SA7i

QH431

r-
C .

NORTH

This book is due on the date indicated
below and is subject to an overdue
fine as posted at the circulation desk.

EXCEPTION: Date due will be
earlier if this item is RECALLED.

150M/01-92— 941680

THE GENETIC AND THE OPERATIVE EVIDENCE

RELATING TO SECONDARY SEXUAL

CHARACTERS

By T. H. Morgan

\

Published by the Carnegie Institution of Washington

Washington, 1919

CARNEGIE INSTITUTION OF WASHINGTON

Publication No. 285

PRESS OF GIBSON BROTHERS, INC.
WASHINGTON, D. C.

CONTENTS.

Part I. page.

Introduction 5

Castration of Sebrights 6

A male Sebright that did not become cock-feathered after castration 10

Transitional feathers 10

Castration of Fi hen-feathered males from Sebright by game 11

Castration of F2 hen-feathered males 13

Hewitt's Sebright hen that became cock-feathered in old age 14

Heredity of hen-feathering 14

Heredity of color in the cross between Sebright and Black-Breasted Game bantam. 17

A. The Fi birds 18

B. Description of F2 birds 19

C. Back-cross of Fi to game 21

D. The number of color factors involved 22

E. Back-cross of Fi 9 to Sebright c? 23

F. Review of the heredity of the color of the plumage in poultry 23

Endocrine cells in ovary and testes of birds 32

Luteal-cells in the testes of the male Sebright 34

Endocrine cells in the testes of mammals 35

Cyclical changes in the interstitial cells in hibernating mammals 36

Hermaphroditism in poultry and the secondary sexual characters 37

Part II.

Darwin's theory of sexual selection 43

Other theories to account for secondary sexual characters 45

Display of the male 50

Part III.

The genetic and the operative evidence 62

A. Evidence from mammals

B. Evidence from birds

C. Evidence from amphibia sr>

D. Evidence from crustaceans

E. Evidence from insects 92

Part IV.

Summary and conclusions

Bibliography

Description of plates 106

3

{ in

THE GENETIC AND THE OPERATIVE EVIDENCE RELATING
TO SECONDARY SEXUAL CHARACTERS.

By T. H. Morgan.

PART I.

There are a few races of poultry that have two kinds of males, one
with the feathering of the ordinary cock, the other with the feathering
of the hen. The Hamburgs and the Campines are perhaps the best
known races of this sort. Convention amongst breeders, in certain
countries, has determined that the cock-feathered bird shall be the
standard, and at other times and places that the hen-feathered males
shall be the show birds. In one breed, at least, viz, the Sebright
bantams, the hen-feathered cock is the only known type. Cock-
feathered Sebrights have never been seen, so far as I know. This
breed is pure for hen-feathering. As shown in plate 1, figure 3, the
male Sebright lacks the long, pointed saddle feathers at the base of
the tail of the common cock, also the peculiar back and neck feathers
(hackles) of the cock bird, as well as the male feathering on the bow
of the wing. His feathers in these parts are almost exactly like those
of the hen (plate 4, fig. 4). The long sickle feathers covering the true
tail are also absent, although the two median ones sometimes occur
in males of this race.

The Sebrights seemed excellent material for studying the heredity
of this type of plumage in the male. In 1911 I began to study this
problem, and crossed Sebrights to Black-Breasted Game bantams.
The latter race was chosen not only because the males have the typical
cock-feathering, but also because the coloration of these birds resembles
very closely that of the jungle-fowl, from which many, perhaps all, of
our domesticated races have sprung.

In dissecting some of the F2 birds from this cross I noticed that the
testis of the male was often more flattened than is the testis of the
typical male bird, that it was often somewhat pear-shaped, and that
frequently it was in part or entirely black. Recalling that male
Sebrights are said to be often partially sterile, the idea naturally
suggested itself that these birds are hen-feathered because the testes
have assumed some of the characteristics of the ovary. It had long
been supposed, and had been finally established by Goodale, that the
presence of the ovary in the female suppresses her potential develop-
ment of plumage, for when the ovaries of the hen are diseased or
removed she develops the plumage of the male. This reasoning led
me to try the experiment of castrating the hen-feathered males in

5

D. H. HILL LIBRARY
Ncrih Carolina State College

6 THE GENETIC AND THE OPERATIVE EVIDENCE

order to see if they would become cock-feathered. The outcome was
immediately apparent; the new feathers were those of the cock bird.
While the " reasoning" that led to the experiment is open to serious
question, nevertheless the "hint" furnished by the unusual condition
of the testis led finally to the discovery that luteal cells were present
in abundance in the testes of the male Sebright like those present only
in the females of other breeds. Whether or not the shape of the testis
of the Sebright, that is sometimes like that of the ovary, is connected
with the unusual abundance of luteal cells in the testis I do not know.
If so, then the hint that came from their shape was not so unreasonable
as appears at first sight.

The birds first operated upon were adult Fi and F2 hen-feathered
birds. The first one done by myself died, but a few, whose testes were
removed by Dr. H. D. Goodale at my request, lived and changed to
cock-feathered birds. Since then I have operated successfully on a
number of Fx and F2 birds, as well as Sebright males. In these opera-
tions I have had throughout the assistance of Dr. A. H. Sturtevant
and for two years the assistance of Dr. J. W. Gowen also. I wish to
express my appreciation of their help and advice, for without it I
doubt whether I could have carried out the work successfully. Since
the main interest attaches to the Sebright experiments, they will be
described first, although they were the last to be performed.

CASTRATION OF SEBRIGHTS.

Except for the similarities of the plumage, the male Sebright differs
as much from the female as do cocks of other races. The rose comb
is very large in the male, small in the female (plate 4, figs. 3, 4). The
wattles also are longer in the male. The cock carries himself erect,
as do the males of other breeds. His spurs are well developed and
he shows the aggressive behavior of his sex. On the other hand, the
shortness of the feathers on the back of the neck (the hackles), the
absence of the pointed feathers on the back and rump, and the usual
absence of long sickles and other tail-covert feathers make him
hen-like. The detailed account of the feathers in these critical regions
will be given when comparisons are made with the feathers of the
castrated birds (plates 6 and 8).

Six males have been successfully operated upon and with one
apparent exception have all given the same results. The birds were of
somewhat different ages ; they had been hatched about July, and were
operated upon about November of the same year, when they were
either half grown or had nearly reached maturity. At the time of the
operation a few feathers were removed from different regions of the
body, and the new feathers that regenerated in the course of 3 or 4
weeks showed all the characteristics of those that came in later to
replace the juvenile or first adult coat. These regenerated feathers do

RELATING TO SECONDARY SEXUAL CHARACTERS. 7

not, therefore, call for special notice. All of the new feathers were in
shape, pattern, and general coloration strikingly different from the
original feathers, some of which were at first still present, the old
feathers of course showing no change.

After completely molting, the appearance of the birds may be
gathered from the photographs (plate 5) and from the colored drawings
(plates 1 and 3). The male now has in all points the plumage of a
typical cock-feathered male bird of other breeds. This is startlingly
apparent in the hackle, back, rump, sickle, and tail-covert feathers.
Instead of the laced feathers that are characteristic of both male and
female, the whole upper surface of the bird appears reddish or yellowish,
the black marginal edging of the feathers having disappeared. A
detailed comparison of the feathers of the different regions will show
how great a change has taken place. (See page 8.)

In plates 6 and 8 the feathers from characteristic regions of the
normal Sebright and of the castrated Sebright are shown in pairs.

One of the first Sebrights that was castrated was a lighter bird than
the others. Its lighter color was partly due to the narrower outer
band of the laced feathers, (plate 6, figure 1,) and partly to the
lighter color of the yellow-brown center of the feathers. The bird had
a single comb, but as this crops up occasionally in some stocks of
Sebrights, it need not be interpreted to mean that the bird was impure
for color factors. After being castrated the bird changed over com-
pletely to cock-feathering and has remained in that condition for two
or more years. As shown in plate 5, figure 2, the plumage is
even more fully developed than in cock birds of some other breeds.
The comb and wattles are, however, shrunken and pale, as in a capon.
The bird is timid and scarcely or never crows. When killed (May
1919) no pieces of testes and no trace of testicular tissue at the old
situs were found.

The details of the feathers are shown in plate 6, figures 1 and la,
where, in each instance, one of the old and one of the new feathers
from the same region are placed side by side. The feathers on the
head and hackle are yellow, even to the base. At the base of the
hackle — the so-called cape — a few feathers have a small black tip.
The feathers of the back are entirely yellow, except that where the
fluff begins there is some dark pigment. The saddle feathers are for
the most part all yellow, but a few have at the base, near the fluff,
black on each side. The tail coverts are long, with a black margin at
their tip. The tail feathers are long, mossy, and have a black tip.
The wing-bow feathers are all yellow, except the black fluff at the base.
The feathers on the crop are mostly yellow with black margin around
the end. Those on the breast lower down are yellow with black tip
and black fluff.

There was another Sebright operated upon at the same time that was
a darker bird (as the original feathers show, plate 8, figs. 1 to 4). It had

8

THE GENETIC AND THE OPERATIVE EVIDENCE

a rose comb. The feathers that were plucked at the time of operation
were replaced at once by new feathers of the cock-feathered type. The
new feathers that came in as the old ones were molted were also cock-
type, and the bird soon assumed the complete characteristic cock-
feathering. The comb was shrunken as in castrated birds (plate 5,

SEBRIGHT.
Plates 6 and 8.

1. On the head (a) the feathers are small,

dull black with lighter margin and
reddish quill.

2. On the hackle (b) the feathers are yellow

bordered with black, expecially at
the base, and at the tip outside of
this border there is a narrow yellow
border (broader at base). The
border is absent at tip.

3. In the middle of the back (between the

wings) the feathers are yellow with
black margin at the tip. At the
base there is some dark color.

4. The saddle is made up of typical laced

feathers with black where the
fluff begins.

5. The tail coverts are short; the upper

ones, especially the short sickles,
are slightly curved. The sickle
feathers extend up only about half
the length of the tail. They are
yellow, laced, and have a black
margin, tending to be lost at tip.

6. The yellow tail is short and erect.

7. The feathers on the wing bow are prac-

tically like those on the back, but
shorter.

8. Over the crop and lower breast the

feathers are laced.

CASTRATE.
Plates 6 and 8.

1 . Feathers entirely yellow and more slender.

Those on each side of the shrunken
comb stand up from the head.

2. Hackles on upper part of neck have a

black base with red tip. The
outer edge, without barbules, is
narrow, then broader than at tip.
Farther down the neck the edge
with barbules is yellow with a
narrow black margin.

3. In the part of the feather with barbules

there is a yellow center bordered
by a broad black band, especially
at base. In the part without bar-
bules the feathers are yellow and
more pointed. This region espe-
cially is deeper yellow than in the
original Sebright.

4. Saddle consists of long, slender laced

feathers, except at tip, which is red.
Barbules are absent along edge of
outer third of feather.

5. Tail coverts long, covering the tail as in

cock birds of other breeds. The
sickle feathers, especially the upper
ones, are much curved, with black
tips; the black margin is largely
gone. The feathers are mossy,
sometimes splotched (this is also
sometimes noticeable in normal
birds) .

6. The tail feathers themselves are almost

twice as long as in normal bird; the
upper feathers are more curved.

7. The feathers of wing bow are like those of

the back of the same bird, but
shorter.

8. The crop feathers are narrower, with a

wider black margin, and a few may
be also entirely black. Feathers
on lower breast much like those in
normal, but a little more pointed.

figure 5). During the spring of 1917 it was noticed that the bird was
going back towards the hen-feathered type, and by the end of the summer
he was in the intermediate condition, as shown by the photograph and
by the individual feathers (plate 8, lb, 2b, 3b, 46) . The comb had begun
to enlarge also. The bird was opened again (1918) and pieces of testis
about as big as peas were found on one side. Evidently a piece of the old

RELATING TO SECONDARY SEXUAL CHARACTERS. 9

testis had been left behind and had regenerated. As it enlarged the new
feathers were affected so that the plumage returned towards the normal
type. The pieces of testis were removed and a few feathers plucked
out. The new feathers that came in were typically cock-feathered,
and, as the molting proceeded during the winter and spring, the bird
became cock-feathered for a second time as shown in photograph
(plate 5, figure 6) and by the feathers in plate 8, lc, 2c, 3c, 4c. Here,
then, is an excellent example of the connection between the gonad and
the condition of the plumage. On opening this bird (May 1919) no
pieces of testes were found. There was a very small whitish lump at
the situs of the old testes, which, when sectioned, showed some gland-
ular-like tissue, not in tubules, and no evidence of testicular tissue.

Three other younger Sebrights of the same stock were successfully
castrated. They were hatched in June or July and castrated in
November of the same year. They remained quite small birds, despite
their elongation due to the long tail and tail coverts that they devel-
oped. One of these birds in his cock-feathered plumage is shown in
plate 3, figure 1." One has died, the other two are alive and markedly
cock-feathered, as shown in plate 6, figure 2a. All three birds were
dark red-brown, much more so than the two preceding cases, especially
the first case. This color difference might be attributed to the earlier
age of the three birds when operated upon, or to a more complete
(or less complete) operation involving perhaps neighboring parts, or
to the birds having a somewhat different genetic composition (t. e.,
modifying factors). - There is no special reason why the operation if
performed early should have a different result on feathers that develop
after the bird is of adult age. Goodale has suggested that there may
be organs in the vicinity of the testis that have some influence on the
kind of plumage produced, and if there are such organs they might be
removed in one bird and accidentally left in another. It would not,
however, be probable that the bird operated on at first had received
one" treatment and the later ones the other treatment. It seems to me
more probable that the birds have come from different genetic strains,
and that this genetic difference gives a more plausible explanation
of the darker cock-feathered plumage. Goodale observed, for the first
time I think, that the largest wing coverts of the castrated cock be-
come longer. I looked, therefore, with some interest at the condition
of these same feathers in the castrated Sebrights. As shown in plate
10, figures 1, la, these feathers are also longer and narrower in the cas-
trated Sebright than in the normal bird.

The true tail feathers of the capon are said to be longer than those of
the cock. This holds also for the tail feathers of all of my castrated
Sebrights. Their true tail feathers are considerably longer than those
of the normal male, as seen by pulling them out and comparing the
two. Their length is concealed while on the bird by the excessively
long coverts that appear after castration.

10 THE GENETIC AND THE OPERATIVE EVIDENCE

In 1916 I operated on a Sebright male that lived for some months,
but died in the summer of 1917. At the time of his death he had
assumed a partial cock-feathering, as shown by the feathers in plate 9,
figure 3, 3a. Dissection showed that some of the testes had been left,
and as is then to be expected, the change was incomplete.

A MALE SEBRIGHT THAT DID NOT BECOME COCK-FEATHERED

AFTER CASTRATION.

One of the males that had been castrated with the others did not
become cock-feathered even after a year. Taking for granted that
the castration had been incomplete, the bird was opened, but as no
pieces of the testes were to be found in the normal position he was
killed and carefully dissected. There were no pieces of testes found
in the normal situs. A small whitish patch of material from this region
was cut into sections, but no testicular material was found in it. Then
a large piece of the back from the region of the attachment of the
testes was prepared, but as yet this piece has not been sectioned. Even
were a small piece of tissue to be found, it would seem unlikely that it
would suffice to hold back all indications of the cock-feathering, for
after incomplete removal of the testis there are nearly always at first
some indications of the lack of material. The most plausible view here
is either that some other gland may have assumed, provisionally, the
function of the missing testes, or else a detached piece has not yet
been found. Glandular cells like the luteal cells of the ovary have
in fact been described by some observers in other organs. of the body.
As yet I have not found time to make a thorough histological study
of the tissues of this bird.

TRANSITIONAL FEATHERS.

In several birds new feathers had begun to develop at the time of
the operation under the influence of the testicular secretion. After
the removal of the testes, these feathers continued to grow and in the
absence of the original conditions changed over to the other type.
The outer end of these feathers shows the original or normal shape and
color, while the inner end shows the new characteristics. Such feathers
have been seen in nearly all of my castrated birds; a few from the
Sebright will suffice by way of illustration. In plate 10, figure 2a,
four such transitional feathers are shown. In a and b two feathers
from the hackle are photographed. The first (a) had begun as a normal
Sebright hackle feather, as seen in the condition of its tip; the rest of
the feather is the same as the feather of the castrated bird. For com-
parison with this feather, two (2b) from the same bird are shown that
began to develop after the testes were removed, i. e., at the same time
as the change occurred in the former feather. At the time the latter

RELATING TO SECONDARY SEXUAL CHARACTERS. 11

feather (b) had not yet completed its full growth. On the bow of the
wing a few intermediate feathers, like the one shown in 2a, were
present. (For comparison with normal and castrated feathers see
those on plates 6 and 8.)

An intermediate feather from the back is shown in 2a. For com-
parison with the old feathers from the same region see plate 6, fig. 2.
An intermediate saddle feather is shown in 2a. For comparison with
normal feathers from the same region see figure 2. A still later feather
from the castrated bird is shown in 2b. The last was not yet com-
plete when removed from the bird.

