosts of about 50 hours
of incubation. Under these conditions a
striking transformation of gonad differentia-
tion was obtained in the host, and in addi-
tion the developing gonaducts [q.v.) were
strongly modified. Ovaries grafted into male
hosts induced differentiation of cortex on
the left testis to such an extent that it some-
times approached the structure of an ovary.
The right testis (which was usually more
distant from the graft) was less modified
but its growth was inhibited. On the con-
trary, implantation of a testis in the same
manner produced no important effect on the
differentiation of the ovaries of the host but
the development of the Miillerian duct was
strongly inhibited indicating that the graft
is endocrinologically active. In their histo-
logic character the effects of gonad grafts
are similar to those produced by crystalline
hormones, differing only, as a rule, in being
more localized in relation to the position of
the graft. The demonstration in this experi-
ment that, physiologically, the ovary is the
dominant gonad in birds is consistent with
the earlier observation that relatively larger
doses of pure hormones are required for the
transformation of ovaries than for testes.

These positive results after so many fail-
ures suggested that the ineffectiveness of
chorioallantoic grafts in the earlier experi-
ments was possibly a matter of hormone
production, failure of the graft to maintain
a sufficient level of the hormone in the blood.
This view is substantiated by the later ex-
periments of Huijbers (1951) who showed
that multiple grafts of well differentiated
testes on the chorioallantois have marked

100

BIOLOGIC BASIS OF SEX

effects on the accessory sex structures of the
host, similar to those induced by intra-em-
bryonic grafts.

4. Sex Reversal in Vitro

More recently the technique of culture in
vitro has been employed by Wolff and his
collaborators with great success to study the
development of embryonic gonads, and it
has been possible to produce typical reversal
of sex differentiation in vitro by two meth-
ods. Prospective ovary and testis of the
chick or duck, isolated at the very beginning
of sex differentiation and placed in close
contact in the culture dish,^ become firmly
fused, facilitating the transmission of hu-
moral influences. As in the case of gonad
grafts in the coelomic cavity, the ovary un-
der such conditions proves to be the domi-
nant gonad (Wolff and Haffen, 1952b). It
readily induces cortical differentiation on
the testis, which becomes an ovotestis, and
may even approximate closely the structure
of a normal ovary of the same age (Fig.
2.15). The same type of transformation oc-
curs in testes after introduction of estradiol
benzoate into the culture medium.

With respect to the histologic character
of the reversal process, the resemblances be-
tween the effects of gonad grafts implanted
in the body cavity, crystalline hormones in-
jected into the whole organism, and the re-
sults of the same procedures applied to iso-
lated gonad primordia in vitro are
extremely close. The in vitro studies demon-
strate again the autonomous character of
the differentiation process, and its flexibility
in the presence of extraneous hormones is
shown to be independent of the organism
as a whole. Hormones in vitro evidently act
directly on the gonad mechanism.

1). TIIK PROBLEM OF SEX REVERSAL
IN MAMM.\LIAN GONADS

1. Bisexual Potentialities in the Emhrijoiiic
Gonads of Mam^nals

In marked contrast with the striking ef-
fects of steroid sex hormones on the differ-
entiation of the gonads of birds and various

" Combinations of gonads in the culture dish
must initially be made at random but the other
gonad of each donor is cultured separately in order
to establish its sex.

species of amphibians has been the failure
thus far to ol)tain comparable effects in
mammalian embryos with the exception of
a single species, the North American opos-
sum, a marsupial. Essentially negative re-
sults have been reported for a number of
species of placental mammal, in which preg-
nant females were treated with relatively
large dosages of sex hormones during the
period of sex differentiation. Experiments
of this type were carried out in the rat by
Greene, Burrill and Ivy (Greene, 1942), and
in the guinea pig (Dantchakoff, 1936, 1937),
the mouse (Turner, 1939, 1940; Raynaud,
1942j , the rabbit (Jost, 1947 a) , the hamster
(Bruner and Witschi, 1946; White, 1949),
and the monkey (Wells and van Wagenen,
1954). With the single exception of the
opossum (to be described later) the modi-
fications induced are minor in character and
are of three types: (1) a general retardation
of growth and development of the gonads,
without obvious signs of sex reversal, which
occurs in both ovaries and testes and may
be produced by either type of sex hormone;'^
(2) a variable degree of hypertrophy of the
medullary elements of ovaries after treat-
ment with male hormone, reported in only
a few cases (Dantchakoff, 1939; Jost, 1947a;
Wells and van Wagenen, 1954) ; and (3) the
occasional persistence of localized patches
of germinal epithelium on the surface of well
differentiated testes, a condition which
sometimes appears after treatment with
either type of sex hormone. IVIinor changes
of this character have not generally been
accepted as convincing evidence of sex re-
versal.

Notwithstanding this array of negative
findings, the failure of the embryonic gon-
ads of placental mammals to respond defi-
nitely to sex hormones can hardly be at-
tributed to an inherent lack of bisexual
potentiality. During the early stages of their
development they show, histologically, the
same evidences of bisexual structure as the
gonads of other vertebrates, although typi-
cally the bisexual phase is of relatively
brief duration and the recessive sex com-

" This effect is of common occurrence and is
best explained as a depression of the gonadotrophic
function of the anterior pituitary, a mechanism
whicli is well established in adult organisms
(Moore and Price, 1932).