It will be noticed that the change after castration involves the color,
the shape, and the presence and absence of barbules in those parts of
the bird that are peculiar in the last respect. The transition in these
characters is quite sharp — as sharp in 'fact as is compatible with the
passage from one structure to that of an entirely different kind without
any discontinuity of growth. Owing to the quickness of the response
shown by the feather, it will be possible to study more in detail the
length of time the secretion remains in the body of the bird after the
testes have been removed.

CASTRATION OF F, HEN-FEATHERED MALES FROM SEBRIGHT

BY GAME.

Hen-feathering is dominant to cock-feathering. As shown in plate
2, the Fx male is almost as completely hen-feathered as is the male
Sebright. There is a somewhat greater color difference between the
Fi male and Fi female than between the Sebright male and female.
Two Fi birds were castrated for me by Goodale. At the time of
operation, in the autumn of 1916, both birds were full grown, (plate
2, figure 1). After molting the old feathers, both birds appeared as
shown in plate 2, figure 4. Each is completely cock-feathered. The
plumage has also undergone a remarkable change in color. In general,
the color change is from yellow and black to reddish yellow. The
greatest change is over the upper surface. The sickle, covert, and tail
feathers are well formed and have now become iridescent black. The
breast has changed least of all. One bird died February 12, 1919.
When opened there was found on the left side a small white lump ;
on the other side almost nothing. The lump was found to consist of
testicular tubules with loose glandular cells on its walls.

The extent to which the change has taken place is best shown by
comparison of individual feathers from identical regions — one before
and one after the new feathers (taken out two years later) have conic
in (plate 7). The contrast between the old and new feathers of the
hackle, saddle, back, and wing-bow are the most striking. In all of
these the new feathers have become red on the exposed portion and the
margin is free from barbules, as in the cock bird. The increase in size

12

THE GENETIC AND THE OPERATIVE EVIDENCE

and change in shape of these feathers is remarkable. Equally great
is the change in the tail-coverts that grade into those of the saddle
at the base. The two median coverts or sickle feathers are longer than
the tail and much arched. They are jet black with a purplish irides-
cence and with a yellow-red shaft. The tail itself has also changed;
it has lost its stippling, and has become black like the coverts. The
increase in length of the tail feathers is as remarkable as the increase
in length of the coverts. A detailed account of these changes follows :

1. The head feathers are yellow with black

base that shows through on the
head.

2. The hackle feathers are yellow with black

base (showing through on neck).

3. The red feathers of the back are penciled.

There is a black band, especially
around tip just inside of the
margin.

4. The saddle feathers are much like those

of the back, much stippled in cen-
ter. The border is more distinct.

5. The tail coverts are similar to those of

the hen.

6. The tail feathers are black, and with

exception of the lower feathers they
are partly stippled.

7. The wing-bow feathers are penciled like

those of the back.

8. Feathers on crop yellow with black spot

at tip; those lower down on breast
have a bigger spot.

CASTRATE.
1. The feathers are entirely red.

2. The hackle feathers are entirely red.

3. Feathers of back are red except for black

at base. Barbules absent at end
and side. Tip pointed.

4. The saddle feathers are red with black

base. They are long and pointed.

5. The coverts are blue-black, with brown

shaft. They are long, pointed,
and curved.

6. The tail feathers are black, not stippled,

and have a black shaft.

7. The wing-bow feathers are red with

black base.

8. Over the crop the feathers are orange-

brown; on the breast they have the
same color and a small black tip.

The Fi bird from which the colored drawing (plate 2) was made
and from which the normal Fi feathers were pulled was lent to Dr.
Goodale in the summer of 1917. The bird died in April 1918, and his
skin was sent to me. He also had begun to change over to cock-
feathering (plate 9, figures 2, 2a). Goodale recorded that the
testis had dwindled to small bodies only about 10 by 5 mm. This
accounts for the change to cock-feathering. For comparison I have
added a third set of feathers to the two former sets, showing the new
hackle, back, saddle, wing, and bow feathers of this bird. The feathers
show that the change is in the same direction as that shown by the
castrated cock, but it has not gone so far in the direction of cock-
feathering. The tail is still short and the feathers are black. The
sickle feathers are not longer than the tail and are stippled. It is
probable that this is the old tail whose feathers have not been molted
since the testis dwindled. In fact, elsewhere the old and the new
feathers are both present, showing that a complete molt had not taken
place. The old feathers still present are practically like those of the
original Fx bird, showing that the change was of recent date, and due
to the decrease in the testis which was probably caused by disease.

RELATING TO SECONDARY SEXUAL CHARACTERS. 13

CASTRATION OF F2 HEN-FEATHERED MALES.

The F2 hen-feathered males from this cross could not be utilized until
they had begun to assume the adult plumage, since before that time
they were like the cock-feathered F2 males. Consequently, the opera-
tion is more difficult and more dangerous to the bird. A good many
birds have died in consequence of the operation, but enough successful
operations (five) were made to show what the color of certain types of
hen-feathered bird would be when changed to cock-feathering.

A hen-feathered male (No. 292) that was darker than the F! male —
in fact, almost black, except for a yellow center in some of the dorsal
feathers that were mossy or penciled — was castrated. The details
of characteristic feathers may be gathered from the feathers in plate 7,
figure 2. A corresponding set of the new feathers after castration, 2a,
are paired with the former. The castrated male in his new plumage
is shown in plate 2, figure 3. His dorsal surface is colored very much
as is the same region in the Fx bird, but the breast is very much
darker, so that the bird as a whole presents a very different appearance
from the FL castrated male. A very small white mass was found
when the bird was killed in place of the old testis, composed, in sec-
tions, of a reticulated mass of cells that look like old broken-down
follicles of testicular tubules with a few cell-layers lining the tubules.

An F2 male (68) also had dark feathers (plate 3, figure 2, and plate
9, figure 1) . The castrated male in his new plumage is represented in
plate 3, figure 3. Here again the upper surface is much like that of the
last castrate, and also like that of the Fi castrate. The breast has
changed much less than the back ; the centers of the feather are brown
with a black margin and a black band at the tip. The exposed
portion of the secondaries and the coverts are not so brown as in the
last bird. The spurs of this bird were bent back, looking like the
horns of a ram. When killed and examined, several small white pieces,
that looked like pieces of testes, were found in the abdominal cavity
near the old attachment of the testis. A histological study showed that
these pieces contained tubular tissue apparently testicular, but with-
out germ-cells.

Another F2 male (Band No. 221) was yellow in general color, the
feathers being irregularly penciled. After castration (plate 3, figure 4)
the bird became red above and deep brown below; the tail and
coverts were black.

A pale-yellow hen-feathered bird (No. 218) was also castrated.
Here also the change was most conspicuous over the upper surface,
not only in a greater depth of color than elsewhere, but in the shape,
etc., of the feathers. On the breast the original yellow color remains,
but is slightly deepened. When killed and opened (May 14, 1919),
a few small, whitish pieces were found. When these were sectioned it
was seen that they were made up, for the most part, of tubules look-

14 THE GENETIC AND THE OPERATIVE EVIDENCE

ing like those of the epididymis and also a few testicular tubules.
At the old situs there were some regenerated lumps, which in sections
appeared to be loose glandular tissue. No germ-cells were present and
the tissue just referred to may be old testicular tubules.

HEWITT'S SEBRIGHT HEN THAT BECAME COCK-FEATHERED

IN OLD AGE.

Darwin records in Chapter XIII of Animals and Plants under
Domestication a change that took place in an old female Sebright :

"Mr. Hewitt possessed an excellent Sebright gold-lace bantam hen, which,
as she became old, grew diseased in her ovaria and assumed male characters.
In this breed the males resemble the females in all respects except in their
combs, wattles, spurs, and instincts; hence it might have been expected that
the diseased hen would have assumed only those masculine characters which
are proper to the breed, but she acquired, in addition, well-arched tail sickle-
feathers quite a foot in length, saddle-feathers on the loins, and hackles on the
neck — ornaments which, as Mr. Hewitt remarks, would be held to be abomi-
nable in this breed."

This is the only record I know of showing the change that takes
place in the Sebright hen when the influence of her ovary is removed.
There can be no doubt from the above description that she changes in
the same way as does the castrated Sebright male.

Concerning the origin of the Sebright bantam Darwin states that
the race "originated about the year 1800 from a cross between a com-
mon bantam and a Polish fowl, recrossed by a hen- tailed bantam, and
carefully selected; hence there can hardly be a doubt that the sickle
feathers and hackles which appeared in the old hen were derived from
the Polish fowl or common bantam; and we thus see that not only
certain masculine characters proper to the Sebright bantam, but other
masculine characters derived from the first progenitors of the breed,
removed by a period of about 60 years, were lying latent in this hen
bird ready to be evolved as soon as her ovaria became diseased."
To-day the problem appears to us in a somewhat different light, since
the secondary sexual characters referred to by Darwin have simply
been kept under for more than a hundred years by the secretion pro-
duced in the ovary of the hen (as in all breeds) and in the testis of the
male Sebright.

HEREDITY OF HEN-FEATHERING.

In 1913 I found that hen-feathering as seen in the Sebright is a
dominant non-sex-linked character. A preliminary statement was
given in the first edition of my book on Heredity and Sex (1913),
which treated the character as a recessive one. This was a mistake
due to a male having been obtained that was like the game race,
which subsequent work showed must have been due to a sperm hav-

RELATING TO SECONDARY SEXUAL CHARACTERS

15

ing been retained in the oviduct of the female during her isolation
period. In the second edition published a few months later the mis-
take, having been found out, was corrected.

If one dominant suffices to produce hen-feathering, the F2 ratio
would be 3 hen-feathered to 1 cock-feathered bird. The numbers
found were 31 to 28. This realized ratio departs too far from a 3: 1
ratio to make it probable that the results are due to a single factor.

The F2 expectation for two dominants, both necessarily present to
produce hen-feathering, is 9 hen-feathered to 7 cock-feathered birds.
If the dominant factors are represented by H and H' and their wild-
type (recessive) allelomorphs by h and h', the expected F2 recombina-
tions are given in the following table :

HH'

Hh'

hH'

hh

HH'. . . .
Hh'

hH'....
hh

/ HH'
\ HH'

Hh'
HH'

hH'
HH'

hh
HH'

/ HH'
\ Hh'

Hh'
Hh'

hH'
Hh'

hh
Hh'

/ HH'

\ hH'

Hh'
hH'

hH'
hH'

hh
hH'

/ HH'

\ hh

Hh'
hh

hH'
hh

hh
hh

There are 9 classes containing both H and H', 6 containing one or
the other, and one containing neither H nor H\ The realized numbers,
31 to 28, are in close approximation to 9: 7.

In classifying the F2 hen-feathered males, an attempt was made to
divide them into two classes, viz, type 1, hen-feathered to the same
extent as the Sebright, and type 2, intermediate between hen and cock
feathering. The line between intermediate and cock-feathering is
sharp, all the intermediates belonging distinctly to the hen-feat hered
group, but the line between the two subdivisions of hen-feathered
birds is not sharp, and occasionally a bird is found that is difficult to
place. These statements hold also for the Fi birds, whose skins I now
have. Five of these are classified as intermediates and one as com-
pletely hen-feathered. The difference between these two classes, then,
is environmental or due to other modifying genetic factors, for which
either the Sebright or the game is not pure. Under these circumstances
it would not be profitable to attempt to find out (without additional
evidence) what genetic differences, if any, lie behind the hen-feat li-
ered and intermediate-feathered birds in the F2 classes.

Concerning the back-cross (Fx by game) the expectation, for one
dominant factor-difference, is 1 hen-feathered to 1 cock-feathered
male. There were obtained 2 hen-feathered (intermediates) to 7 cock-

16

THE GENETIC AND THE OPERATIVE EVIDENCE

feathered birds. The numbers are too small to be significant, taken by
themselves. The expectation for 2 dominants, both essential to hen-
feathering, is 1 to 3, and this is in agreement with 2 to 7 as found.
It seems, then, more probable from the evidence of the F2 and of the
back-cross combined that there are two dominant factors present in
the Sebright that make the male hen-feathered, and since the race
breeds true to hen-feathering, both factors must be present in homo-
zygous condition unless an undetected lethal destroys some of the
classes.1

Gold Male

(Belgian imported
from Brussels.)

X Silver Female
(Product of English
type males for 7
generations in my
own yards )

I
6 Silver Males
(Belgian type)

6 Gold Females

B

Silver Cross-Bred X Gold Cross-Bred

Male
(offspring of A)

Female
(offspring of A)

4 Silver 2 Gold

Males Males

(Belgian (Belgian

type) type)

2 Silver
Females

5 Gold
Females

Silver Male

X Gold Female

6 Silver Males
(English type)

6 Silver Females

D

Silver Cross-Bred X Silver Cross-Bred

Male
(Offspring of C)

Female
(Offspring of C)

Silver Males Silver Females Gold Females
(English type)

E

Silver Male (offspring of C)

Gold Female (pure)

Silver Males Silver Males Gold Males Gold Males Silver Females Gold Females
(Eng. type) (Belgian type) (Eng. Type 1912) (Belgian type)

F

Silver Male (offspring of C)

X

Silver Female (pure)

Silver Males (English type)

Silver Females

Gold Females

Smith and Haig have reported the following curious case of hen-
feathering. Smith had a breed of White Leghorns with cocks of two
classes — those that assumed cock plumage at 6 months and those
that are like the hens for 8 months, after which they slowly assumed
the cock-feathering. The difference is hereditary and appears to
segregate. Possibly this breed had one factor at least for hen-feathering
that is effective for young birds, but not for older ones, or some of the

1 The expectation for 1 dominant and 1 recessive factor is so nearly the same as for 1 dominant
alone that for the numbers obtained no difference between the two cases could be detected.

RELATING TO SECONDARY SEXUAL CHARACTERS. 17

birds pass through a stage when they produce an internal secretion
that disappears later. But it is also possible, and perhaps more
probable, that the young birds, not cock-feathered, have remained
longer in the juvenile stage than the others, so that they might be said
to be falsely hen-feathered.

The results published by the Rev. E. Lewis Jones in 1914, describing
crosses between two breeds of Campines, one called Belgian (which
has hen-feathered males), the other English (that has cock-feathered
males), are summarized in the table on page 16. They show the domi-
nance of hen-feathering with some probability. The table given there
is the original, to which the author has kindly added the numbers here
prefixed to some of the classes. The numbers are not large enough in
all cases to be satisfactory, but the dominance of the hen-feathering
is, I think, apparent, as well as its non-sex-linked transmission. The
golden female in C must have been English type, or at any rate hetero-
zygous for English-type feathering, for if Belgian her sons would have
been Belgian type.

Punnett and Bailey (1914) have published the result of a cross with
hen-feathered Silver Sebrights and Hamburgs. The dominance of
hen-feathering in the male is shown in the figures that illustrate their
paper, but as the paper deals solely with the inheritance of weight the
account of inheritance of hen-feathering was deferred to a later paper,
that has not yet appeared.

HEREDITY OF COLOR IN THE CROSS BETWEEN SEBRIGHT AND
BLACK-BREASTED GAME BANTAM.

The cross between the Sebright and the Black-Breasted Game
bantam was undertaken primarily to study the inheritance of hen-
feathering. The Sebright was chosen, on the one hand, because this
race is pure for hen-feathering, whereas in other races, such as the
Campines, both kinds of males are known. The hen-feathered birds
of such races are, I believe, frequently not pure for hen-feathering. The
game race was chosen because the cock has the typical plumage of the
wild bird, Gallus bankiva, and although his feathers are remarkably
short, they show the characteristic cock-feathered type.

Only secondarily was the experiment concerned with color inheri-
tance. The two breeds differ so markedly in coloration and pattern
that the very complex results that appeared in F2 were to be expected.
In addition to the differences involving hen-feathering versus cock-
feathering, and Sebright plumage versus game plumage, the game is
strongly dimorphic in the plumage, while in the Sebright the colora-
tion of the two sexes is closely similar. But the castration experiments
have shown that this difference is the result of hen-feathering in the
Sebright cock, and that the race carries the same potential dimorphism
as do other races of poultry.

18 THE GENETIC AND THE OPERATIVE EVIDENCE

The game cock is shown in plate 1, figure 1, and plate 4, figure 1.
The wattles and comb had been removed from the bird. The yellow-
red back and saddle are to be noted. The upper tail coverts and sickle
feathers are black, as is the tail. These parts are shorter in the game
than in other races, being one of the points selected for. The dorso-
anterior edge of the wing is black, this color meeting across the middle
of the back. Below this black area comes the red wing bow, followed
by a double row of blue-black feathers. The exposed portions of the
secondaries are brown, of the primaries black with green margin. The
breast and entire lower surface is black. The legs are greenish, the
bill black and yellow, the iris yellow.

The hen of the Black-Breasted game (plate 1, figure 2) is light
yellowish-brown. The back, saddle, and wing coverts are golden
brown, finely penciled with darker brown or black. The hackle is
penciled; it has a yellow border (without barbules); the back is more
brown, the forepart of the breast is salmon, the more posterior parts
lighter salmon. The sides of the body under and below the wings are
stippled gray.

The Sebright male is represented in plate 1, figure 3. Photographs
of the male and the female are given in plate 4 figures 3 and 4. Most
of the feathers have a yellow center and a black border. Such feathers
are said to be laced. The details of the different regions are shown in
the feather plates, 6 and 8.

A. The Fi Birds.

The Fi birds were remarkably uniform. The sexual dimorphism
is slight, as a comparison of the male and female in plate 4, figures 5, 6,
will show. In the female the body feathers are penciled but very
mossy, and this holds for the male too, except that in the hackle, back,
and saddle, a change in color accompanies the change in shape, as seen
in the individual feathers in the feather chart (plate 7, figure 1). If
there are any sex-linked factors involved in the cross, we should expect
different types of Fi hens in the direct cross and its reciprocal, because
in one case the Fi hen gets her single X chromosome from one father,
and in the other case, the reciprocal cross, from the other. Unfortu-
nately no careful comparison can now be made, because the crosses were
carried out in different years and the changes due to age may have
affected the color sufficiently to obscure such slight difference that may
have existed. But the effects of such factors, if present, are very small,
since the birds seemed to be the same, regardless of the way in which
the cross was made. In the F2 counts, although an attempt has been
made to keep apart the birds obtained in the two crosses {%. e., the di-
rect and the reciprocal crosses) , it is very doubtful if the two groups
show any significant differences.

RELATING TO SECONDARY SEXUAL CHARACTERS.

19

B. Description of F2 Birds.

All together there are 72 hens, 29 hen-feathered males, and 26 cock-
feathered males, as shown in table 1 :

Table 1.

F2 from Sebright 9 by Game cf.

Fj from Sebright cf by Game 9 .

Females.

Hen-feath.
d* type 1.

Hen-feath.
d* type 2.

Cock-feath.

Females.

Hen-feath.
<? type 1.

Hen-feath.
d" type 2.

Cock-feath.

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

15
1
3
3
6
6

11
5
1
2

3

2

1
5

4
2

1

4

1
1
2

1

1

2

1

1?

1

1

1
2
1

1
1

2
1

3
1

2

1

2
2
1

1
2

2
2
2

1

1

1

2

1
1

1

1

53

7

8

13

19

7

7

13

F2 Hens.

A. 8 females that are like Fi; 7 others resemble them below, but have stippled back and

rump feathers (4 of these have yellow necks, and 3 black necks like those of Fi).
Here, then, are two or three subdivisions, or perhaps main classes.

B. 1 female is very close to game, having the characteristic stippling above and salmon

breast below. She is darker colored than the game females, therefore more like
the Leghorn female.

C. 3 birds resemble the Sebright in plumage, but would not pass muster for real Sebrights.

D. 3 others have spangled breast feathers like the Sebright, but a great amount of stippling

on the back.

E. 6 birds are yellow on breast, with stippled back.

F. 6 others are yellow, stippled birds with a little yellow penciling on the breast.

G. 11 black birds with some stippling on the back of the wing, and sometimes with

traces of yellow in the hackle.

H. 6 other birds are dark, but not as black as the last. There is some stippling, especiallv
on breast. The hackle is always striped. (4 of these have yellow necks above
and below; 2 have dark necks.) The series of feathers photographed in plate 10,
figure 4, from a bird of class H, show all gradations between a spangled and a
barred condition. It is practically certain that the barring seen here (as well as
that under class M) is quite different from that of the Barred Plymouth Rock.

7. 1 yellow hen with a little black (as a band) on wings and tail.

J. 2 other yellow birds with a little black penciling on the back, the tail, and with long
wing feathers.

Fj Males.

K. 2 intermediate males with a black-splotched red breast and black tail. The saddle is
coarsely stippled. The corresponding male-feathered bird is red above.

L. A Sebright-like bird with black stippled feathers on back; the rest of the plumage
heavily laced. Posterior part of breast and thigh black. The tail stippled.

20 THE GENETIC AND THE OPERATIVE EVIDENCE

M. Yellow neck, the back, tail, and lower half of each secondary coarsely stippled. Pos-
terior part of breast barred; cape and anterior breast penciled. The series of
feathers photographed in plate 10, figure 5, from the breast of a bird in class M,
shows all possible gradations from a penciled to a barred condition.

0. 1 cock-feathered, Red-Breasted game with somewhat stippled feathers. The lower half
of each secondary is penciled (Hamburg type).

P. 1 intermediate male with yellow hackle that is black striped. He has a peculiar saddle,
the general color of which is reddish brown. Each feather has a faint black edge,
and is clear yellow along shaft; the rest of the feather is finely dusted on a yellow
background (plate 10, fig. 3). Breast feathers (in front part) are laced with
an outer black band edged with yellow. In the posterior region of the breast the
feathers are broadly laced. Wing-bow and coverts red, laced with black. The
exposed edges of primaries and secondaries are red-brown, the covered parts black.
The tail is black.

Back-Cross Hens.

A. Two dark Fi types. The breast is between stippled and penciled, the head is black.
C. (1) Sebright type. Very dark with much stippling. Some penciling on back. Breast
dark; neck like that of Sebright.

(2) Sebright type like (1), but not such clear yellow. Secondaries and tail feathers

and coverts stippled (with black tips).

(3) Yellow Sebright. Neck and breast yellow with black base and tip to feathers.

Cape, breast, and wings (except bow) penciled to barred.
C. to D. (1) Pale yellow, breast spangled, back lightly penciled, tail same. Secondaries
yellow and little stippled. Upper web of primaries stippled.

(2) Breast spangled, rest as in (1).

(3) Same as (2).

L. Dark Sebright. Back-feathers broadly laced and a little penciled. Neck black with
yellow centers to feathers and yellow edges (reversed Sebright).

All of the preceding hens except A are in general Sebrights. The
last three are pale stippled Sebrights.

Back-Cross Hen-Feathered Cocks.

A. One cock like Fi male, but rather paler on back.

C. Four cocks. Light Sebrights, but spangled, in general, instead of laced. Feathers

clear, not stippled.
G. One black.

L. Dark Sebright. Back and rump black. Feathers with narrow center, not stippled.
Q. Dark Sebright nearer to hen C (1). Thoroughly stippled with game tail. Neck and

breast dark Sebright. Probably a new class nearer to (C).
R. Two cocks. Pale yellow instead of reddish, and much less black than are other yellows.

No class of hens to match.1

In regard to color inheritance the preceding 19 birds are too few
to add anything of significance to the other results, except that they
serve to emphasize the dominance of the factors making for Sebright
coloration. The hen-feathered cocks confirm the other results as to
the dominance of the factor or factors in question.

There can be little doubt that some of these classes are complex.
They almost merge into each other and in one part of the body individ-

1 There is one other bird, not given in the above list, that is pure Sebright except that his legs
are yellow. Until I find out by further breeding of the Sebright stock whether yellow legs are
present in it, this case must remain doubtful. On the basis of a two factor color-difference one
Sebright (as to color) is expected in 16 birds, and one in 64 on a three factor basis. Some
Sebrights had been raised along with the back cross, hence the possibility of contamination.

RELATING TO SECONDARY SEXUAL CHARACTERS. 21

uals may grade off into one class, in other parts into other classes.
An almost continuous series of types might be arranged from black
to pale yellow.

The difficulty of matching the hen-feathered males to their genetic
mates is almost insuperable. In table 1 an attempt was made to put
these males with their respective females. The difficulty is, of course,
greater for the cock-feathered birds, even with the castration evidence
(that is too meager at present for the purpose) , but a few of the males
may be placed with certainty, and the rest guessed at.

One bird appears to be a hen-feathered game male resembling in
many respects the female game, but darker and redder. There is more
shafting on cape and wing-bow. The breast is unusually dark-salmon.
The hackle is darker than is the game female. Upper wing-coverts
broadly laced with black. (Plate 10, fig. 3.)

The occurrence of this hen-feathered jungle-fowl is so unique and
the coloration of the bird so interesting that I have added to the plates
three feathers of such a bird, viz., a stippled saddle feather, a feather
from the back, a hackle feather, and a wing covert with stippled center
and a black border. The neck hackle departs somewhat from the
hackle of the jungle-fowl hen, but in the same direction as does the
neck hackle of the Sebright cock from his hen.

Looking over the F2 group, the most noticeable thing is the large
number of blacks (E and G), all of which are stippled. Probably the
factor came from the game, because group E was present in the back-
cross as well as in F2, and because these black birds are always stippled.
The yellow color (land J) may have come from both, each breed having
then a black factor that, as a pattern, covers over most of the yellow.
It is difficult to distinguish penciling from stippling in the F2 yellows.
Without figuring each of these types, their description in detail is not of
much value. The skins will be deposited for reference in the Zoological
Laboratory of Columbia University.

C. Back-Cross of Fj to Game.

As the back-cross of the Fj to the game might appear more likely to
reveal the kinds of germ-cells present in the individual, the results
from such a cross may be given before discussing the genetic data. If
it were certain that the "game" contained all of the recessive factors
that are involved in the experiment, this method of testing the result
would be ideal, but there is no way of determining a priori whether
this is the case. The question will be taken up later. The pre-
sence of two kinds of males with corresponding but largely uncor-
related differences in their plumage makes their classification as a
group impossible. It is simpler, therefore, to put the females into their
classes first, after which the hen-feathered males may be expected to
fall into the same groups (or nearly so), while the identity of the cock-

22

THE GENETIC AND THE OPERATIVE EVIDENCE

feathered males, i. e., their class relationship can only be determined
for the classes that resemble the Fx and the Pi birds. The F2 hen-
feathered males can in part be further identified by means of the
evidence that castration of these types affords.

Two of the F2 classes of hens can be identified in this back-cross,
viz, (a) 4 hens like the Fi birds, (6) 3 hens like the game; (c) there were
3 other hens with plain yellow, i. e., not stippled backs. The upper
surface was like that of the game female, but much lighter. The first
two classes (a), (b) might be again split into two types. There were
only two hen-feathered males, one nearly like the Fi male, the other
blacker; they probably belong to different classes.

Of the 7 cock-feathered males, one was like the Fi castrated males;
another had a similar back, but a darker and differently marked
breast; 2 were game-cock type; 3 were odd birds much like the game
cock above except for absence of black, with reddish heads without any
black. The males may be approximately classified as follows :

Back-cross Fi 9 by game cf.

Hen-feath-
ered cf.

Interme-
diate d1-

Cock-feath-
ered cT.

F (or K) . .

1
1

1
2
2
2

(C)

B

A

7

Four or five types may then be recognized in this rough grouping.
None of the groups seem uniform and probably might be split again.

D. The Number of Color Factors Involved.

The theoretical expectation for two pairs of factors calls for 4 classes
in the back-cross, but this assumes that the parent type used for back-
crossing contains all (here 2) recessives. But this simple assumption
can not be true in this case, for the Fx bird would have been like the
Sebright. On the 3-factor assumption the expectation for the back-
cross is 8 classes, but this would apply only if the game were the triple-
recessive form, which, again, it is not, as shown by the F] cross. But if
the dominance of one or more of the Sebright color factors is incomplete,
then either a 2 or 3 factor assumption might apply to the back-cross.

If only 2 pairs of factors are present we should expect to recover the
game type once in 16 cases in F2. But, as will be shown, only 1 game
was recovered out of the 49 F2 females. This result fits better with a
3-factor assumption, for even with the small number in the back-cross
the indications are that more than 4 classes are present.

RELATING TO SECONDARY SEXUAL CHARACTERS. 23

In the F2 birds at least 11 classes may be distinguished, and some
of these appear composite. For 3 factors the maximum number of
possible classes (including heterozygotes) is 27. We can recognize at
least 11 F2 classes amongst the females alone, and a few others are
doubtfully present in the males.

In favor of the view that the heterozygous classes are here different
from the homozygous, the following evidence may be utilized :

(1) The Fx birds are entirely different from either parent and they
are heterozygous for all the factor differences between the two types.
The only alternative explanation for the intermediate condition of Fi
would be that each race carries one or more completely dominant
factors. But the latter view is improbable because more of each parent
type would then be expected in the F2 generation.

(2) In the F2 generation the Fx type is not as frequent as would be
expected on the view that the heterozygotes could not be distinguished.

E. Back-Cross of Fx 9 to Sebright cf.

It is possible to add, now, while this paper is passing through the press
(June 1919), the results of a back-cross of 4 Fi females to a Sebright
male carried out during the summer of 1918. The birds being now
mature their permanent colors are evident. Making the back-cross
in this direction is much less advantageous than the reciprocal de-
scribed above, because the Sebright contains most of the dominant
color factors. The group of birds obtained appeared to be less varia-
ble in color than those from the other back-cross, and one can see at a
glance that more of them approach the Sebright type; some quite
closely.

All of the males are hen feathered, as expected. No evidence was
found that two types of males exist, which would have been expected
if the two types noted in F2 had any hereditary significance. If, then,
as the F2 results suggest, two factors for hen-feathering are present
both are dominant, and no genetic distinction is found between in-
dividuals in which one or both of the dominaut factors are duplex
or simplex.

There were 9 adult hens and 10 hen-feathered cocks. An attempt
is made below to refer them to their corresponding F2 classes.

F. Review of the Heredity of the Color of the Plumage of

Poultry.

In poultry there are perhaps more different colors and color-patterns
than in any other species of domesticated animals. The genetic work
has advanced far enough to show that many of the differences depend
on Mendelian factors. It is probable that, in addition to the main
factors, there are many contributory, minor, or modifying factors that
give the finer details to "show birds."

24 THE GENETIC AND THE OPERATIVE EVIDENCE

It is generally supposed that the wild bird from which some at least
of the domesticated races have come is Gallus bankiva of India and
Indo-China, or else one or another of its subspecies. In any case,
the wild type of coloration is approximately known, since the known
wild races are colored alike in all essential respects. Even were the
color of the wild type not known, the original plumage could be
deduced with some degree of probability from the atavism that
appears when some of the races are hybridized. It is interesting to find
that many of the new plumage characters are dominant to the wild
type. The same relation also holds rather generally for other characters
of poultry, such as the comb, etc.

Amongst the uniform or single-colored races, the whites, blacks,
reds, and buffs have been studied. Bateson and Punnett were the first
to show that the white of the White Leghorn is dominant. They also
showed that the white of the White Rose Comb bantams is recessive.
Another white, that of the White Silky, is also recessive, but due to a
different factor from the white factor of the Rose Comb bantams; for,
when these two whites are bred together they give colored birds in the
first generation. Hurst showed later that the white of the Leghorn
is dominant over the black of the Hamburg and the buff of the Cochin.
The dominance is often not complete, since tints of black or of buff or
even patches of these colors may occur. The latter may be confined
to the head, neck, and breast. The black plumage of the Hamburg is
dominant over the buff of the Cochins, but incompletely so, as the
black background may be marked and shaded with brown. Whether
we are dealing here with one pair of factors, or two pairs, could only
be determined by an F2 ratio; whether it is 3: 1 or 9: 3: 3: 1.

The blue color of the Andalusian is known not to be a simple color,
but to be a fine mosaic of splashed white and black. The color is
produced in birds that are heterozygous for splashed white and black,
or at least for certain kinds of white and black. This relation was first
demonstrated by Bateson and Punnett (1902 and 1905) and later
Saunders (1906). It appears also from certain crosses made by Daven-
port that some of the whites (such as that of the Leghorn) and black
(such as that of the Minorca) may at times also give some blue birds
when crossed. Whether there are also other races with dominant
white color different from that of the Andalusian white (and the same
holds for black races also) or whether a special (recessive) white was
present in this cross when the blue appeared, was not made out by
Davenport.

Lippincott has recently studied the Andalusian cross and obtained
essentially the same results as his predecessors. He calls attention to
an interesting fact in the splashed whites, namely, that the color
splashes are blue when thay are found in those parts of the body where
the color is blue in the Andalusian. Although the Andalusian is always

RELATING TO SECONDARY SEXUAL CHARACTERS 25

spoken of as a blue bird, the hen only is entirely blue, while the male is
black above and blue below. The splashes on a white male correspond
to the black and blue of the Andalusian male, and are black if above
and blue if below.

Lippincott found also that the blue birds differ from the black in two
characteristics, viz, in the blues the pigment is in larger masses, i. e.,
it is more clumped, leaving more white between the clumps than in
the blacks, and in the blues the pigment is absent in the extremities
of the barbules. If the clumping and the condition of the barbules are
treated as separate entities, each gives a 3: 1 ratio. Lippincott con-
cludes, therefore, that the Andalusian cross is a 2-factor case. If each
of these characteristics was independent of the other in the sense that
some birds had clumped pigment and others deficiencies in the barbules,
then one might conclude that he was dealing with a 2-factor case; but
if these two characters are only different aspects of the same gene, and
when one is present the other is also, the situation is not different from
those that are very common, viz, two or more effects produced by the
same genetic factor.

Davenport has recorded results of crossing several breeds of different
colors (1906 and 1909). The white of the Leghorn was found dominant
to the black of the Minorca breed, although the hybrids, "at least the
females," had some black feathers. This white was also found to be
dominant to the mottled Houdan and to the "Red-backed game."
On the other hand, a male Tosa with wild-type plumage by recessive
White Cochin female gave " barred " males in Fij the barring coming
in, no doubt, from the Cochin and although not at the time recognized
by Davenport as sex-linked inheritance, the statement that barring
is "associated with maleness" (as already pointed out by Darwin)
indicated that the barring that appeared within the cross was probaUy
the sex-linked barring shown by other breeds.