HORMONES IX DIFFERENTIATION OF SEX

101

.»■

c'tMi

Fig. 2.15. Sex transformation in the testis of the duck, isolated in vitro at the beginning
of sexual differentiation and cultured in close contact with an embryonic ovary (Wolff and
Haffen, 1952b). A. The ovary of such a combination, showing the thick covering germinal
epithelium and the vacuolated condition of the medullary region. This young ovary is es-
sentially normal in structure. B. Intersexual condition induced in the testis under the in-
fluence of the ovary. The heavy germinal epithelium representing the cortex is as well
developed as in a normal ovary ; the meduUaiy region retains largely the compact structure
of a testis, but signs of vacuolation are appearing. This gonad is an ovotestis. C and D
represent, respectively, the ovary and the completeh' transformed testis in another experi-
ment. The two gonads in this case show almost identical structure, featuring the thick cortex
and vacuolated medulla of an ovary. In these experiments the other testis, cultured alone,
developed normal testicular structure. (From Et. Wolff and K. Haffen, Arch. x\nat. microscop.
et Morphol. exper., 41, 184-207, 1952.)

102

BIOLOGIC BASIS OF SEX

Primary sex cords

Rete cords

Germinal epithelium

A

Germinal epithelium

Medullary cords

(primary sex cords)

Cortical cords ^Germinal epithelium

(secondary sex cords)

Fig. 2.16. Diagrams illustrating schematically the main features of gonad differentiation
in amniote embryos with reference to the origin of the medullary and cortical components.
A. Origin of the primary sex cords (medullary cords) from the germinal epithelium. B.
Gonad at the indifferent stage of sexual differentiation ; the well developed primary sex cords
i-epresent the male or medullary component, whereas the germinal epithelium represents,
potentially, the cortical component. C. Differentiation of a testis consists in the further
development of the primary sex cords, and the reduction of the germinal epithelium to a
thin, serous membrane, accompanied by development of the tunica albuginea. D. Differen-
tiation of an ovary consists in reduction of the primary sex cords to medullary cords of the
ovary, whereas the cortex is formed by continued development of cortical cords from the
germinal epithelium.

poneiit is often weakly represented (Fig.
2.16). In the ovary, medullary cords repre-
senting the male component are present but
tend to become vestigial in most species.^
In the embryonic testis the germinal epi-

'^ Notable exceptions should bo mentioned in the
case of certain species such as the mole (Godet,
1950) and the desman or "water shrew" (Peyre,
1955) in which, as a normal condition, the ovarian
medulla is so strongly developed as to resemble
a testis, and so active physiologically as to pro-
duce strong masculinization of many parts of the
genital tract. A somewhat similar development
and hypertrophy of the medullary component also
occurs in the fetal ovary of the horse (for the lit-
erature on this unusual condition see Cole, Hart,
Lyons and Catchpole, 1933; Parkes, 1954).

thclium usually disapi)ears early, and with
its involution the potentiality for cortical
development is permanently lost. On the
other hand, hermaphroditismus verus not in-
fi'ec|iu>ntly occurs as a developmental anom-
aly iu many mammalian species (including
man), indicating the existence of a basic
bipotentiality. As an example, an extensive
literature dealing with this subject in ro-
dents has recently been summarized by Hol-
lander, Gowen and Stadler (1956) and Kirk-
man (1958). Also it must be remembered
that the classsical example of an embryonic
gonad transformed by the action of a sex
hormone is found in the freemartin. In some

HORMONES IX DIFFERENTIATION OF SEX

103

freemartin gonads morphologic transforma-
tion may be extreme, although the resulting
testis-like structure is histologically abnor-
mal and is almost invariably sterile (Wil-
lier, 1921 ).»

2. Bisexual Potentiality in the Embryonic
Ovary of the Rat

One of the best known cases illustrating
a well marked capacity for bisexual differ-
entiation in a mammalian gonad is provided
by the embryonic ovary of the rat. The gon-
ads of rat embryos have been isolated at
various stages, both before and during the
period of histologic differentiation, and
transplanted to various locations in adult
hosts of both sexes, normal and castrate,
beneath the capsule of the kidney (Buyse,
1935; Mclntyre, 1956), subcutaneously
(Moore and Price, 1942), to the omentum
(Holyoke, 1949) and into the anterior cham-
ber of the eye (Torrey, 1950). In general,
differentiation of the transplanted gonad
proceeds without reference to the sex or the
hormonal status of the host (certain minor
exceptions will be noted later) ; however,
there is a great difference in the behavior of
testis and ovary after transplantation with
respect to their capacity for autonomous
differentiation. There is virtual agreement
among all investigators that the testis pri-
mordium from the beginning of its develop-
ment possesses a remarkably stable organi-
zation, and develops more or less normally
in the various foreign environments, even
when isolated before the beginning of histo-
logic sex differentiation.^*^ The case of the
ovary is entirely different; its organization
appears to be extremely labile and it is in-
capable of fully autonomous development
until a relatively late stage of differentia-
tion, after a well formed cortex is present.
Before this stage (w^iich according to Tor-
rey is reached about the 17th day of devel-
opment) ovaries in a high percentage of
cases either do not develop at all, or are

^ For an important exception in which a free-
martin testis is well supphed with germ cells and
essentially normal in appearance see Hay (1950).

" Torrey considers that the self-differentiating
capacity of the testis probably dates from the lay-
ing down of the early gonadal blastema, i.e., the
material of the primary sex cords, which occurs as
early as the eleventh day of development.

prone to undergo spontaneous reversal, due
apparently to incapacity of the prospective
cortical component to develop effectively in
abnormal tissue environments. On the other
hand, the medullary component of the trans-
planted ovary suffers no such handicap and
frequently assumes the lead in development.
The embryonic ovary of the rat thus pro-
vides a flexible system for the study of the
morphogenetic capabilities of the cortical
and medullary components at different
stages of development and under different
experimental conditions.

When transplanted early in development
prospective ovaries may give rise (Buyse,
1935) to structures of four types: (1) poorly
developed grafts of indeterminate sex, (2)
atypical ovaries of retarded development,
(3) ovotestes in which both sex components
are readily identifiable, and (4) rudimen-
tary testes. As a group, ovaries are ad-
versely affected by transplantation. Some
fail entirely to develop the specific struc-
ture of gonads (type 1, above) and those
that do give rise to ovaries that are greatly
retarded (type 2). On the other hand, the
medullary component of the prospective
ovary resembles the testis in possessing con-
siderable powers of self-differentiation.
Thus in many cases the two components de-
velop together, resulting in an ovotestis; in
still others the cortical element fails com-
pletely to survive and the medulla alone
develops, giving rise to a rudimentary testis.
The development of types 3 and 4 is favored
by the fact that cortical differentiation is
almost always severely repressed. This may
have the effect of releasing the medullary
element from an inhibition normally im-
posed by the dominant cortex. In all cases,
however, the phenomenon of reversal was
found to be unrelated to the sex of the host]
it seems to occur spontaneously, as it were,
in consequence of a disturbance in the nor-
mal balance between cortical and medullary
systems.

A similar behavior is seen wdien entire re-
productive tracts of rat embryos, including
the gonads, are transplanted subcutaneously
into hosts of various ages, male or female,
and into castrate hosts of both sexes (Moore
and Price, 1942). Again it was found thai
the sex or the hormonal status of the host
has no apparent influence on the result.

104

BIOLOGIC BASIS OF SEX

Testes develop normally except that in cas-
trate hosts there is some hypertrophy of the
interstitial tissue, presumably in response
to the gonaclotrophin of the host (a phe-
nomenon also reported by Jost, 1948b I.
Again many prospective ovaries give rise to
gonads in which both cortex and medulla are
well differentiated, the hypertrophied med-
ullary cords sometimes approaching the
structure of testis tubules. Cells resembling
the interstitial cells of the testis are also
found around these transformed medullary
cords in grafts developing in castrate hosts.
On the whole the ovarial cortex is better
developed than in the experiments of Buyse,
because perhaps the gonads were usually
older and better differentiated at the time
of transplantation.

Similar forms of develojiment are found
when embryonic gonads are transplanted to
the omentum of adult hosts (Holyoke,
1949). Testes develop in a virtually normal
manner regardless of the sex of the host;
this author, however, describes certain ef-
fects which appear relatively late, after the
testis has acquired its characteristic tubular
structure. These are: (1) repression of tu-
bule growth in some cases, with degenera-
tive changes, a condition which was ob-
served only in grafts growing in female
hosts; (2) an increase in the amount of in-
terstitium present, similar apparently to
that reported by Moore and Price. Trans-
planted ovaries display the same variability
as in the foregoing experiments; cortical
development is adversely affected and some-
times fails altogether. In all cases in which
cortical structure could be well identified
there was also more or less hypertrophy of
the medulla, sometimes to the point where
the gonads were classified as ovotestes. This
latter condition was reported only in male
hosts, and this is perhaps the only change
that might be interpreted as a reversal of
sex conditioned by the sex of the host. On
this point the findings differ from those of
Buyse, of Moore and Price, and of Torrey,
all of whom reported similar changes but
without relation to the host's sex.

A further study of the problem was made
by Torrey (1950). In his experiments the
embryonic gonads were transplanted to the
anterior chamber of the eye, using hosts of
various ages and of both sexes. He con-

firmed the main conclusions of Buyse and
of Moore and Price (Holyoke's study was
not then available for consideration),
namely, that regardless of the stage at
which the primordium is isolated, testes are
capable of an autonomous and virtually
normal development, irrespective of the type
of host in which they develop, whereas
ovaries vary greatly in the state of differen-
tiation attained, depending on the stage of
development at which they are isolated.

Torrey paid particular attention to the
importance of developmental age in the fate
of grafted ovaries. When transplanted before
the appearance of a definite cortical zone
(zone of secondary sex cords), prospective
ovaries show little capacity for development
of cortex; on the contrary (as found by
l^revious workers), there is a marked tend-
ency to hypertrophy of the medullary com-
ponent, leading in some cases to testis for-
mation. This tendency is not influenced by
the sex of the host; rather, it seems to be
inherent in the state of organization of the
primordium at the time of transplantation.
In the young ovary the medullary compo-
nent (primary sex cords) is already in
existence; the cortex does not appear as a
discrete tissue until much later, and only
after a well defined cortex is present is the
gonad capable of development as an ovary.
The fate of the transplanted ovary appears,
then, to be primarily a iiiatter of the self-
differentiating capacities of the elements
already formed and present in the primor-
dium at the time of its isolation. The de-
velopment of these elements is influenced
also by the temperature of the eye chamber,
a point not directly involved in the present
discussion.

In the foregoing experiments it has been
emjihasized that the hormonal environment
provided by the host seems to have no im-
portant influence on the sexual differentia-
tion of the transplanted gonads (for a par-
tial exception as noted above see Holyoke) .
The question thus arises whether the hor-
mones of embryonic gonads might be more
effective. An answer to this question was
sought by Mclntyre (1956) who trans-
planted the eml)ryonic testis and ovary to-
gether beneath the capsule of the kidney of
adult castrate hosts of both sexes. The gon-
ads were ])laccd in close contact in order

HORMONES IN DIFFERENTIATION OF SEX

105

to determine whether hormones or other
diffusible substances might be produced,
capable of modifying the differentiation
process. As a control procedure, ovaries were
transplanted alone, or in association with
nongonadal tissues, into noncastrate male
hosts.