In Davenport's cross of White Leghorn by Minorca two bines
appeared (as stated above), indicating that the same factors were
here present that in the Andulusian white and black strain gives the
same result,1 but why only some of the Fi appear as blue, while others
are not blue, is not yet made clear, unless two factors for white were
present. White of the Leghorn breed was found not to be as com-
pletely dominant over buff as over black. Black was found dominant
over the wild-type (Black-Breasted game), but red is present in 1 ■".
birds also to some extent in those places where red is found in the game.
Lacing, as shown by the Dark Brahma, is dominant to the plumage
of the Tosa. Penciling also is said to be dominant, as shown in females
of the cross between the Dark Brahma and Tosa fowl.

In his later paper (1909) Davenport gives fuller information in
regard to some of the Fx cases reported in his first paper, as well as the

1 Provided that the blue classification was based on the adult plumage and not on down r,.\nr.

D. H. HILL LIBRARY
Norfh Carolina State College

26 THE GENETIC AND THE OPERATIVE EVIDENCE

F2 results. Thus, in the cross of Silky to Minorca, that gives black
Fi birds, the F2 count gave 210 black, 57 game, and 95 white — approxi-
mately the expectation for two pairs of factors, one of them giving
white (9:3:4). Silky by White Leghorn gave white Fi's, but the males
developed red on the wing bow and saddle when they became mature,
and the female a faint blush of salmon ("red") on the breast. In F2
there were whites, games, and blacks, approximating to expectation for
three pairs of factors, one being a dominant white (52: 9:3). Silky
by Buff Cochin gave a washed-out buff, but with the jungle coloration
partly developed in the tail (black) and hackles and wing bow (redder
buff). Davenport represents the Buff Cochin as having lost the jungle
patterns and coloration, while the Silky retains it. The heterozygous
condition of the genes for the wild-type color in Fi is made responsible
for the part development of color. The White Silky is represented as
carrying the factor for black (N), hence in F2 both black and game-
colored birds are expected and they were obtained. When Black
Cochin is crossed to Buff Cochin, the Fx males are in general like the
game (black and red) while the females are black (except for some red
on the hackle). In this case Davenport represents the Black Cochin as
showing a factor for jungle-fowl pattern, but lacking the color that is
assumed in his other formulae to go with this pattern. What is meant
by this change is not quite clear to me, unless Davenport supposes there
is an independent factor for the jungle-fowl pattern which may be
filled in by other colors determined by other factors. But were there
enough Fi birds to exclude the possibility that jungle-fowl birds would
not appear in this cross?

Davenport has reported a cross between a female White Cochin and
a male Tosa (wild type) from which the daughters were Tosa, except
that the shafting was broadened, and the saddle feathers and proximal
secondaries were obscurely barred (black and buff) ; the sons were also
like the Tosa, but every feather was repeatedly barred (see above).
In F2 there were 15 white, 25 game, and 16 barred birds. Davenport
concludes that "barring is clearly heterozygous and confined to the
male sex," and in a footnote he adds that the sex-linked barring factor
of the Plymouth Rock is different from that of this Cochin-Tosa cross,
but Goodale informs me that the barring that appeared in this cross
is probably the same as that in Barred Rocks.

As pointed out, an interesting feature of color inheritance in poultry
is the large number of cases of sex-linked inheritance. It might seem
probable here, as in the case of Drosophila, that this is due to a well-
recognized difference between sex-linked and autosomal characters,
namely, that a recessive mutation in one of the sex chromosomes of a
sperm-cell of the male bird will have a chance of showing its effect
immediately if that sperm-cell unites with an egg without a Z to form
a daughter, whereas it would not immediately show up in the offspring

RELATING TO SECONDARY SEXUAL CHARACTERS. 27

if the mutation were autosomal.1 In consequence the recessive mutant
would have a greater chance of being observed and selected if it
appeared in a sex chromosome. But dominant sex-linked characters,
however, have the same chance as dominant autosomal ones and the
question turns therefore on the kinds of characters shown in the cross.

The first indication of sex-linkage in fowls was furnished by evidence
that Spillman published in 1903 on information supplied by poultry-
men — information that has been proven subsequently to have been
accurate. Spillman pointed out clearly the similarity between the
facts he quoted and the then known cases of sex-linkage in the canary
and in the currant moth. The case referred to by Spillman was a cross
between Barred Plymouth Rock and Black Langshan. Goodale and
I repeated the cross, using both Plymouth Rock and American Dom-
inques, publishing the results in 1912. In addition to the Fi results
evidence was obtained for the F2 generation. The theory was also
tested by back-crossing. The results of such a cross that are typical
for all cases of the sort are briefly as follows : Plymouth Rock cock by
Langshan hen gives Fi barred sons and barred daughters. These
inbred give F2 barred cocks and barred and black hens (2:1:1).

In the following schemes the sex chromosomes are represented by Z
and W, while the exponents stand for the factors involved, viz, B for
barred and b for not-barred, which here means a black bird.

Barred $ Black 9

?i zBzB zV

zBzb zV

F> BarredcT Barred j

-ft -ft zV zV

Barredc? Barredd* Barredj Blacky

In the reciprocal cross, a black cock was mated to a barred hen.
The sons were barred, the daughters black (F,). These inbred gave
(F2) barred males and females, black males and females in the ratio
of 1:1:1:1. The chromosome scheme of inheritance is as follows:

B&ckd"

n

zV zV

1 Barred^" Black j

zBzb z*zb zV z\v

Barredc? Black c? Barred<j> Black j

1 If the recessive mutation occurs first in the Z chromosome of an egg of the female it will not
appear in the next generation; then if it has passed into a male, half hi.s daughters will show it.
The single factor-pair involved is carried by the sex chromosomes Z Z.

28 THE GENETIC AND THE OPERATIVE EVIDENCE

One back-cross test consists in mating the Fx barred males ZBZb
(from both crosses) to a pure black female. The expectation is for
equal numbers of barred and black males and females, and the result
was realized. The Fx barred hen of the first cross (ZBW) back-crossed
to a black cock is expected to give only barred males and black females,
and this result also was obtained. The explanation of the last cross,
based on the sex chromosomes, is as follows :

Sack <? £ Barredj

Before these experiments were finished Goodale had made other
crosses involving the barring factor, and had obtained results that
showed the sex-linked inheritance of this factor (1909). For example,
he crossed Buff Rock male (not barred) to white Plymouth Rock
females. The sons were barred and the daughters not barred. The re-
ciprocal cross gave barred sons and daughters. A White Rock male
(carrying barring) mated to a Brown Leghorn female gave barred sons
and daughters. Reciprocally, the chicks were of two kinds as to their
down, viz, black chicks and chicks with the down pattern of the barred
rock. All these results with Barred Plymouth Rocks show that they
carry a sex-linked dominant factor for barring. Its wild-type allelo-
morph would be game-color (jungle-fowl), but since, when the dominant
barring is absent in some of the individuals in these crosses, they are
black, it would seem to follow that another dominant factor, one for
black, that is not sex-linked, is also present.

Pearl and Surface have also carried out crosses with Plymouth
Rocks on a much larger scale. Their results conformed in every way
to the foregoing. They crossed Barred Plymouth Rocks and Cornish
Indian games. The plumage of the male of the latter race is black
with dark red on the back and wing-bows; the females are also black
laced with mahogany ground-color on back, breast, wing, and tail
coverts. When the male game is mated to the barred hen the sons are
barred and the daughters are black. In the reciprocal cross both sons
and daughters are barred. The back-cross tests conformed to expecta-
tion. The results were the same as those already stated above for the
Langshan-Rock cross.

Sturtevant crossed Columbian Wyandottes and Brown Leghorns.
The Fi sons were alike, whichever way the cross was made. They were
fairly typical Wyandottes, which race carries therefore more of the
dominant plumage characters (two or three?). There were two types
of daughters, depending on the direction in which the cross was made.

RELATING TO SECONDARY SEXUAL CHARACTER-. 29

When the father is Wyandotte, the daughters are like him (except for
stippling of the Leghorn type). When the father is Brown Leghorn
the daughters are somewhat stippled red birds. In the former case
the daughters getting their Z chromosome from their Wyandotte
father resemble him; in the latter case the daughters getting their Z
chromosome from their Leghorn father look more like him. Their
failure to look exactly like him must be due to autosomal factors
derived from the Wyandotte mother that dominate other autosomal
factors from the father.

Hagedoorn crossed Black Breasted Game bantams (like those used
in my Sebright crosses) to Brown-Breasted bantams. In the latter
the black breast feathers of the male are bordered bjr lemon ; the hens
are nearly black. Black-breasted male to ''brown-red" female gave
both black-breasted sons and daughters. In the reciprocal cross all
the sons were black-breasted (like the mother) and all the daughters
were brown red like the father. Evidently the factor here for Brown
Breasted game is sex-linked and recessive. In this case the new
mutant sex-linked character is recessive to the wild type.

Davenport (1912) crossed Brown Leghorns to Dark Brahmas. In
the cross and its reciprocal all the sons are alike. Two dominant sex-
linked factors were found,1 viz, the white background characteristic of
the Dark Brahmas and the red upper wing-coverts (and back) charac-
teristic of the Brown Leghorns. On the other hand, the daughters
differ in the two crosses, in each case resembling their father in their
hackle color.

When two sex-linked characters are involved in a cross it is possible
to determine by suitable matings whether an interchange between the
chromosomes that bear them has taken place. In the case of the sex
chromosomes only one sex, the male, has both like chromosomes, viz,
ZZ, and we expect from analogy with the Drosophila work that
crossing-over would be found between the sex chromosomes only in
the male. Goodale has recently (1917) made the important discovery
that in poultry crossing-over takes place between the sex chromosomes
(ZZ) in the male, but not in the female (ZW or ZO). This relation,
therefore, is the reverse in birds and flies, for, in the one, crossing-
over takes place in the female and in the other in the male. Whether
this difference extends also to the other chromosomes in birds as it
does in flies is as yet not known.

Several years ago some crosses between gold and silver Campinee
were reported by Rev. E. Lewis Jones. The results are consistent with
the view that a sex-linked factor pair is responsible for this difference
in color, although the author does not apply this view to his results.
The results may be seen in the table on page 16, to which Jones has

1 One may be either sex-linked or sex-limited so far as the evidence gOW.

30 THE GENETIC AND THE OPERATIVE EVIDENCE

prefixed the number of individuals. The cross also involved hen-
feathering versus cock-feathering, which appears here (as in other
cases) to be a non-sex-linked dominant factor. As stated above there
are in the results a few apparent inconsistencies with this interpre-
tation, due possibly to heterozygous females having been used in the
crosses.

Lefevre crossed Silver Spangled Hamburgs and Brown Leghorns.
The spangling was found to be a sex-linked dominant factor. A span-
gled cock bred to a Leghorn hen gives spangled sons and daughters; a
spangled hen by a Leghorn male gave spangled sons and not spangled
daughters. The daughters do not transmit spangling. Other factors
may obscure the results, especially factors for black, or the localization
of the pattern. Lefevre says "it would seem probable that multiple
factors for black, introduced by the Brown Leghorns, are present, and
that these factors may have a cumulative effect, with the result that
pigmentation is developed to varying degrees of extension." Whether
the factors for black spoken of as coming from the Leghorns are
dominant wild-type factors that have mutant allelomorphs in the Silver
Spangled Hamburg is not entirely clear from the quotation.

Baur gives in his Introduction to the Study of Heredity (1914,
pp. 202-203) some results (unpublished) that Hagedoorn had obtained
by crossing gold and silver races of Assendelver birds. The factor
is sex-linked and is no doubt the same factor reported by Jones for
gold and silver Campines and by Sturtevant for Columbian Wyan-
dottes. Silver dominates gold and the sex relations are the same as
those already reported by others for poultry, viz, the male is ZZ, the
female ZW. Gold hens by a heterozygous silver1 gave 162 silver cocks,
163 silver hens, 168 gold cocks, 160 gold hens, expressed graphically
(g for gold, s for silver) :

Zg-W9XZ — Z^cT
Z*Z»— Z"Z"— Z3W— z»w

Silver Gold Silver Gold
male male female female

When a silver hen was united to a gold cock there were 246 silver
cocks and 243 gold hens — crisscross inheritance.

Summary.

From the standpoint of the Brown Leghorn type representing the
wild type, the following colors and patterns represent dominant
mutations from that type:

Dominants.
White of White Leghorn. Barring of Plymouth Rock.

Silver of Dark Brahma. Black (?) of Plymouth Rock.

Black of Minorca. Buff (or red).

Lacing of Brahma.

1 No mention is made by Baur that a heterozygous male instead of a pure silver male was used,
although the male is made heterozygous in the formulae.

RELATING TO SECONDARY SEXUAL CHARACTERS. 31

Each of these (in heterozygous condition of course) is dominant;
in some cases completely so, in others incompletely dominant. At
three different loci in the sex chromosome a dominant mutation has
occurred ; at three loci in other chromosomes dominant mutant changes
have also occurred.

Recessives.

White of Rose Comb bantam. Brown of Brown-breasted game.

White of Silky. Penciling.

White of White Rock.

Whether the recessive white that is sometimes found in dominant
White Rock stock is different from both of the other recessive whites
is not known. There are, then, 5 or 6 recessive characters that are not
sex-linked and 1 recessive sex-linked character.

Owing to the relatively large number of color dominants in poultry,
some unnecessary confusion has arisen concerning the relation of the
dominants to the wild type, and especially to other mutant characters
to which they are said to be dominant, in the sense, however, of being
epistatic. An imaginary example will illustrate this. For example,
if at some locus in the wild type a mutation occurred that gave a
dominant black (i. e., a black that shows up when one gene for it is
present) and at the same time this black also showed up even when
other recessive mutant characters were present in homozygous form,
then Fi birds would be black when black is crossed to such pure
recessive stocks. Such cases have indeed been described as dominant,
but a knowledge of F2 would have shown at once the error of such a
system. For, if black had been a real dominant, the F2 would have
given 3 blacks to 1 of the other type (such as the wild type), but if the
case were one of epistasis, then there would have been 9:3:3:1
classes in F2 (or some modification of that ratio). In this sense, then,
epistasis may be defined as a result that appears when one member of
the pair of genes produces its effect regardless of the constitution of the
individual with respect to another gene (or other pairs of genes). It
is curious at least to note that in the case of dominant white the term
epistatic has been much less often used than in the case of black.
Theoretically the two situations are exactly alike, but because black
could so obviously conceal things beneath it, while white is not thought
of as doing so, it seemed "natural" to make such a distinction. In
reality it is not a question of covering up at all, but a case of a dominant
character (white or black) preventing other colors from appearing.

In the case of recessive white the situation is somewhat different
and no one, so far as I know, has gone so far as to speak of such a white
as epistatic, although when the animal is white it certainly hides, when
completely effective, all the other effects of color-producing factors,
but allows them, to "show through" in some of the cases. This
means not that they do "show through," but that they only develop

32 THE GENETIC AND THE OPERATIVE EVIDENCE

to a "lower" degree. The difference between dominant and recessive
whites rests on the fact that in one case one member of a pair of fac-
tors gives white and in the other both members are necessary. But
obviously such a distinction is not important, and if it were worth
while the case might be argued for recessive whites being also epistatic.
The whole tangle goes back to a false interpretation of presence and
absence of characters and presence and absence of factors. As I have
gone over this ground recently in my paper on the Theory of the Gene,
I need not repeat here what I tried to make clear there.

ENDOCRINE CELLS IN OVARY AND TESTES OF BIRDS.

The occurence of gland-like cells with an internal secretion in the
ovary and testes of fowls has been described by a number of writers
and denied, at least for the testes, by others. The work of Boring and
Pearl has done much to bring this question to a satisfactory solution,
for they have tested out and made use of the best reagents that their
predecessors had discovered and have used a much greater amount of
material. As they have reviewed very fully the literature of the subject,
it will not be necessary to go over the ground again in detail.

In the follicles of the ovary there are present, according to Boring
and Pearl, groups or nests of cells lying among the connective tissue of
the inner theca. The cells are about three times as large as the ordi-
nary connective-tissue cells of the ovary. The cytoplasm is clear and
vacuolated, "only occasionally containing a few acidophile granules
which stain with the fuchsin in Mallory's stain or the eosin of Mann's
stain, while the real interstitial cells are crowded with granules."

When the egg is set free from its follicle, the latter collapses and the
rupture becomes closed. A mass of cells collects in the center of the
collapsed structure which develop yellow pigment. The cells, lying in
the puckered edge of the follicle, may also develop such yellow color.
The cells that produce the yellow pigment come from the nests of cells
that lay originally mainly in the theca interna. Either by migration
or by division they come to fill up the central cavity. The yellow
substance in the cells is not fat, since it does not dissolve in the clearing
oils, nor can it be protein, for it does not take acid stains as normal
secretion granules of protein. It does not dissolve in HC1, HN03,
or H2S04, nor in strong KOH, although the latter turns the pigment a
bright red color. Many other substances were also tried by Boring and
Pearl, but none of them dissolved the yellow pigment, which reacts
in this respect in the same way as does the yellow pigment in the
luteal cells of the mammal. The similarity in the nature of the pig-
ments in the two cases is an argument in favor of the view that the
cells that produce the pigment are the same in both groups. In the

RELATING TO SECONDARY SEXUAL CHARACTERS. 33

mammal the yellow corpus luteum is a large, gland-like organ that
develops after the ovum is discharged; in the bird there is also a yellow
spot on the ovary, due to the pigment in the collapsed follicle, but it is
smaller and much less conspicuous than in the mammal. The evidence
concerning luteal cells in the testes of the bird is conflicting. One of
the difficulties in the situation is the identification of the cells, which
are sometimes regarded merely as the general connective-tissue
stroma of the testis that is undoubtedly present; at other times special
secretory cells are discerned embedded in the connective tissue, as
individual cells or in islands. Boring states (1912) that in newly
hatched chicks about half of the tissue of the testes is interstitial con-
nective tissue; the other half consists of tubes or cords whose principal
function is the development of the germ-cells. In the paper of 1912
Boring reached the conclusion that there are no "interstitial cells in
the testes of the domesticated chicken in the sense that this term has
been previously used," and states that no evidence has been found that
an internal secretion of any kind is formed by any cells of the interstitial
tissue.