The results in most respects correspond
with those already described. The testis was
found to develop normally regardless of the
sex of the host or the presence of a contigu-
ous ovary. The behavior of grafted ovaries
differed, however, according to whether they
were associated with an embryonic testis,
or developed alone or with other tissues
(control operations). In the jEirst case the
ovaries were strongly modified along the
lines previously described. Some differenti-
ated poorly or hardly at all, others showed
fairly good development of the cortex with
some primary follicles, but in the medullary
area tubular structures resembling testis
tubules were found. Still others had a few
well formed follicles in the cortex, but again
tubular structures were present in the me-
dulla which sometimes contained ovocytes.
The two last mentioned categories would
seem to correspond to the "ovotestes," or
ovaries with "transformed medullary cords"
described by previous writers. In contrast,
however, ovaries grafted alone, or with non-
gonadal tissues, were found to differentiate
in an almost normal fashion. It is at this
point that the findings depart from those
of other investigators. The conclusion was
reached that the modifications observed in
ovaries associated with embryonic testes are
due to a substance produced by the testis,
and considered to be of the nature of a
medullary inductor.

However, both with respect to the sever-
ity of cortical inhibition and the degree of
masculinization of medullary structures,
these modified ovaries do not appear to
differ significantly from those described by
previous investigators when ovaries were
grafted alone. The significance of the results
rests then upon failure to obtain similar
changes in the control ovaries. The age and
state of differentiation of the control gonads
at the time of transplantation is important.
They were from donor fetuses of 15 days
development, and although older than any
used by Buyse they were still within the

period during which Torrey found the ovary
to be extremely labile in its differentiation.
According to Torrey only a small proportion
of ovaries aged 15 to 16 days developed as
such (5 of 17 cases) whereas at an age of
17 days — after the cortical zone is estab-
lished — ovaries are obtained almost without
exception (10 of 11 cases). Further experi-
ments are needed to clarify this matter.

In all of the foregoing experiments there
was general agreement (for a partial ex-
ception see Holyoke) that the hormonal en-
vironment provided by adult hosts of both
sexes, intact or castrate, has no important
influence on the sexual differentiation of the
transplanted gonads, even in the case of
ovaries which are still in a labile state. The
reason for this is not clear. Possibly it is a
matter of insufficient concentration of the
host hormone (compare the case of the
chick, p. 99), or it may be that after trans-
plantation a certain interval, perhaps a crit-
ical one, elapses before vascularization
makes the graft accessible to the hormones
of the host. On the other hand, it has been
pointed out that in the embryos of placental
mamtnals pure hormones thus far have pro-
duced no significant changes in the dif-
ferentiation of the gonads. The cjuestion
remains whether there is an essential differ-
ence between the sex hormones of adults and
the hormones or sex-differentiating sub-
stances elaborated by embryonic gonads.

3. Experimental Transjormation of the Tes-
tis in the Opossum

Up to now the clearest experimental dem-
onstration of sex reversal in the gonad of
any mammal, and the only one to be pro-
duced by a steroid hormone, has been ob-
tained in the gonads of young opossums
(Didelphis virginiana) . In this species the
embryonic testes, if taken in time, are read-
ily transformed into ovotestes or even into
"ovaries" of remarkably normal histologic
structure by the action of estradiol dipro-
pionate (Burns, 1950, 1955a, 1956b). The
hormone is administered at short intervals,
beginning at a stage of development corre-
sponding to stage B in Figure 2.16. This is
the condition found at birth in litters born
at stage 34 of McCrady's series (McCrady,
1938). It is characterized by the presence
of well developed primary sex cords which

106

BIOLOGIC BASIS OF SEX

are just in process of separation from the
overlying germinal epithelimii. The germi-
nal epitheliimi, however, is still present as a
layer of low, columnar cells and at this stage
represents, potentially, the female compo-
nent of the young testis. Normally the ger-
minal epithelium does not survive long after
birth. In the course of the first day of post-
natal life irreversible changes occur which
lead to its rapid involution. It is the pres-
ence at stage 34 of a viable germinal epi-
thelium which makes possible the subse-
quent conversion of the testis into an ovary,
since it is this layer which must produce
the cortical zone of the transformed gonad
by the proliferation of secondary sex cords.
In so doing it plays precisely the same role
as the corresi)onding layer in the develop-
ment of the ovary.

Treatment with estradiol dipropionate has
been carried out over varying periods up to
an age of 30 days postpartum or somewhat
longer, after which survival becomes diffi-
cult (Burns, 1939b). At the present time
46 male fetuses have been studied histologi-

cally, comprising all the surviving males of
13 litters. Without exception, every speci-
men shows histologic modifications of the
type described below, the stage of transfor-
mation attained varying only with the
length of treatment. The process of trans-
formation consists at first of a gradual in-
hibition and suppression of testicular differ-
entiation, accompanied by persistence of the
germinal epithelium. At the same time the
differentiation of the interstitial tissue is
severely repressed (compare A and B, Fig.
2.17). Atrophy of the interstitial tissue has
also been described by Raynaud (1950) in
the testes of mouse embryos treated with
estrogen. After an interval (which varies in
different experimental litters and is appar-
ently influenced by dosage) the germinal
e})it helium again becomes active, producing
secondary sex cords which form a cortical
zone of varying thickness depending on the
length of treatment (Fig. 2.181.

The first essential in obtaining transfor-
mation of the testis is the timing of the first
treatment, which must not be later than

Fig. 2.17. The effects of female hormone on differentiation of the testis in young opossums.
A. Normal testis about 10 days after birth. Note tlie thin, .serous character of the epithehum
covering the testis (originally the germinal epithelium), the presence of a distinct tunica
albuginea, the prominent testis cords (prospective tubules), and the richly developed inter-
stitium. B. Testis (at somewhat higher magnification) of a young male aged 14 days, modi-
fied by the action of the female hormone estradiol dipropionate. Note the greatly reduced
condition of the testis cords and interstitial tissue, the thick, spongy character of the tunica
albuginea, and especially the survival of the germinal epithelium long after the stage at
which it normally undergoes involution.