It is not necessary to discuss whether or not connective-tissue cells
are present in the testes of birds, for is it generally conceded that they
are found at least in certain stages, but it is important to look into the
question as to whether among these interstitial cells there are others
that have an endocrine function. Mazzetti gives pictures of such
gland-cells between the seminal tubules of the cock bird, but says that
they are rare, "even though this bird has very marked secondary
sexual characters" (Boring and Pearl). It maybe remarked parenthet-
ically that if they had been more abundant the bird might have had
no secondary sexual plumage since it will be pointed out below that such
glandular cells may have as their special function the suppression of
these characters.

According to Des Cilleuls, interstitial cells are first found in males
about 30 days old and at this time the secondary sexual characters put
in their appearance. If, as will be shown in the sequel, he means by
interstitial cells the endocrine cells that suppress the development of
the male plumage in the female, the appearance of these cells at this
time would be significant; but if he implies that their occurrence in
the male incites the development of the secondary sexual characters,
his interpretation is open to serious doubt. Reeves found interstitial
cells in testes of cocks 3, 5|, 9, and 18 months — more in the earlier
stages.

In a later communication by Boring and Pearl the whole question
is taken up again with improved methods, etc. Previously 21 male
birds had been studied, just hatched to 12 months old. More Mo-
tions of this same material were made which were stained according

34 THE GENETIC AND THE OPERATIVE EVIDENCE

to Mann's and Mallory's methods. In addition, a whole new series
of preparations was made. A few interstitial cells, i. e., granule con-
taining-cells were found in newly hatched chicks, but not in any of
the 60 mature birds examined.

LUTEAL-CELLS IN THE TESTES OF THE MALE SEBRIGHT.

Finding that the testes of F2 hen-feathered birds were often flat and
pear-shaped instead of rounded and cylindrical, as in ordinary cocks,
and that they were often black in color, suggested, as already stated,
that the testes of the Sebright might be hermaphrodite in some element.
It seemed not impossible that egg-cells might be found. I made a
considerable number of sections of the testes of these birds and
examined them under the microscope; not finding any egg or egg-like
bodies, the slides were laid aside, but the idea that in some other way
the Sebright's testes might correspond to the ovary of the female next
recurred to my mind. Consequently, when in the summer of 1918 I
had some new material derived from a castrated Sebright male that
had partly regenerated its testes and was again going back to hen-
feathering, and pieces from one of the old testes of a castrated bird,
I asked Miss Boring, who was then in Woods Hole, to make some
preparations and examine them to see if she could detect any such
elements in them as she had found in the female. Miss Boring
reported the occurrence of luteal cells in the testes from hen-feathered
males, and the results have been published in a brief preliminary paper
(1918). The abundance of these clear cells, supposedly gland-cells
with endocrine influences, in the testes of hen-feathered birds is in
sharp contrast to their absence in the normal adult cock birds. It
seems to follow, therefore, that the hen-feathering in the Sebrights is
due to the presence of these cells, whose function is the same as of the
similar cells in the female, i. e., the suppression in both of cock-feather-
ing. Castrating the Sebright produces its effect by the removal of
these cells that are responsible for the suppression of cock-feathering.

The occurrence of luteal cells in young stages of other races of poultry
raises the question as to whether in these races the first or juvenile
plumage, that resembles that of the hen rather than that of the cock,
may not also be due to an internal secretion from these cells, or whether
this juvenile plumage is only the plumage of a characteristic stage in
development. Castration of young chicks ought to settle this point.
Such castration experiments have been made by Goodale. The
absence of any reference to any effect on the juvenile plumage in
these early castrated birds probably meant that they did not develop
precociously cock-feathering, and he writes me that he examined them
carefully and that their plumage is like that of the normal chicks.
Geoffrey Smith has reported the occurrence of two kinds of males

RELATING TO SECONDARY SEXUAL CHARACTERS. 35

in a race of Leghorns, the males of one of which become cock-feath-
ered before the other. May not this difference depend on the length
of time endocrine cells remain or begin to develop? A histological study
of the two types would be of the greatest interest.

ENDOCRINE CELLS IN THE TESTES OF MAMMALS.

In man and other mammals it has long been recognized that in
addition to the germinal cells of the testis there are also present other
cells, sometimes called interstitial cells, that, so far as known, have no
immediate function in connection with the germ-cells, or at least that
have other important functions outside the relation to the reproductive
organ. That some internal secretion from these cells has an important
influence on the secondary sexual characters rather than anything done
by or produced by the germinal cells has been very clearly shown by
evidence derived from three separate sources, namely, from the
operation known as vasectomy, from an exceptional condition known
as cryptorchidism, and more indirectly from X-ray treatment. Vasec-
tomy involves either cutting the vasa deferentia in such a way that
the cut ends do not reunite. In consequence of the closure of the
outlet of the testis the germinal cells slowly degenerate, and finally
completely disappear. How such an effect is produced we do not know.
That this result does take place is borne out by the unanimous testi-
mony of all those who have successfully performed the operation.
Ancel and Bouin showed (1903) that breaking the continuity of the
vas deferens suppressed spermatogenesis in 8 to 12 months. Both
the Sertoli cells (the nourishing cells of the germinal epithelium) and
the interstitial cells persist. Such animals remain sexually active and
their secondary sexual characters are not affected. Marshall states
that in the hedgehog the remarkable periodic enlargement of the
testis takes place even after vasectomy, although the germ-cells have
disappeared.

In mammals the testes fail at times to pass through the inguinal
canal, and, in consequence of their retention in the body-cavity, the
germ-cells fail to develop. On the other hand, the interstitial cells
of the testis develop normally. Cryptorchid individuals show the
normal secondary sexual characters of their species. How retention
of the sperm should give rise to the same result as cutting the duet,
viz, absorption of the germinal cells, is not known. A possible solution
may be found in the pressure exerted on the testes, both when retained
in the abdomen and when their outlets are stopped by tying or cut tint:
the ducts.

Finally, it has been long known that continued or repeated exposure
to X-rays or to radium causes the destruction of the germ-cells, but
leaves the interstitial cells intact and presumably functional. Destruc-

36 THE GENETIC AND THE OPERATIVE EVIDENCE

tion of the germ-cell by X-rays has no effect on the secondary sexual
characters.

This threefold evidence demonstrates that in the male of the mam-
malia most, perhaps all, of the secondary sexual characters that are
affected by castration are not affected by the destruction of the germ-
cells. This conclusion supports very strongly the view that the inter-
stitial cells are the cellular element in the testes that influence through
internal secretion the development of the secondary sexual characters
of the male.

Equally important are the results that relate to the accessory organs
of reproduction, such as the glands that open into the vas deferens
(prostate, Cowper's gland, etc.) and the copulatory organs also. In
the castrated mammals these organs diminish in size. On the other
hand, after destruction of the germ-cells in the testes (or even when
they fail to develop as in cryptorchid individuals) these accessory parts
are unaffected. In birds, as will be shown, the situation is entirely
different.

CYCLICAL CHANGES IN THE INTERSTITIAL CELLS IN
HIBERNATING MAMMALS.

The changes that take place in the interstitial cells in mammals that
hibernate and in which there is a definite rutting season following
hibernation have been examined by several workers. The mole has
been studied by Regaud (1904), Lecaillon (1909), Tandler and Grosz
(1911); the marmot by Hauseman (1895) and Gaugini (1903); the
hedgehog by Marshall (1911); and the woodchuck by Rasmussan
(1917). In the mole the interstitial cells are most abundant when the
tubules in which the spermatogenesis is taking place are least devel-
oped, and vice versa. In the hedgehog the increase in both tissues takes
place at the same time. In the woodchuck both tissues increase rapidly
after hibernation (during March and April), after which the spermato-
genesis continues actively for the two following months (May and
June), while the interstitial cells retrograde rapidly during April and
remain at a low level for the rest of the year. Retrogression in the
germinal epithelium begins in July, after the rutting season is past.
It appears from this evidence that the activity of the two tissues does
not always run the same course. Since the secondary sexual characters
of the male, which are not well developed in these animals, are not so
far as known affected by the condition of the testes, the evidence does
not have any very direct bearing on our present topic. How far the
sexual behavior of these mammals is determined by the quantity or by
the activity of the interstitial cells is not very clear from the evidence,
although there is a very noticeable increase in the amount of this
tissue just before and during the rutting season. In the mole also the

RELATING TO SECONDARY SEXUAL CHARACTERS. 37

interstitial cells begin to increase just before the mating season, and
the increase continues for several months after mating has taken place.
It is difficult to judge how great or how little the change amounts to
unless the whole organ is considered, for the relative volumes of the
seminal tubes and the interstitial tissues does not give a measure of the
total volume of these tissues, since the testes may decrease greatly
in size when the seminal tubes retrograde, and the apparent increase
of the interstitial cells at the time may not increase the total amount
of that tissue present.

Probably more important than the ratio of interstitial tissue to
tubules is the activity of the former. Rasmussan states that in the
woodchuck the interstitial cells not only increase in number immedi-
ately after hibernation, but the increase in amount of this tissue is
largely due to increase in the cytoplasm, in which there appears an
accumulation of fatty globules in the more peripheral parts of the cells.
In the central cytoplasm an abundance of fine lipoid granules develops.

Marshall has made some interesting experiments on the hedgehog
at different seasons. Castration in March prior to the breeding-
season has an influence on the accessory generative organs (vesicuke
seminales, prostates, and Cowper's glands). They remain in the same
undeveloped stage in which they were at the time of operation. If
castration is carried out very early in the breeding-season, when the
accessory reproductive organs are about half developed, their further
enlargement is prevented. In so far as the accessory organs rank as
secondary sexual organs, their complete development is thus shown
to depend on the testes. Transection of the vasa deferentia before the
beginning of the breeding-season affects somewhat the enlargement of
the testes, but produces no effect on the accessory organs.

HERMAPHRODITISM IN POULTRY AND THE SECONDARY SEXUAL

CHARACTERS.

Several hermaphrodite birds have been described (Brandt, 1889;
Shattock and Seligman, 1906; Pearl and Curtis, 1909; Smith and
Thomas, 1913; Bond, 1914; etc.). The most recent and complete
account of such birds is that by Boring and Pearl. They examined in
all 8 hermaphrodites, or at least 8 birds that showed in their plumage,
or other secondary sexual characters, peculiarities of both Bexes. Five
of the birds came from Herr Houwink in Meppel, Holland, who had a
stock in which there appeared, in 1911, two hermaphrodites out of 80
birds, and in 1912, three out of 80 birds. These were the birds studied
by Boring and Pearl. In addition, when Pearl saw Horr Houwink'a
birds in 1910, " there were then on hand a considerable Dumber of these
supposed hermaphrodite birds." An anatomical study of the Holland
birds showed that one of them was nearly a normal female; three, the

38 THE GENETIC AND THE OPERATIVE EVIDENCE

authors say, were ''evidently undeveloped females. They have infantile
oviducts and embryonic ovaries." It should be added that there was a
tumor more than twice the size of the ovary attached to or part of the
ovary. If the ovary itself was affected by the tumor, or the tumor was
a part of the ovary, the slightly unusual condition of the birds might
be accounted for. Of the other 3 birds, 2 are also suspected to have
ovarian tumors, while in the third bird streaks of a secretion which
resembles the substance of the tumor of the other two were found.
The change towards male plumage in these 5 birds is probably due
either to the incomplete development of ovary or to the effect of the
tumor on the ovary. Although luteal cells are described as present, it
seems probable that their total number might be less than in a normal
bird, and hence their insufficient secretion would fail to suppress the
development of male plumage. From this point of view these birds
are no more hermaphrodites than is a hen with her ovary taken out.

The remaining Holland birds were entirely different. On the left
side there was an ovary in an inactive condition ; on the right side there
was a testis, producing spermatozoa. Sections of the testis show
that it is normal, consisting of a mass of tubules with very little con-
nective tissue between them. In both ovary and testis there are "a
few nests of luteal cells near the surface. The ovary contains eggs, but
is abnormal to some extent." The authors state:

"In external appearance it is more like a male than the others, which fact
correlates well with the active condition of the testis and inactive diseased
ovary, with only one corpus luteum scar. The interstitial cells can scarcely
be held accountable for the male secondary sex characters, as the only ones
in an active secreting condition are a few in the ovary."

It is not quite clear what is meant in this quotation by the statement
that the interstitial cells can scarcely be held accountable for the male
secondary characters unless to suggest that they cause the development
of these characters in the male, as they are supposed to do in mammals —
a view that the authors do not seem at other times to hold.

Another hermaphrodite (Atwood's black) had an infantile oviduct
and an ovotestis. A second bird, too, had an ovotestis — mostly
testis — as well as a rather large oviduct. Collections of luteal cells
are described between the tubules of the testicular portion. If, as
suggested by the Sebright cases, these cells tend to suppress the female
plumage, their presence here in excess might at least be made to account
for the female part of the plumage of this bird. Comparing the last
two birds (that showed active sex-behavior as males) with the best
of the Holland birds, Boring and Pearl point out that the active sex
behavior of the two former can not be due to "interstitial cells that are
absent in these but present to a slight extent in the former." They
then add " .... though the differences can not be laid to the lutear cells,

RELATING TO SECONDARY SEXUAL CHARACTERS. 39

as they are present in all three." That the relative amounts of the
latter or their activity might still be accountable for the difference
would not seem entirely excluded from the evidence so far as it is given.

A fourth hermaphrodite (Dexter's) laid 12 eggs and had a large
coiled oviduct. There was present "a large, lobulated reproductive
organ on the left," which proved to be an ovotestis. Several ovarian
tumors were present and there was testicular tissue.

It is fairly evident, then, that four of these birds described by Boring
and Pearl were females with abnormal ovaries. The incomplete
development of the latter, or their abnormal condition due to tumors,
may sufficiently explain the occurrence of male secondary sexual
characters. That these tumors affect, to different degrees, such charac-
ters is expected from what is shown by imperfectly spayed females of
normal breeds.

There are a few statements in the summary of this paper that call for
comment. The statement that the "development of comb, spurs,
and wattles does not stand in direct quantitative relation to the Bex
of the gonad," appears to be only intended as a statement of fact based
on the author's observation. But in what sense is there an expectation
that they should stand in such relation beyond the obvious fact that
in the cock the comb and wattles are larger than in the hen, and that
spurs are generally present only on the cock. But if the expression
"sex of the gonad" implies the germ-cells it is not at all certain that
there is any expectation of a quantitative relation, and there is some
probability at least that other cells than the sex-cells are involved in
the development of combs, wattles, and possibly spurs. A castrated
cock has a small comb resembling that of the female bird. On the
other hand, removal of the ovary sometimes leads to an increase in the
comb and wattles. Here we have, to say the least, a paradoxical
situation, for the result looks superficially as though something in the
ovary keeps down the hen's comb, while something in the testes kei
up the cock's comb, yet when the ovary is removed the hen develops
a cock's comb; when the testes are removed the cock develops a hen 'a
comb. The real meaning is, I think, that the genetic complex for
femaleness (one Z or else ZW) stands in itself for a full-sized comb,
while the genetic complex for maleness (two Z's) stands in itself for
small comb.

Boring and Pearl state that "body-shape and carriage have a genetic
relation to the sex of the gonad." This statement means, I think,
that the amount of testicular matter present stands in some direct
relation to the shape of the body and carriage of the male. ( lastratioo,
both of the normal cock and the Sebright, seems to change the carri-
age somewhat and perhaps the shape. Both lose something of the
peculiar attitude of the male, but I have not been able to my own

40 THE GENETIC AND THE OPERATIVE EVIDENCE

satisfaction to analyze what this means. As has been pointed out, and
as the pictures show, the castrated Sebright changes his attitude, but
whether this is a change due to his new contour, or to a new balance
resulting from a large tail, or to a let-down resulting principally from
effects on the nervous system, is difficult to determine. The same
statements apply in part to the castrated cock of ordinary breeds, but
not to the same degree, since the change after castration, in feathering
and in carriage at least, is slight.

The conclusions that the "amount of lutear cells or pigment (?) is
in precise correlation with the degree of external somatic femaleness
exhibited by the individual" is of especial interest in connection with
the Sebright evidence. It is difficult, however, to gather from the
body of the paper what the absolute amount of luteal cells is that is
present, for even in some of the more male-like birds with an ovotestis
the description leads one to suppose that there may be as much luteal
material present as in some of the more female birds with infantile
ovaries or cystic tumors.

Pearl and Curtis (1909) described "a case of incomplete herma-
phroditism" in a Barred Plymouth Rock fowl. Externally the bird
looked like a hen, but "the head and neck resembled these parts in
the cockerel," especially the comb and wattles. The bird was never
seen to tread a hen, nor did it ever crow normally. An ovary and
oviduct were found on the left side, the former no larger than that of a
laying hen after removal of the large yolks. No eggs were visible on
its surface. On the right side a testis (9 mm. by 6 mm.) and vas
deferens were present. No eggs were found in the ovary, and it gave
every indication of being in a degenerating condition, with no eggs
or egg follicles in it. The testis had no "normal seminiferous tubules,"
but indications of cellular rods were present. The organ is in all
probability a degenerating testis.

A Leghorn 2 years old has been described by Shattuck and Selig-
mann (1906) that had the full-developed comb and wattles of the
cock, but the former drooped slightly to one side as in the hen. Well-
developed spurs were present. The plumage was mainly female, with
neck-hackles moderately developed, and with "saddle-hackles" prac-
tically absent. The tail, though not typically female, lacks sickle
feathers. The bird excited no notice from other birds of either sex.
A large left oviduct and the distal end of a right oviduct were present.
Two vasa deferentia were also present. In the left side a flattened sex-
gland (3 cm. high) was found, made up of testicular tubules. Two
small ova were found in its posterior end. The right gonad was also
tubular (testis).

The occurrence of real testicular tissue in one of the Holland birds
and in three others described by Boring and Pearl, as well as in one

RELATING TO SECONDARY SEXUAL CHARACTERS. 41

described by Pearl and Curtis, and in another by Shattuck and Selig-
mann calls for special comment, since the presence of both testicular
and ovarian tissue in the same bird is the essence of hermaphroditic a.
In general there are two ways of looking at such a result. Either the
sex-determining factors have been changed so that in one part of the
body, where the reproductive organs are laid down, one condition can
prevail, in other parts other conditions; or a mixup of the sex chromo-
somes has taken place. Until we get some more evidence concerning
such cases it is useless to speculate, although the former view might
seem the most probable of the two if the Holland birds of Herr Hou-
wink's flock were in a high degree true hermaphrodites.

But in fact three of the four described by Boring and Pearl were
due to tumors of the ovary, which, if they suppress the normal develop-
ment of this organ, would be expected to call forth the appearance of
the secondary sexual characters of the cock. If the likelihood of
developing a tumor were inherited, the frequent occurrence of hen-
feathered birds in this flock would be explained. However, one true
hermaphrodite in 4 birds is surprisingly high for a chance result, since
hermaphrodite birds are very rare.

The second interpretation suggested above is one that has been
advanced and established by genetic evidence in Drosophila, viz,
dislocation of the sex chromosomes. In the case of birds the male is
supposed to be duplex for the sex factors (ZZ), the female simplex
(ZW), and consequently the chromosome-dislocation hypothesis must
be worked out contrawise in birds and insects. We should have to
suppose that such birds start as males (ZZ), and that at some division
of the cells of the embroyo one of the Z's became lost (left at the cell-
wall for example). All the cells that got ZZ would be male ; all that got
Z would be female. If the reproductive region included cells of these
two kinds, an ovotestis would result. The rest of the body should be
the same, or nearly so, since the soma of male and female birds is alike
whether ZZ or Z, except in so far as it is affected by the secretions from
the ovaries (in most races of poultry), or from the testes if the race
be Sebright, Campines, or Hamburgs. Birds with ovotestis might,
nevertheless, be expected, on this view, to show at times an inter-
mediate condition of the secondary sexual characters, according to
how much internal secretion is produced in the ovotestis. In other
words, the chromosome loss might involve much more extensive
regions than the reproduction organs, but show its effects first in that
organ and then indirectly other parts of the body be affected by the
luteal cells of the testis. There is one rather good piece of evidence
that seems opposed to this interpretation. In the hermophroditea the
oviduct is present in all cases. Its conspicuous presence in the four
hermaphrodites would seem, therefore, to indicate that the birds

42 THE GENETIC AND THE OPERATIVE EVIDENCE

started as females (ZW), which is inconsistent with the dislocation
hypothesis. The alternate would be that in all these cases the Z part
always included the region of the oviduct, which seems improbable.

There is another possibility, viz, that in birds a sex-factor is carried
by the W chromosome, and ZW is a female not because of one Z, but
due to the presence of W. If so, then one Z or two Z's might give the
same result, viz, female. If a bird started as female, (ZW) and chromo-
somal dislocation occurred, then the Z parts would be female and the
male part W. Until we get evidence on this point it is not worth
elaborating. Without genetic evidence from hybrids, the interpretation
of hermaphrodites in birds can have at present only a speculative
interest. We may hope some day to get the same kind of evidence
as in the case of Drosophila. Hermophrodite hybrid pheasants that
have been often described might seem to furnish a hopeful field, for
they appear to be quite common and to show characteristics of both
races. As yet, however, no one has, I think, succeeded in finding a
simple interpretation of the results. It is also not unlikely that many
of the pheasant cases are not true hermaphrodites, but due to failure
of normal development of the reproductive gland, which gives an
intermediate or mixed type of secondary sexual characters.

RELATING TO SECONDARY SEXUAL CHARACTERS. 43

PART II.

DARWIN'S THEORY OF SEXUAL SELECTION.

Darwin seems to have felt the necessity of giving some other explana-
tion for the secondary sexual differences between the male and female
than that such differences were only a by-product or concomitant of
sex itself. His reason for searching further was probably a part of the
general point of view he had reached in regard to the utility of special
structures of animals, namely, that their presence finds its explanation
on the basis of utility. Believing as he did that most of the adaptat ions
of plants and animals have been built up by the accumulation of small
steps, it must have appeared to Darwin inconceivable that the highly
developed ornamentation exhibited in the secondary sexual characters
could have been simply the by-product of sex itself, especially when
the ornamentation may have been entirely absent in males of closely
related species. To-day we are not, I think, so oppressed with the
difficulties of the situation, for we have become familiar with the fact
that very slight genetic differences may cause very great differences
in the end-product. In a -word, the problem seems less formidable to
us than it did to Darwin.

Darwin appealed to three processes to account for the facts: (1) to
natural selection between the members of the same sex; (2) to choice
on the part of the "other" sex; (3) to the " inheritance of use." Since
each of these appeals to a different procedure, let us take them up
separately.

Competition of the males with each other for the female would,
Darwin said, lead to the survival of those males best endowed with
organs of offense and defense. The spurs of the cock are weapons dan-
gerous for other birds; the horns of the bull and those of deer are used
for offense and defense; the mane of the lion is a protection against the
teeth of other lions. It is true that these same weapons and shields
serve for attack and defense outside the species; but since the female
lacks them or has them less developed, they would not seem necessary
for survival of the individual against aggression from without. They
have developed, then, through competition within the specie-.

Several objections of greater or less weight have been urged against
Darwin's interpretation. It has been pointed out that the combats
within the species are seldom fatal and that the defeated rival finds
another mate. If, as a rule, there are as many females as males within
the species and monogamy is the rule, all males will find partners
sooner or later, all may have offspring, and the offspring have squally
good chances of survival. Under these circumstances it is not to be
expected that the combat would be likely to lead to the production of
males with longer spurs or larger horns.

44 THE GENETIC AND THE OPERATIVE EVIDENCE

Darwin realized this difficulty and tried to meet it by another
assumption, viz, that the better endowed males would also be more
likely to have more offspring. How could this be made probable?
Darwin suggested that the strongest males would be in position to
mate with the first females to reach maturity, and if these were more
likely to have offspring, either because of maternal endowments that
made them also more prolific or because the earlier broods would have
a better chance of getting food, etc., then the successful competitor
would sooner or later impress his advantages on the race.

At other times Darwin suggested that the exceptional vigor that
led to the greater development of the character in question would
itself be of value and through transmission to the offspring lead to
advance in the development of the other character in question. But
here the argument shifts to another field of inquiry and survival is
ascribed to greater vigor, while the secondary sexual character is carried
along in its wake as a sort of correlated effect.

It will be conceded, I think, that such pleading does not help the
argument, but exposes rather its inherent weaknesses. There is,
however, a line of defense that is permissible. If monogamy is not the
rule, if the male captures or attracts several females and keeps a harem,
as do the fur seals and walruses, or rules a herd as does the bull, or has
a flock as does the cock, or mates more frequently with random females
than do some other males, then the advantage of his more developed
weapon might lead to more offspring. If it could be shown that such
intraspecific weapons prevail more frequently within polygamous spe-
cies, a fair argument for natural selection might be made. I do not know
whether such a census has been taken as yet, but it is true, I think,
that in most polygamous groups we find weapons of offense very highly
developed. The fur seal has a harem and the male is greater in size,
in strength, and in the development of his tusks than is the female.
Similarly for the walrus. The bull drives away rival bulls from the
herd until through age or injury, or through the development of a
better fighter, he is replaced. If the better endowment is due to a
genetic factor, we should expect natural selection to keep the race at
the highest possible level that variation supplies material for. If, then,
we confine the application of natural selection to cases of this sort, the
explanation is as valid as is the theory in other fields. Such a con-
clusion becomes weakened when an attempt is made to apply it to
other groups of animals in which it appears improbable that the
secondary sexual characters of the male have any obvious value as
organs of offense. There are families of beetles, for example, in which
the development of the horns of the male are as striking as are those of
the ram or the stag. The males of these beetles are not known to fight
with each other, nor are they polygamous. It may seem that we must
look here for some other explanation, which, if found, might suffice to

RELATING TO SECONDARY SEXUAL CHARACTERS. 45

cover also the case of birds and mammals. In answer to this criti-
cism it may be argued that it is also possible that the other explana-
tion when found need not necessarily apply to the higher animals,
where the laws of combat may still give the true explanation. On the
whole, I think that, for our present purpose, it will suffice to state
it is consistent with the theory of natural selection to accept prorision-
ally this part of Darwin's theory for those species in the higher groups
in which polygamy holds, conceding, however, that even here it may
have to be altered when fuller knowledge is gained.

We are more concerned with that special feature of Darwin's theory
of sexual selection that is applied to those cases where the characters
are supposed to owe their special development to selection by the
individuals of the opposite sex. It is assumed that the female chooses
the better endowed males, because of the strong appeal he makes to
her sense-organs. Here we must employ perforce or for brevity's sake
the terms used in human psychology, and run the risk at every turn of
imputing to other animals the emotions and acquired associations
which man himself utilizes. Even granting that other animals pos-
sess somewhat similar emotions to ours, there still remains always the
danger, in the absence of real evidence, of imputing to them the par-
ticular emotion that we call "feeling for beauty"; and the greater
danger of imputing an esthetic sense so highly developed that the choice
falls in the long run on the suitor better ornamented than his rivals.

OTHER THEORIES TO ACCOUNT FOR SECONDARY SEXUAL

CHARACTERS.

Wallace has always been an opponent of Darwin's theory of sexual
selection in so far as it is based on female choice. As already stated,
he believes that the difference between the plumage of the male and
female in birds is due to natural selection keeping down the ornamen-
tation and high coloration in the female, because these would be
expected to expose the female while sitting on the nest to the attacks
of enemies, more especially of hawks. In support of this view he
points to a long series of species which build exposed nests and in them
the female is plainly and inconspicuously colored, while he also points
out that in such birds as parrots, toucans, woodpeckers, hangnests,
and starlings, which nest in holes or have covered nests, the female
is often as highly colored as the male. It can not be denied that
he makes out rather a strong case in support of this view, despite the
fact that there are other birds, like the Baltimore oriole, that have
covered nests and in which the sexes are very markedly different.

Wallace tries to meet cases like the last one by assuming that the
covering keeps off the rain ; but, if so, why are the sexes st ills., different ?
In the case of other highly colored birds, such as jays, magpies, hawks.
and crows, Wallace believes that these birds are all aggressive, hence

46 THE GENETIC AND THE OPERATIVE EVIDENCE

can protect their nests if attacked. As a further support of his
view, Wallace points out that in the few cases where the female is
more highly colored than the male (as the dotterel, species of phalarope,
an Australian creeper) the male incubates the eggs.

Wallace's suggestion still leaves unexplained the ornamentation of
the male, which he tries to account for as the direct result of the greater
vitality of the male. He tries to show that excessive ornaments and
high coloration develop especially in those parts of the body to which
there is an unusual supply of blood or where nerves and blood-vessels
emerge to go to the skin or to the muscles.

"If we have found a vera causa for the origin of ornamental appendages of
birds and other animals in a surplus of vital energy, leading to abnormal
growths in those parts of the integument where muscular and nervous action
are greatest, the continuous development of these appendages will result from
the ordinary action of natural selection in preserving the most healthy and
vigorous individuals, and the still further selective agency of sexual struggle
in giving to the very strongest and most energetic the parentage of the next
generation. And, as all the evidence goes to show that, so far as female birds
exercise any choice, it is for 'the most vigorous, defiant, and mettlesome male/
this form of sexual selection will act in the same direction, and help to carry
on the process of plume development to its culmination. That culmination
will be reached when the excessive length or abundance of the plumes begins
to be injurious to the bearer of them; and it may be this check to the further
lengthening of the peacock's train that has led to the broadening of the
feathers at the ends, and the consequent production of the magnificent eye-
spots which now form its crowning ornament.

"The display of these plumes will result from the same causes which led to
their production. Just in proportion as the feathers themselves increased in
length and abundance, the skin-muscles which serve to elevate them would
increase also; and the nervous development as well as the supply of blood to
these parts being at a maximum, the erection of the plumes would become a
habit at all periods of nervous or sexual excitement. The display of the
plumes, like the existence of the plumes themselves, would be the chief external
indication of the maturity and vigor of the male, and would, therefore, be
necessarily attractive to the female. We have, thus, no reason for imputing
to her any of those esthetic emotions which are excited in us, by the beauty of
form, color, and pattern of these plumes; or the still more improbable esthetic
tastes, which would cause her to choose her mate on account of minute differ-
ences in their forms, colors, or patterns."

Wallace says, referring to the immense tuft of golden plumage in
the best known birds of paradise (Paradisea apoda and P. minor) that
springs from a very small area on the side of the breast, that Mr. Frank
E. Beddard, who has kindly examined a specimen, says that "this
area lies upon the pectoral muscles, and near to the point where the
fibers of the muscle converge towards their attachment to the humerus.
The plumes arise, therefore, close to the most powerful muscle of the
body, and near to where the activities of that muscle would be at a
maximum. Furthermore, the area of attachment of the plumes is just

RELATING TO SECONDARY SEXUAL CHARACTERS. 47

above the point where the arteries and nerves for the supply of the
pectoral muscles, and neighboring regions, leave the interior of the
body. The area of attachment of the plume is, also, as you say in your
letter, just above the junction of the coracoid and sternum." " Orna-
mental plumes of considerable size rise from the same part in many
other species of paradise birds, sometimes extending laterally in front,
so as to form breast shields. They also occur in many hummingbird-,
and in some sun birds and honey-suckers; and in all these cases there
is a wonderful amount of activity and rapid movement, indicating
a surplus of vitality, which is able to manifest itself in the development
of these accessory plumes."1

There are two serious defects in such an attempt to explain the facts.
In the first place, it has been shown in several cases that have been
studied that it is not the lessened "vitality" of the female but thesuppres-
sion caused by the ovary that keeps down the development of the full
plumage in that sex. In the second place, the anatomical influences
appealed to are imaginary rather than real, for it is by no means
apparent that the local exits of blood-vessels and nerves to muscle-
are at all correlated with the location of the ornamental parts, in the
skin. Even when larger blood-vessels run to the region of excessive
development of feather ornaments it may well be that they go there
because the ornaments in question use them for their nourishment; in
other words, Wallace puts the cart before the horse. The top of the head .
where crests so often develop, the throat coloration and throat shields
of hummingbirds and birds of paradise, the two long tail feathers of
several species of hummingbirds, etc., do not arise, so far as known,
from regions which are conspicuous for a rich supply of blood and
nerves. Wallace's appeal to underlying organs such as muscles that
supposedly influence the special development of the feathers in the
skin above does not strike one as a fortunate appeal to physiological
principles.