HORMONES IX DIFFERENTIATION OF SEX

101

Fig. 2.18. A. Testis of an opossum aged 20 days, converted into an ovotestis by the action
of estradiol dipropionate. Internally the reduced and disorganized medullary region is seen,
separated from the external cortical zone by the well defined, fibrous tunic layer. The large,
irregular ca\ity in the upper center, lined by a heavy eiuthelium, is the rete testis. B. De-
tail at higher power of the cortex, showing the structure of the cortical cords (which are
sterile) and the highly developed germinal epithelium.

stage 34 if the germinal epithelium is to be
preserved. In earlier experiments, in which
treatment was begun after stage 35, there
were no significant effects on the differen-
tiation of the gonads, even in cases where
almost complete transformation of the ac-
cessory sex structures had occurred. It is
now evident that in these experiments the
first application of the hormone came too
late to prevent involution of the germinal
epithelium, thus precluding development of
a cortex. Also of importance is the dosage,
which must be kept at a low level to secure
a good result. This point is crucial for the
survival of germ cells in the developing cor-
tex. High dosages always result in com-
plete sterilitij of the cortical zone, even when
this layer is otherwise well developed (Fig.
2.18). With lower doses, however, germ cells
are found in limited numbers in the cortex,
sometimes sparsely scattered, sometimes in
small groups, and not infrequently these
cells display the cytologic characters of
young ovocytes (Fig. 2.20). Often there is
a considerable growth of the cytoplasm and
well formed primordial follicles are seen
(Fig. 2.20B).

Treatment with relatively low doses of
estradiol (of the order of 0.2 to 0.3 fxg. per
day) from birth to an age of 20 days pro-
duces a remarkable transformation of the
testis, which retains hardly any normal fea-
tures (Burns, 1956b). Rather, it presents
the appearance of a somewhat atypical
ovary (Fig. 2.19B). The only remnant of
testicular structure is a small, central nodule
at the junction of the rete canals, and the
massive cortical zone is covered externally
by a thick germinal epithelium. The cortex
of the transformed testis contains germ cells
in considerable numbers, including a few
large ovocytes. Views of cortical areas in
gonads of this group are shown in Figure
2.20.4 and B. In more recent experiments,
using still lower doses and a longer period
of treatment (thus far the longest experi-
ment has extended to an age of 33 days post-
partum ) , the result is even more striking,
the structure of the cortex in many cases ap-
proximating that of normal ovaries. Always,
however, certain remnants of testicular
structure ]icrsist in the medullary region
(Fig. 2.21, compare A and B). The number
of ovocytes in the cortex is enormously

108

BIOLOGIC BASIS OF SEX

Fig. 2.19. A. The testis of a normal male opossum aged 20 days for comparison with that
of another male, B, treated for 20 days with a low dosage of estradiol dipropionate as de-
scribed in the text. Only a remnant of testis structure survives as a nodular mass in the
medullary region, representing straight tubules at the point of junction with the rete canals.
Note the well developed cortical zone witli numerous germ cells and a heavy germinal epi-
thelium.

greater than in the preceding experiment.
It is not clear whether this is due mainly to
the lowered dosage or to what extent it is a
result of multiplication of ovogonia present
in smaller numbers in the younger gonads
(see Fig. 2.195). In any case, dosage in
some manner influences the survival and
multiplication of the gonia. Although the
cortex of the transformed testis is always
well developed, there is great variability in
the extent to which testicular structures
have survived in the medullary zone. Some
gonads exhibit well preserved male sex cords
and present the picture of a typical ovotes-
tis, whereas in others (Fig. 2.21B} only
traces of the male elements remain in the
form of degenerate sex cords and patches
of fibrous tissue. In these cases transforma-
tion is all but complete.

Since this work is still in progress inter-
pretation must be tentative. It seems that
the primary effect of the female hormone
is a strong inhibition of the testis, affecting
both the primary sex cords and the inter-
stitium. Both influences are apparent at an
early stage (Fig. 2.17). Inhibition of testicu-

lar differentiation i:)resumably i-)ermits sur-
vival of the germinal epithelium, to be
followed later by a renewal of activity pro-
ducing secondary sex cords and the cortex.
This course of events may simply be the
result of release from an inhibition nor-
mally imposed by the differentiating testis.
Counterparts are seen, for example, in the
spontaneous development of the medullary
component in transplanted rat ovaries when
cortical differentiation is interfered with, in
the development of the rudimentary cortex
in Pleurodeles after the medulla has been
suppressed by testosterone (p. 94), or in
the development of the heterotypic sex com-
ponent after castration in the toad or the
newly hatched chick. In the light of cur-
rent knowledge regarding the secretory
capacity of the embryonic testis (for dis-
cu.ssions see Jost, 1953, 1957; Burns,
1955b, and later in this chapter) it is proba-
ble that the primary condition for survival
of the germinal epithelium is suppression of
the interstitial tissue; cortical differentia-
tion is presumably the consequence of es-
cai)e from an inhibition normally exerted

HORMONES IX DIFFERENTIATION OF SEX

109

fel!t".^*.Z '" WS *3i?^~ k'

l't%^

.♦3^"*

*A-. \"%.

^Mj

Fig. 2.20. A. View at higher magnification (X 1000) of the cortex of ;i traiisiorme.l te.stis
containing gonia, and other germ cells showing the early meiotic prophase stages of young
ovocytes. B. Cortex of another transformed testis of the same experimental group showing
the formation of primordial follicles (X 1000).

by the testis hormone. For this there is no
direct evidence; however, in typical ovo-
testes, with a well developed cortex, the
tubular elements may also at times be very
well preserved but the interstitium is de-
generate. This interpretation does not re-
quire positive stimulation by the female

hormone to promote cortical differentiation,
but it does not exclude the possibility that
this may occur. The germinal epithelium of
transformed testes is strongly hypertrophied
in comparison with that of normal ovaries
of the same age (Figs. 2.18B, 2.195, and 20) ,
a condition which is seen also in the ovaries

no

BIOLOGIC BASIS OF SEX

Fig. 2.21. A. The normal ovary of a young female aged 30 days; note the highly developed
cortex and the absence of conspicuous structures in the medullary area. B. Transformed
testis of a young male treated with estradiol dipropionate for a period of 33 days ; for details
see text. Observe the remarkable development of the cortex associated, however, with dis-
tinct remains of testicular structure in the medulla.

of females receiving estradiol. Also, preco-
cious growth of a certain number of follicles
commonly occurs in the cortex of trans-
formed testes (Figs. 2.195 and 2.20). This
effect, however, may be exerted indirectly
in response to gonadotrophic stimulation.

It should be noted that in the only other
case of an embryonic mammalian gonad
transformed by hormone action, that of the
freemartin, the reversal involves the con-
version of ovary to testis. In the opossum
the situation is reversed. In those cases of
comjjlete, or near complete, transformation
in amphibians, in which the sex chromosome
complex has been determined, it appears
that it is always the homogametic sex that
is readily transformed (Gallien, 1955). This
generalization would seem to apply also in
birds and in the case of the freemartin. The
opossum, however, is certainly an exception;
the male in this species is heterogametic
(Painter, 1922; Tijo and Puck, 1958; Gra-
ham, 1956). In certain fishes at least sex
genotype is apparently of no conseciuence
.-^ince functional sex reversal proceeds
eciually well in either direction (Yamamoto,
1953, 1958).

V. The Role of Hormones in the

Development of the Accessory

Sex Structures

The heterogeneous character of the vari-
ous structures comprising the genital com-
plex of the embryo has been previously
emphasized as providing a basis for great
variability in their behavior under experi-
mental conditions. On the basis of embry-
onic origin and morphologic relationships
the accessory sex structures fall into three
principal groups: (1) the embryonic sex
ducts and related structures which are taken
over from the primitive nephric system; (2)
derivatives of the cloaca or the urinogenital
sinus, derived at an early stage from the
primitive gut; and (3) external organs of
sex. Because of the great diversity of the
so-called secondary sex characters in verte-
brates, and because as a rule they become
sexually differentiated only in jiostpubertal
life, these structures can be considered only
in special cases.

Two distinct stages can be recognized in
the development of the accessory sex struc-
tures: an early phase which is independent
of sex, and wliich follows a \-ii-tuallv identi-

HORMONES IN DIFFERENTIATION OF SEX

111

cal course in all individuals, and a later
phase, the period of sexual differentiation,
which is chiefly hormone conditioned. Dur-
ing the first phase the primordial structures
necessary for the development of both sexes
are laid down and develop in similar or
identical fashion up to a certain point, at
which stage each embryo possesses, morpho-
logically and for a certain time, the capacity
to develop into an individual of either sex
(Fig. 2.22A). In this early, indifferent phase
of development the sex primordia show lit-
tle reactivity to hormones, capacity for re-
sponse evidently requiring a certain degree
of maturation in the reacting organs or tis-
sues. The onset of sexual differentiation of
the accessory structures follows the appear-
ance of sexual differentiation in the gonads,

and this is the phase of development in
which control by hormones is predicated.

A. DIFFERENTIATION OF THE
EMBRYONIC GONADUCTS

A complete account of the origin and nor-
mal development of the embryonic sex ducts
has been given by Willier (1939) and Burns
(1955b) . In the embryos of most vertebrates
both sex ducts are present and equally de-
veloped throughout the sexually undifferen-
tiated period. In many amphibians this
primitive condition is retained throughout
larval life or indefinitely ; the male, or Wolf-
fian ducts, function as nephric ducts in both
sexes and in females they are permanently
retained in this capacity. The Milllerian
ducts persist throughout life in the males of

VAS DEFERENS

MES0NEPHR05\5

EPIDIDYMIS

VAGINAL CANAL

BULBAR GLAND

1 H PHALLUS

V i URINOGENITAL

^'i-*^ SINUS

|1 > BULBAR GLAND

POUCH YOUNG - ± 10 DAYS

FEMALE a MALE-±35DAYS

Fig. 2.22. Early (io\ clopnicnt and sexual differentiation in the genital tracts of young
opossums. A. The lusrxu.il stage of development in a female embryo ±10 days of age, show-
ing the paired gonaducts of both sexes, the sexually indifferent stage of the urinogenital
sinus, and the undifferentiated genital tubercle or phallus. B. Male and female at about 35
days, when sexual differentiation is far advanced, showing the structures which develop from
the primitive sex ducts, and dimorphic development of the sinus region. The phallus shows
chiefly a difference in size, without marked morphologic divergence. (From R. K. Burns,
Survey Biol. Progr., 1, 233-266, 1949.)

112

BIOLOGIC BASIS OF SEX

many species as complete if somewhat rudi-
mentary canals. In amniote embryos, on the
other hand, the ducts of the genetically re-
cessive sex are typically transient struc-
tures. Sexual differentiation consists in the
retention of one sex duct with development
of its derivative structures, whereas the
other either disappears completely or sur-
vives only in a more or less vestigial condi-
tion (Fig. 2.22B). Under these circum-
stances experimental reversal of sex, to be
successful, must be properly timed with re-
spect to the state of development of the
heterotypic duct. Once regression has been
determined it is impossible to preserve the
duct.