Hudson, in his interesting book, "The Naturalist in La Plata," has
also criticized Darwin's theory of sexual selection. He has brought
together a considerable number of interesting observations that go to
show that the displays — dancing, singing, and combats — of males and
females have no relation to mating. Many of them involve birds
already mated, sometimes several males participating, sometimes
males and females together. Some of the tourneys he describes are
more elaborate than the mating instincts themselves, yet arc not con-
cerned with mating. He attempts to explain them as overflow phe-
nomena, i. e., as expressions of the high vitality of the males, especially
at this time. If he is right, then elaborate exhibitions of these kind-
have evolved that have no special connection with mating. Are we

1 For activity and pugnacity in hummincliinls, ne Tropical Nature, pp. ISC, MS.

48 THE GENETIC AND THE OPERATIVE EVIDENCE

called upon for a different explanation for other differences that dis-
tinguish the sexes? One example will suffice to bring out a curious
emotional (?) display that, elaborate as it is, has no apparent connec-
tion with mating (p. 269) :

"The lapwing display, called by the natives its 'dance' or 'serious dance' —
by which they mean square dance — requires three birds for its performance,
and is, so far as I know, unique in this respect. The birds are so fond of it
that they indulge in it all the year round, and at frequent intervals during the
day, also on moonlight nights. If a person watches any two birds for some
time — for they live in pairs — he will see another lapwing, one of a neighboring
couple, rise up and fly to them, leaving his own mate to guard their chosen
ground ; and instead of resenting this visit as an unwarranted intrusion on their
domain, as they would certainly resent the approach of almost any other bird,
they welcome it with notes and signs of pleasure. Advancing to the visitor,
they place themselves behind it; then all three, keeping step, begin a rapid
march, uttering resonant drumming notes in time with their movements ; the
notes of the pair behind being emitted in a stream, like a drumroll, while the
leader utters loud single notes at regular intervals. The march ceases; the
leader elevates his wings and stands erect and motionless, still uttering loud
notes; while the other two, with puffed-out plumage and standing exactly
abreast, stoop forward and downward until the tips of their beaks touch the
ground, and sinking their rythmical voices to a murmur remain for some time
in this posture. The performance is then over and the visitor goes back to
his own ground and mate, to receive a visitor himself later on."1

Cunningham, who has brought together many interesting cases of
secondary sexual differences in his book on "Sexual Dimorphism in the
Animal Kingdom," attempts to show that the development of the
secondary sexual characters of the males are due directly to the use of
certain parts of the body during courtship — the use of the parts leading
to the enlargement and excessive growth of the parts. The effects
are believed by him to be inherited, and he tries, furthermore, to show the
way in which such acquired characters could be inherited. He makes
use of the modern idea of hormones — substances that are elaborated
in many organs of the body, whose effects are often most conspicuously
produced in other parts of the body. He imagines these hormones to
be collected in the germ-cells and transmitted to the next generation,
where their presence contributes to the further development of the
special region (when it develops) that corresponds to the region in
its parent in which the hormone was made. His speculation meets
in the first place with the general objections inherent in Lamarck's
theory — objections so well recognized to-day that I need not go over
them here. His special appeal to the hormone theory makes use of that
theory in a way to which it was never intended to be put, by assuming
that an internal secretion formed in one organ can be stored up in
another organ, eggs and sperm — an assumption not only unsupported
by any evidence, but, as I have stated, one quite foreign to the hor-

1 The Naturalist in La Plata, W. H. Hudson, London, 1892, pp. 269-270.

RELATING TO SECONDARY SEXUAL CHARACTERS. 49

mone theory. In fact, Cunningham's suggestion is nothing more than
Darwin's old idea of pangens, which, being imaginary, could be en-
dowed with all desirable properties. But one can not invoke a chemi-
cal substance, even a hormone, and then at the critical moment endow
it with special virtues.

A rather unique explanation of the origin of secondary sexual charac-
ters is made by Stolzmann. His argument runs as follows: (1) There is
a great excess of males in birds; (2) the males left over after mating
are useless to the species, since they can not propagate and they con-
sume food needed by the reproducing part of the population ; (3) the
conspicuous coloration of the male has been evolved in order that he
could be seen more readily by birds of prey and the objectionable
excess of males removed; the comb of the cock has developed in order
that he may be the more easily killed by other cocks.

Stolzmann's account of the origin of the plumes of the birds of
paradise should be immortalized in the literature of the subject:

"Nous comprendrons aussi facilement la presence de longues plumos chez
les males de nombreuses especes, comme p. e. chez les oiseaux de paradis, chez
les veuves (Vidua) et chez l'engoulevent africain (Cosmetornis). Telles
plumes ont probablement pour but de relantir le vol des males. J'ai constat*-
chez la Loddigesia mirabilis (oiseaumouche pe>uvien), que le view male
poss£de l'aile quelques millimetres plus courte que le jeune male ou la femelle.
Cet avortement des remiges provient assurement a cause de deVeloppement
extraordinaire de retrices externes chez le vieux male de cet oiseaumouche.
Si done d'une part les retrices allongees rendent le vol plus difficile et d'hautre
les ailes plus petites diminuent sa v61ocit6, le vol du male doit etre plus lent
que celui de la femelle, le poids du corps restant la meme. Le develop-
pement extraordinaire soit des remiges soit des rectrices, en relantissant le
vol des males, rend leur role plus difficile, en facilitant en meme temps celui
des femelles. Nous pouvons prendre comme exemple le Cosmetornis, qui,
comme tous les engoulevents, se nourrit d'insectes, qu'il attrape au vol. ( hez
cet oiseau quelques plumes des ailes se developpent extraordinairement pendant
l'epoque de reproduction, en retardant visiblement son vol. II est done facile
a remarquer, qu'alors le male, ayant les mouvements plus lounls, n'est pas en
etat de se procurer la meme quantity d'insectes qu'auparavant ; ainsi done la
femelle a plus de chances de trouver une nourriture plus ahondante." l

Equally worthy of perpetuation is Stolzmann's explanation of
dancing and singing birds :

"Toutes les reunions des males, leurs danses bizarres, leur chant, enfin, ne
servent pas probablement a s6duire les femelles, mais pour distrain" les males,
ce qui rend plus faciles les besognes maternelles des femelles et au surplus les
protege contre l'assiduite nuisible des c61ibataires. Darwin lui-meme con-
state le fait, qu'ordinairement pendant les reunions des males, quand
derniers sont trop occupes par le combat ou la danse. la femelle B'echappe awe
un d'eux pour copuler. Ainsi done dans ce cas e'est bien la selection D&turelle
et non la selection sexuelle, qui agit pour la conservation d'equilibn sexuel." ■

Proceedings of the Zoological Society of London, 1885, p. 431, QMJqUM IwaQMI sur le
dimorphisme sexuel. Jean Stolzmann.

50 THE GENETIC AND THE OPERATIVE EVIDENCE

DISPLAY OF THE MALE.

The antics of male birds at the mating season, their courtship
so-called, has played an important role in Darwin's theory of sexual
selection. The behavior of many birds at this time is of such a kind as
to suggest that the male is exhibiting his plumage before the female
for the "purpose" of influencing her choice. The whole parapher-
nalia of human psychology is imported into the situation and both the
consciousness of the male, his intentions so to speak, and the supposed
esthetic response or choice of the female is invoked. Even though it
be granted that the words that we must make use of, borrowed from
human behavior, are such as to imply much more in the direction of
consciousness and purpose than is desirable, and that most of the
behavior of animals should be stated in a more roundabout and
objective way, yet the theory will only work out on the assumption
that the female chooses in some sense the more brilliant or ornamental
(or effective) male, whether she is "conscious" or unconscious of
intention. I doubt if anyone to-day would care to defend seriously
the theory on the grounds of consciousness or esthetic value of the
exhibition, despite the fact that Darwin's language often takes this
turn and the less-guarded statements of some of his disciples, such as
Romanes, show little hesitation in anthropo-morphologizing the entire
situation. It is, however, not necessary for the working out of the
theory that this complication be introduced into it, for if the female
is more likely to mate with a more brilliantly colored than a less
brilliantly colored male, the theory may be made to apply regardless
of whether she is "conscious" or not of the difference to which she
responds.

But there are weighty arguments against such an interpretation of
the behavior of the male and female during courtship. In the first
place, there is almost no direct evidence to show that the female mates
with the more ornamental male. As this is the all-essential require-
ment of the theory, the almost complete absence of facts in its support
leaves the theory resting on a theoretical assumption. It can scarcely
pass unnoticed that while there exists a large mass of data describing
the secondary sexual characters, there is practically nothing in this
accumulation to show that the female makes her selection on differ-
ences in coloration or ornamentation. And on the other hand, there is
some evidence showing that the female is ready to succumb to the
aggressiveness of the male rather than that she "chooses" him.

The behavior of the male under sexual excitement is often described
to be of a kind to exhibit before the female his peculiar adornments.
That the "purpose" of his exhibition is to show himself off before the
female may be conceded, with reservations as to what is meant here
by "purpose." That the male is conscious of the probable results of his
conduct is scarcely probable the first time he courted ; but that he may

RELATING TO SECONDARY SEXUAL CHARACTERS. 51

have found out the most probable result after the first attempt through
"associative memory" is in accord with what the study of "animal
behavior" has shown to be possible. In this sense purpose would mean
a line of conduct that experience had shown to lead to a certain and.
Anticipation or far-sightedness would henceforth characterize such a
reaction. Here, however, we venture on very dubious grounds. Hut
the display of the male may be purposeful in a much simpler sen.-*'.
His activity may be an inborn reflex to visual or other sensory stimuli
that is a part of his attack on the female, or possibly a series of reflexes
that we may register under the old unanalyzed terms of "desire and
fear." The action calls forth a responsive reflex in the female, for
the sexual act is not entirely active on one side, passive on the other,
but consists of a series of interreactions on the part of each sex, which,
if they pursue a given course, leads to the final mating. The mutual
responses appear to follow an automatic course in many cases if the
individuals are sexually ready to mate and the environment is pro-
pitious. Types of behavior of this kind must be familiar to anyone
who has observed domesticated and semi-domesticated animals. The
purpose of the display may mean no more than a reaction that leads
to a result propitious to the perpetuation of the species if the situation
is ripe for such an outcome.

This conclusion still leaves open the question as to whether the
display is more likely to be successful, if certain special characters
possessed by the species are exhibited. In the absence of any sufficient
evidence to show that this is so, and in the light of the very great danger
of projecting "our human standards" into the world of other animals,
and in view of the fact that related species without such marks are as
successful in maintaining themselves, I can not but think that at
present we have a good deal to lose in the way of scientific procedure
and nothing to gain of scientific value in accepting Darwin's inter-
pretation of sexual selection based on the display of the male as fur-
nishing an opportunity to the female to make the "best" selection
amongst her suitors on the basis of his adornment.

An excellent opportunity to study the problem as to "choosing"
by the female is furnished by the mutant races of Drosophila. BOOM
of which, differing in a single mutant gene, have wings as different in
coloration as black, yellow, or gray, and eyes as differently colored as
white, vermilion, or red. Sturtevant put a yellow female with a gray
(wild-type) male and a yellow male. The male that Inst mated was
noted and the trio discarded. The female "chose" the gray malec
times and the yellow only 8 times. In the control combination, when' a
gray female "chose" between the same two kinds of males, she took
the gray male 60 times and the yellow male 12 times. In both cases it
"appears" that the female " prefers " the gray male, but this deduction
may give an entirely wrong impression as to what is taking place, for

52 THE GENETIC AND THE OPERATIVE EVIDENCE

the result would be the same in kind if the gray male were more active
and mated quicker. This was tested by putting a gray and a yellow
female with a gray male and then for control a gray and a yellow female
with a yellow male. The result was as follows :

RedcfJGray9 25 y u (Gray 9 12

Ked & \Yellow 9 31 le OW ^ \Yellow 9 30

Here the gray male mated slightly oftener with the yellow female
than with the other, whereas the yellow male mated much oftener with
the yellow female than with the gray one. Both results are explicable
on the view that the yellow female, being less active, is more easily
captured by the yellow male than is the gray female. This view fits
in also with the former experiment, where the yellow male is much less
successful than the more active gray male. Such a conclusion gives a
more consistent explanation of all the facts than does the theory of
female choice, for on the latter we must suppose that the yellow
females prefer the gray males and the yellow male prefers the yellow
females, etc.

The following results were obtained by Sturtevant when red and
white eyed flies were competing :

„ , , /Red 9 54 p . „ /Red 61 53

Redc? \White 9 82 Red$ \Whitec? 14

wuu ^/Red9 40 w,;. 0 /Red c? 62

White c? (white ? Q3 Wlute ? \White cf 19

The outcome can be interpreted in the same way as the yellow-gray
competition. The red male wins by virtue of his greater activity, while
the white female is chosen more often, especially by the white male,
because of her passivity (or weaker resistance) . It may be claimed that
these results do not show that the female does not choose, for such
choice, if made, would be swamped by another condition of the experi-
ment, viz, the greater aggressiveness of one kind of male and greater
passivity of the other kind of female. This, of course, is true, but the
experiment still shows that in these flies other influences are so much
greater than "choice" by the female, if it exists, that the postulated
effect of the latter practically disappears from the situation.

Mayer's experiments with the large moth Callosamia promethea
furnish important information as to the factors involved in mating.
The results are all the more significant from our present point of view
because the colors of male and female are in this species markedly
different. The wings of the male are black, those of the female reddish
brown; the antennae of the male are large and bushy, those of the
female small and slender. Mayer found that the males are attracted
by the female from some distance when the latter are put into a glass
jar covered by only coarse mosquito-netting, but if the same jars are
turned upside down the males are unable to find the female. Females

RELATING TO SECONDARY SEXUAL CHARACTERS. 53

concealed in loose cotton attracted males. Females were put into a
box with an open chimney at one end, the other open end being covered
by mosquito-netting. A current of air blew into the open end and out of
the chimney. The males flew to the end of the chimney from which
the air came and fluttered about in the neighborhood. Males are
attracted to places where a female has been kept even several hours
after her removal. The male finds the female through the sense-organs
in his antenna?, for a male whose abdomen has been cut off and the sides
of whose thorax are covered with shellac will still fly to the female,
but if his antennae be coated with any substance he no longer seeks the
female. If the eyes of the males are blackened they will mate with
females "in the normal manner."

Mayer cut off the wings of females and glued male wings in their
places, so that the female looked like a male. Males readily mated
with these females. The wings of males were cut off and female wings
glued in their place. Mating occurred "with normal frequency, and
I was unable to detect that the female displayed any unusual aver-
sion" to such males. Males with female wings pass unnoticed by other
males.

In a later paper (1901) Mayer and Soule describe how, when the wings
of the male were painted scarlet or green, the males were accepted
as readily as normals in competition with them. Experiments were ata i
made by them with the gipsy moth. Wingless males met with more
"resistance" from the female than do normal males, but when the eyee
were covered the wingless males succeeded as often as the normal
males, but the number of observations on which this statement is
were far too few to be of any value, and there are several other obser-
vations that make any such conclusion from the evidence highly
uncertain.

That it is the odor of the females that attracts the male can not be
doubted. It might still be claimed that the female chooses amongst
her suitors the darkest males, but the evidence gives no grounds for
inferring such a choice, and since she will even accept males with
female wings when they attempt to mate with her, it does not appar
probable that the color of the male is a factor in the result, or at least
if it is, then it must be entirely subordinate to the sense of smell in
finding the female and of touch after he arrives. These is little or
nothing in the behavior of these moths, or in that of the silkworm
moth, according to Kellogg, to suggest that vision plays any significant
role in courtship.

Concerning the genetic situation in insects, there are only a few casrs
that have been studied. The most instructive are those in which more
than a single kind of male exists (two or three), one of which may b*
like the female, the other quite different. The best worked out cases
are Papilio memnon and P. polytes. De Meijere and Punnett have

54 THE GENETIC AND THE OPERATIVE EVIDENCE

shown from the breeding data that it is possible to frame an explanation
of such a sort that the aberrant female differs from the female resem-
bling the male in only a single genetic factor — one not sex-linked
(i. e., not carried by an X chromosome), but autosomal. The gene
would be of such a sort that it affects the female only — producing no
visible effect on the male. Such a conclusion, if established, helps,
theoretically at least, toward simplifying the situation in other species,
for it shows that genetic factors occur whose influence is on one sex
alone; hence the difference between the male and one type of female
does in such cases result from a single gene present in both but causing
them to be differently colored. There would be no need, then, to assume
that the difference had been slowly built up by selection, but rather
that the difference arose at some time by a single mutant step. The
incorporation of the step in the species would then follow if the effect
of the gene were useful in mating or if it had some other primary
significance for the welfare of the species, the different effect produced
on the male and female being only an unimportant by-product of its
action. On the other hand, it should be emphasized that because a
single factor difference between the two kinds of females will explain the
genetic results, it does not necessarily follow that the difference did
arise as a single mutation. The foregoing argument does no more
than imply that the difference in question may have arisen in this way,
and if so, that the situation, as it exists, would be the more easily
comprehended.

In insects and spiders, where dimorphism is as marked as in birds,
the mating habits have been studied by a number of naturalists. Here
alou there are nuinexuus accounts of the display of the male during
courtship. The account given by Dr. and Mrs. Peckham are particu-
larly detailed and call for careful consideration on account of their
well-recognized accuracy in observational work. Moreover, as a
result of their observations, along with those of Montgomery, Petrunke-
witsch, and others, we have really fuller information concerning the
courtship of spiders than of birds and of mammals.

In the great majority of species where the sexes are different the male
is more brightly colored or more ornamental. For example, in a group
such as the Attidae of France, where both sexes are known, the Peck-
hams state that in 26 cases the male is more conspicuous than the
female; in 55 cases the sexes are alike, or if they differ the male is more
conspicuous. It appears that in other genera there are cases where the
female is more conspicuous than the male. The Peckhams state that
possibly as many as 250 species are in this condition. Those females
with brighter colors than the males are usually well armed by strong
spines. When very young they are like the males and begin to assume
the adult form and color when they are a quarter to a third grown.
Whether the change depends on changes in the ovary is not known.