1. The Miillerian Ducts: the Effects of Fe-
male Hormones

The effects of female hormones, whether
produced by grafted ovarian tissue or ad-
ministered in pure form, may be stated gen-
erally as follows: in female subjects as a
rule they accelerate sexual differentiation,
inducing a precocious hypertrophy of the
Miillerian ducts which with large doses may
become extreme. In males female hormones
cause persistence of the Miillerian ducts fol-
lowed by differentiation in varying degrees
depending on timing of treatment, dosage,
and the special status of the duct in the
species under consideration. There are many
deviations from this pattern, however, aris-
ing in part from basic group or species dif-
ferences and in part from the many experi-
mental variables."

In most amphibians the IVIiillerian ducts
of both sexes respond readily to sex hor-
mones during larval life. In Triturus {Tri-
ton) , after castration, both sex ducts remain
indefinitely in a more or less undifferenti-
ated condition, thus providing an ideal basis
for sex reversal. Grafting gonads into cas-
trates of either sex readily induces differen-
tiation of the appropriate duct (de Beau-
mont, 1933). Furthermore, in the males of
various species which have undergone com-
plete sex reversal the later development of

" For reviews and references to a large literature
covering amphibians, birds, and mammals see
Humphrey, 1942; Wolff, 1938; Willier, 1939; Rav-
naud, 1942; Greene, 1942; Moore, 1947; Jost, 1947a,
1948a, 1955; Ponse, 1949; Burns, 1949, 1955b; Stoll,
1950.

the ]\liillerian ducts is always in accordance
with the altered sex of the gonad, and the
ducts may eventually become completely
functional (see e.g., Humphrey, 1942; Ponse,
1949; Gallien, 1955). The reaction of the
Miillerian ducts to female hormones (estra-
diol, estrone) varies in different species and
is greatly influenced by the stage at which
treatment occurs and by dosage as well. In
Ambystoma the ducts show a marked hy-
pertrophy in females and the response in
males is almost as great. The backward
growth of the incomplete ducts is also ac-
celerated (Burns, 1938a; Foote, 1941).
Large doses, paradoxically, may arrest the
backward extension of the duct (as do male
hormones, q.v.) but the part already laid
down becomes greatly hypertrophied (for a
summary see Gallien, 1955).

The effects of the female hormone in bird
embryos are particularly striking (Wolff,
1938, 1950; AVillier, 1939; Gaarenstroom,
1939; Stoll, 1948). In male embryos both
oviducts persist and hypertrophy as does
also the right duct of the female, which nor-
mally undergoes involution (Figs. 2.23 and
2.12). Once established these effects are per-
manent, development continuing even after
hatching (Wolff, 1938). However, the period
of susceptibility to the hormone is limited.
Retention and permanent development can
be assured only by treatment up to the
seventh day of incubation (this is the so-
called "stabilization effect" of Wolff) ; later
treatment is without effect for the preserva-
tion of the ducts, irreversible changes having
occurred which determine their regression
with finality (Wolff, 1953b). The hormone
of the embryonic ovary has the same effects.
Ovaries grafted into the body cavity of
male embryos cause persistence and devel-
opment of\he oviducts (Wolff, 1946). The
effect of the hormone appears to be a direct
one since it occurs independently of any ef-
fect on the gonads.

In mammalian embryos the effects are in
general similar, but marked species differ-
ences have been found. Female hormones
cause accelerated development of the Miil-
lerian duct derivatives in females, and with
high dosages oviducts, uteri, and vaginal
canals all show great hypertrophy. In male
embryos the ducts are frequently retained
and also differentiate regionally into ovi-

HORMONES IN DIFFERENTIATION OF SEX

113

A B

Fig. 2.23. The effects of sex hormones on development of the sex ducts in chick embryos.
A. Normal female embryo of 18 days, showing development of the left oviduct with shell
gland, and retrogression of the right. B. Genetic male embryo, 18 days, treated with 2.0 mg.
estrone. Both oviducts are present and greatly hypertrophied. Compare with C for the nor-
mal condition. C. Normal male embryo at 17 days of incubation. Note the paired Wolffian
ducts and absence of both oviducts. D. Genetic female embryo at 17 days, after treatment
with 1.0 mg. androsterone, showing absence of both oviducts and extreme hypertrophy of the
Wolffian ducts. For the normal female anatomy compare with A, and for normal size of
male ducts see C. (From B. H. Willier, in Sex and Internal Secretions, 2nd ed., The Williams
& Wilkins Co., 1939.)

duct, uterine tube, and vaginal canal
(Greene, 1942; Burns, 1939b, 1942a and b;
Moore, 1941; Raynaud, 1942; Jost, 1947a).
Details of structure depend on the state of
development of the rudimentary Miillerian
ducts in the males of the species in question.
The vagina may be defective or absent en-
tirely in males of certain species, as in the
opossum, in which the duct is usually incom-
plete, without a connection to the urino-
genital sinus. In other species the effects are
slight or lacking entirely (Raynaud, 1950,
the field mouse; White, 1949, the hamster;
Davis and Potter, 1948, man).

The effects of male hormones. The effects
of male hormones on the ]\liillerian ducts
are more variable; strong inhibitory effects
are obtained in many species and under
proper experimental conditions; but with
large doses stimulating or "paradoxical ef-
fects" often appear, such as have been de-
scribed in the case of the gonads. The
time of administration of the hormone is
important. Both in chick embryos and in
larval amphibians treatment with androgens
before the appearance of the duct, or during

the formative period, may result in total
suppression.^^ In amphibians, in which the
]\Iiillerian duct develops slowly, a particu-
larly interesting situation is found. Early
administration of testosterone propionate
prevents development entirely (Burns,
1939c; Foote, 1941, Amhystoma) or may
leave only the ostial rudiment (Gallien,
1955, Pleurodeles) ; however, treatment dur-
ing the period of formation may result in
suppression of the unformed portion of the
duct, while the part already laid down per-
sists and with large doses may even be
strongly hypertrophied (Mintz, 1947). Here
is a striking paradox in which different re-
gions of the same structure (which are, how-
ever, developmentally of different age)

" See, for example, for amphibians, Burns, 1939c ;
Foote, 1941; Hanaoka, 1941b: for the chick, Wil-
lier, 1939; Wolff, 1938, 1950; Gaarenstroom, 1939;
Stoll, 1948; Huijbers, 1951. Exceptions must be
noted, however, in a few cases: the field mouse
(Raynaud, 1950); the hamster (Bruner and Wits-
chi,'l946, White, 1949); man (Davis and Potter,
1948) in which no clear effects were observed,
whether because of true species differences or other
experimental variables is not clear.

114

BIOLOGIC BASIS OF SEX

react in an opposite way to the same treat-
ment.

In chick embryos also, early treatment
with male hormone may completely sup-
press development of the Miillerian ducts
(Figs. 2.23 and 2.12; see also Gaarenstroom,
1939; Stoll, 1948, 1950) and a similar effect
is produced by grafts of the embryonic
testis (Wolff, 1946; Huijbers, 1951). Again,
however, suppression of the ducts depends
on certain rather precise conditions, the
dose must be adequate and the hormone
must act at the proper stage of development.
They can be suppressed completely before
the 6th or 7th day of development (Stoll,
1948; Huijbers, 1951) but later treatment is
ineffective. Thus (as was found also for the
''stabilizing effect" of female hormone on
the Miillerian ducts of male embryos) there
is a limited period of development during
which the ducts are susceptible to inhibition
by male hormones. In contrast with these
clear-cut results, however, a paradoxical
hypertrophic effect of certain male hormones
(androsterone, dehydro-androsterone and
related compounds) has been reported on
the ^Miillerian ducts of chick embryos after
rather large doses (Willier, 1939; Wolff.
1938; Wolff, Strudel and Wolff, 1948).

In the embryos of mammals effective in-
hibition of the Miillerian ducts by male hor-
mones has not been found, but suppression
of regional parts of the duct sometimes oc-
curs. The ostial portion is suppressed in the
hedge-hog (Mombaerts, 1944) and the vagi-
nal segment (the last part to be laid down)
is frequently inhibited in female opossums
(Burns, 1942a, b) and in mice (Raynaud,
1942). In the mouse and in the rabbit fail-
ure of the posterior ends of the ducts to
unite to form vagina and corpus uteri has
been reported (Raynaud, 1942; .lost, 1947a).
In female opossum embryos treated with
testosterone propionate the vaginal canals
are absent in about half of all cases and
arc always absent in males (in which, as
noted earlier, the terminal portion of the
Miillerian duct is lacking) . However, a para-
doxical stimulation of the Miillerian duct
and its derivatives also takes place in opos-
sums of both sexes when large doses (25 to
100 ^g. per day) are employed (Fig. 2.24;
see also Moore, 1941, 1947; Burns, 1939a.
1955b) an effect which completely disa]^)-

pears when the dose is lowered to ±5 /xg. or
less (Burns, 1942a, b).

In contrast with the failure of androgenic
hormones to inhibit effectively the Miillerian
ducts of mammalian embryos it is known
that they are normally inhibited by the
hormone of the fetal testis. In castrated
male fetuses of the rabbit the ducts persist
instead of regressing and develop almost as
well as in normal females; conversely, the
embryonic testis when grafted into a female
fetus inhibits the Miillerian duct in the
vicinity of the graft (Jost, 1953, 1955). On
the other hand, a crystal of testosterone pro-
pionate implanted in the same manner lacks
this inhibiting power. Testosterone also fails
to inhibit development of the Miillerian
ducts in castrate males although in all other
respects it fully compensates for the ab-
sence of the testis (Jost, 1947b, 1953, 1955).
This discrepancy has led to the suggestion
(Jost) that in mammals another substance
may be required for the inhibition of the
Miillerian ducts. In the fetal rat, on the
other hand (Price, 1956), neither testoster-
one nor the presence of the fetal testis in-
fluences the differentiation of the ]\Iiillerian
ducts in genital tracts when isolated at an
age of 17.5 days and cultured in vitro. In
this case, however (as suggested by Price),
it is likely that development of the Miille-
rian ducts has already been irreversibly de-
termined before the time of explantation.
This question will come up again in a dis-
cussion of the stage at which irreversible
determination occurs in the rat.

The effects of castration on development
of the Miillerian ducts. Although the effects
of steroid hormones on the development of
the Miillerian ducts are on the whole con-
sistent with theory (failure of the ducts of
mammalian embryos to be inhibited by male
hormone is a notable exception) such results
do not constitute evidence that sex hor-
mones are present and active in the normal
differentiation of sex. A direct test of this
question is provided by castration of the
eni])ryo. The effects of castration in am-
phibian larvae have been previously men-
tioned (p. 112); after removal of the gon-
ads both sex ducts fail to differentiate
fui'tluM-, ]X'rsisting indefinitely in the condi-
tion in which they were at operation. In re-
cent years this difficult operation lias been

HORMONES IN DIFFERENTIATION OF SEX

115

B

D

Fig. 2.24. Diagrammatic representation of the effects of relatively large doses of testos-
terone propionate, administered from birth to an age of 50 days, on the development of the
sex ducts in young opossums. A. Sex ducts as they appear in a normal male at 50 days; the
vas deferens (W