RELATING TO SECONDARY SEXUAL CHARACTERS. ">.">

The mating behavior of Saitispulex,a, species in which the males
and females are much alike, is described by the Peckhams as follows:

"On May 24th we found a mature female and placed her in one of the larger
boxes, and the next day we put a male in with her. He saw her as she stood
perfectly still, twelve inches away; the glance seemed to excite him and lie at
once moved toward her; when some four inches from her he stood still and
then began the most remarkable performances that an amorous male could
offer to an admiring female. She eyed him eagerly, changing her position
from time to time so that he might be always in view. He, raising his whole
body on one side by straightening out the legs, and lowering it on the other by
folding the first two pairs of legs up and under, leaned so far over as to be in
danger of losing his balance, which he only maintained by sidling rapidly
toward the lowered side. The palpus, too, on this side was turned back to
correspond to the direction of the legs nearest it. (Fig. 13.) He moved in a
semi-circle for about two inches and then instantly reversed the position of the
legs and circled in the opposite direction, gradually approaching nearer ami
nearer to the female. Now she dashes toward him, while he, raising his first
pair of legs, extends them upward and forward as if to hold her off, but withal
slowly retreats. Again and again he circles from side to side, she gazing
toward him in a softer mood, evidently admiring the grace of his antics. This
is repeated until we have counted 111 circles made by the ardent little male.
Now he approaches nearer and nearer and when almost within reach whirls
madly around and around her, she joining and whirling with him in a giddy
maze. Again he falls back and resumes his semi-circular motions, with his
body tilted over; she, all excitement, lowers her head and raises her body ><>
that it is almost vertical; both draw nearer; she moves slowly under him, he
crawling over her head, and the mating is accomplished.

"After they have paired once, the preliminary courtship is not so long.
When this same pair mated a second time, there was no whirling movement,
nor did the female lift her body, as at first." (pp. 37-38). l

The courtship of another species, Dendryphantes capitatus, in which
the sexes are entirely different, is described as follows :

"The males of capitatus are very quarrelsome, sparring whenever they meet,
chasing each other about, and sometimes clinching. It is a very abundant
spider with us, so that we often put eight or ten males into a box to sec them
fight. It seemed cruel sport at first, but it was soon apparent that they wen-
very prudent little fellows, and were fully conscious that 'he who tights and
runs away will live to fight another day.' In fact, after two weeks of hard
fighting we were unable to discover one wounded warrior. When the males
are approaching each other, they hold the first legs up in a vertical direction.
Sometimes they drop the body on to one side as they jump about each other.
These movement are very quick, and they are always ready for a passage at
arms. When courting the females they have another movement. They
approach her rapidly until within two to five inches, when they stop and extend
the first legs directly forward, close to the ground, the legs being slightly
curved with the tips turned up. (Fig. 18). Whether it be intentional or not .
this position serves admirably to expose the whole of the bronze and white
face to the attentive female, who watches him closely from a little distance
(Fig. 19.) The males also give their palpi a circular movement, much as a
person does when washing his hands. As he grows more excited, he lies down

1 George W. and Elizabeth G. Pcckham. Observations on Sexual Si-lection in Bpfckn of the
Family Attidffi. Nat. Hist. Soc. of Wisconsin, Vol. I, 1KS9, pp. 46, 47.

56 THE GENETIC AND THE OPERATIVE EVIDENCE

on one side with his legs still extended. These antics are repeated for a very-
long time, often for hours, when at last the female, either won by his beauty or
worn out by his persistence, accepts his addresses." (Pp. 45, 46.)

In another species, Dendryphantes elegans, both sexes are brilliantly
colored.

"The male is covered with iridescent scales, his general color being green;
in the female the coloring is dark, but iridescent, and in certain lights has lovely
rosy tints. In the sunlight both shine with the metallic splendor of humming-
birds. The male alone has a superciliary fringe of hairs on either side of his
head, his first legs being also longer and more adorned than those of his mate.
The female is much larger, and her loveliness is accompanied by an extreme
irritability of temper which the male seems to regard as a constant menace to
his safety, but his eagerness being great, and his manners devoted and tender,
he gradually overcomes her opposition. Her change of mood is only brought
about after much patient courting on his part. While from three to five
inches distant from her he begins to wave his plumy first legs in a way that
reminds one of a wind-mill. She eyes him fiercely and he keeps at a proper
distance for a long time. If he comes close she dashes at him and he quickly
retreats. Sometimes he becomes bolder and when within an inch, pauses,
with the first legs outstretched before him, not raised as is common in other
species; the palpi also are held stiffly out in front with the points together.
Again she drives him off, and so the play continues. Now the male grows
excited as he approaches her, and while still several inches away whirls com-
pletely around and around; pausing, he runs closer and begins to make his
abdomen quiver as he stands on tip-toe in front of her. Prancing from side to
side, he grows bolder and bolder, while she seems less fierce, and yielding to the
excitement lifts up her magnificently iridescent abdomen, holding it at one
time vertically and at another sideways to him. She no longer rushes at him,
but retreats a little as he approaches. At last he comes close to her, lying flat,
with his first legs stretched out and quivering. With the tips of his front legs
he gently pats her; this seems to arouse the old demon of resistance, and she
drives him back. Again and again he pats her with a caressing movement,
gradually creeping nearer and nearer, which she now permits without resis-
tance until he crawls over her head to her abdomen, far enough to reach the
epigynum with his palpus". (Pp. 46-47.)

If we lay no emphasis on the implied emotional elements in the
behavior of the spiders in this description — terms of emotion borrowed
direct from human psychology — there still remain the several types
of apparently significant reactions associated with courtship. The
statements leave no room for doubt that vision plays an important
role in the complex reflexes that lead gradually to successful mating.
The Peckhams insist that the display of the male is always of a kind
to bring before the female the special adornments of the male in what-
ever part of the body they may he. The chance of subjective inter-
pretation here is so great that unless the results are carefully checked
up by studies of the attitudes assumed by males in species in which
the males are without ornament, their interpretation must be taken
with the greatest reserve. Assigning, as our authors do, so much by
gratuitous implication to the emotional side of the picture prejudices

RELATING TO SECONDARY SEXUAL CHARACTERS. 57

one, perhaps too greatly, against accepting a special (even an implied
intentional) exhibition of the specially ornamented parts. On the
other hand, if it be conceded that the conspicuousness of the male is
an element in the reaction, the very special adornments visible from
the front might be supposed to enhance the effect produced in the
female. Similar displays of special ornamentation in the male have
been described both for birds and insects, but here, too, the question
has been raised as to whether such exhibitions are more than an
accidental accompaniment of the posturing of the male, for the same
kind of behavior is known to occur in other cases where the male is
unornamented and resembles the female. Had such a male special
ornamentation it would no doubt appear to us that his behavior was
"calculated" to display his ornaments.

Dr. and Mrs. Peckham point out that their observations are entirely
inconsistent with Wallace's interpretation of the origin of secondary
sexual characters. They find no evidence in favor of his view that the
male possesses greater "vital activity." On the contrary, the female
is the more active and pugnacious of the two. They also object to
Wallace's statement of a total absence of any evidence that the female
notices the display of the male. In spiders the females "observe" the
males with close attention during their courtship. They point out also
that, in spiders at least, as the female gradually becomes adult, a male
if preferred will have a chance of mating with several females, "and
as the mating season lasts for two or three weeks the more brilliant
males may easily be selected again and again." In regard to Wallace's
argument as to the distribution of accessory plumes in humming birds,
the Peckhams point out that —

"The pectoral muscles reach their highest development in the humming-
birds, the diurnal birds of prey, and the swallows, and wo may, therefore,
fairly use these groups to test Mr. Wallace's explanation of breast plumes. In
the swallows and birds of prey we find no such appendages, in spite of their
further claim to them, on the ground of great vigor and activity. As to the
humming-birds, we find in the genus Aglceactis six species with more or less
developed breast-plumes, which are also found in nine other species, scattered
through different genera — in all, only fifteen species out of four hundred and
twenty-six; while we find in fifty-six species the lengthened and modified tail-
feathers, which, according to Mr. Wallace's view, should be peculiar to the
Gallinacese.

"Again, there are elongated feathers from the throat or from the side of the
neck in thirty-five species, while seventeen have crests from the top of the
head, and seventeen, downy puffs from the tarsi." *

From this brief survey of the family we see that, contrary to what
we should expect from Mr. Wallace's theory, although the breast
muscles are the seat of the highest activity, breast plumes are the least
frequent of all the forms of ornamental plumage.

1 Loddigesia mirabilis has the tail about three times as long as the body. Similar modifiwitiom
are found in the genera Sappho, Cynanthus, Lesbia, Stegnura, Di.icura. OowWi'a, ct al.

58 THE GENETIC AND THE OPERATIVE EVIDENCE

"We may fairly say, then, that the humming-birds completely refute the
proposition that there is any relation between the development of color and
accessory plumes and 'surfaces where muscular and nervous development is
considerable.' m

What is true for birds is even more obvious for spiders where the
special ornaments are not confined to parts of the body with high
muscular development, etc. The writers make the very pertinent
criticism that while Wallace objects to assuming the emotional states
in females, he is less careful in regard to the males' emotions when he
speaks of the display "under the influence of jealousy or sexual excite-
ment The males, in their rivalry with each other, would see

what plumes were most effective; and each would endeavor to excel his
enemy as far as voluntary exertion would enable him."2

"If the males have so complex an emotion as jealousy, and further, if they
are conscious of the value of the plumes, may it not be asked why the female is
unable to 'see what plumes are most effective?' The mental state in the male
is without meaning unless we suppose the female to be affected and pleased."
(Peckham, loc. cit., p. 144.)

In regard to another interpretation of the courtship, the Peckhams
point out:

"Mr. Pocock has suggested that the attitude of observant interest on the
part of the female spider might be taken to indicate that she was preparing to
spring upon her mate and devour him; or that it might simply mean that she
was warily guarding herself from his approach. Neither of these suppositions
is admissible. In some species the male is not attacked by the female, and
when she does wish, as frequently happens, either to avoid or to destroy him,
her attitude is totally different. In the former case she turns about and runs
rapidly away, or suspends herself by a thread of web. In the second, there is a
contraction of all the muscles, the legs are drawn together, and in this crouch-
ing position she creeps slowly toward him, as she might if he were a fly, only
with something more malignant in her aspect. When she takes this stand the
male incontinently flees. When, on the contrary, the female is interested in
the male display, she seems perfectly absorbed in watching him, the muscles
are all relaxed, unconscious of herself she directs her glance now here, now
there, as he moves about; as he continues his mad antics, her appearance gives
every indication of pleasurable excitement, and as he comes closer and closer,
she yields herself to the impulses which he has awakened in her, and, as in
pulex, joins in his dance and whirls around and around as though intoxicated.
We claim, then, to have completely answered Mr. Wallace's first objection."
(Peckham, loc. cit., pp. 145, 146.)

Finally, in regard to the specific character of the display of the
males, the Peckhams make the following significant statement:

1 Among the most remarkable of this wonderful family are the nine species of coquettes (Loph-
ornis), which have elongated feathers, with metallic tips, springing from the sides of the neck;
some have also beautiful crests. (George W. and Elizabeth G. Peckham., Additional Observa-
tions on Sexual Selection in Spiders of the Family Attidae., Nat. Hist. Soc. of Wisconsin, 1889,
vol. I, pp. 141, 142.)

2 Tropical Nature, p. 210. The italics are ours.

RELATING TO SECONDARY SEXUAL CHARACTERS. 59

"The spider has four pairs of legs, and all are equally available for display
or locomotion, and since all the movements are slow and on the ground they
are entirely open to observation and study, and we arc thus in a portion to
decide by facts whether their activity is simply an outlet for superfluous
energy, and therefore meaningless, or whether there is a purpose in it. If the
purpose of the antics is only to let off energy, then wo should expect one- pail
to be flourished around quite as often as another, and that the pair flourished
should as frequently be one that was not ornamented as one that was; and,
moreover, their movements ought not to be of such a nature as to display the
color or ornament more frequently than the law of chance would explain. If
the spider almost always moves the ornamented legs, and in such a way, too,
as to bring out their beauty, it would seem to us, to say the least, highly
improbable that the dance of the spider was merely a meaningless overflow of
surplus energy. Such an explanation leaves much that needs explanation.
The facts are, that the best foot is put forward; and this is just what Darwin's
theory requires and explains. Under Mr. Wallace's view the facts are inex-
plicable. The better to show that these movements are not simply meaning-
less outlets of high vigor, we illustrate the several positions by figures taken
from nature (figs. 7-12). The figures would seem to prove that the legs that
are ornamented or contrasted in color are also the legs that are usually flour-
ished; that where none of the legs have special ornament, then all are used; or,
as sometimes happens, when an unornamented leg is used the movements are
of such a character as to display some ornament that would otherwise have
been more or less hidden from the female." (Peckham, loc. cit., p. 147.)

In the tarantula, Petrunkewitsch finds that sight plays no role in
mating — that it is due entirely to accidental contact between the male
and female. Here the sexes are closely alike, except for a pair of hooks
on the front legs of the male, by means of which he grasps the mandibles
of the female, holding them during the elaborate process of trans-
ference to her genital opening the sperm that he has already collected
in the genital spoon on his palpi. The hooks serve to guard the
male against injury or death, wrhile at the same time they aid him in
the act of coitus.

In a common spider, Mceiia villata, two kinds of males exist. Both
have been seen to mate with the same female. No preference is given
to either type. The difference between them, according to Painter,
is connected with or caused by an additional pair of chromosomes in
the gray male. The two types may therefore have no connection with
sexual selection, but be directly due to a difference in the chromosome
group.

Montgomery, who made observations on the courting habits of
several species of spiders, states that his ''general theoretical con-
clusions were quite different from those of the Peckhams." It turns
out, however, that his objection to their view is based entirely on their
assumption that the male is conscious of his display and that the
female is guided by an esthetic sense in selecting the more beautiful
male. It should be pointed out that even after the removal of tl

60 THE GENETIC AND THE OPERATIVE EVIDENCE

gratuitous assumptions as to the cause of the evolution of the male
and female, enough still remains in Montgomery's own observations
to include his results on courtship under Darwin's theory of sexual
selection. For example, Montgomery says:

"The adult male is excited simultaneously by fear of and desire for the
female, and his courtship motions are for the most part exaggerations of ordi-
nary motions of fear and timidity. By such motions he advertises himself to
the female as a male, but there is no proof that he consciously seeks to arouse
her eagerness by esthetic display — there seems to be no good reason to hold

that the female is actuated in her choice by sensations of beauty

Thus my opinion was opposed to Darwin's theory."

Now, it is obvious that if a more brightly colored male has a better
chance of " advertising himself" to the female all the essential require-
ments of Darwin's theory are fulfilled, regardless of whether the male
is conscious of his ornamentation or the female makes use of an " esthe-
tic sense." In another passage (p. 173) Montgomery concedes all
that any modern critical advocate of Darwin's theory is likely to ask:

"We have previously seen that conscious aesthetic choice by the female
probably does not account for such male characters [secondary sexual charac-
ters with their 'conspicuous color markings']; that they are accordingly,
probably not due to sexual selection. These characters of the males may be
most readily explained as being conceived by simple natural selection. Pecu-
liar ornamentation would be selected because unusually greater sex recogni-
tion therefore prompted mating."

It is evident that Montgomery has only shifted the situation, although
to advantage, I think, but is essentially in accord with Darwin's theory
of sexual selection, despite his protest to the contrary. The difference
lies in Darwin's and especially in the Peckhams' use of the term
" choice," "aesthetic sense," etc., to stand for the fact that the female
more promptly mates (as Montgomery prefers to put it) with a male
peculiarly ornamental.

The most critical observations on sexual selection that have been
made in the group of insects are those by Sturtevant on the pomace
fly. The courtship is described as follows :

"The first and most noticeable act in courtship occurs when the male, being
near the female, extends one wing at about right angles to his body, and
vibrates it for a few seconds. The wing is then returned to the normal position
and the process is repeated, usually with the other wing. But between times
there is a scissors-like movement of the wings repeated several times. This
vibrating of the wings is often repeated many times, and may be done in any
position relative to the female, though the male always faces her. Usually,
in fact, he swings quickly around her in a semicircle once, or oftener, during
the process. Soon the male begins to protrude his genitalia and, if the female
remains quiet, to lick her posterior end. Some white matter now protrudes
from her ovipositor, and other males in the same vial are usually observed to
become excited now and begin courting, indicating odor as a cause of sexual
excitement. If the female runs or flies away the male is excited, moves his

RELATING TO SECONDARY SEXUAL CHARACTERS. 61

wings jerkily, and walks around rapidly, but seems unable to follow the female
accurately or to locate her quickly. The penis is directed forward l>y bending
up the abdomen underneath, towards the thorax, and is jerked toward the
female (the male always standing facing her at this stage), but not always
toward her genitalia, as I have seen it strike her in the eye. (The male in
this case, however, had white eyes, and so was perhaps blind. Normally the
aim is accurate.) If it does strike the mark the male mounts on the female's
back, between her wings. Mounting never takes place until after the actual
copulation has occurred, in which respect Drosnphila differs from some related
flies (e. g., Muscidse, Anthomyidae, Sepsidaj, Borboridae, and Ephydridse, so far
as my observations go). In these forms the male flies and lights on the female,
after which copulation may or may not take place, probably depending upon
the way the female responds." !

To test whether the wings have any significance in courtship, the
wings of a male were clipped off and he was put into competition with
a normal male of the same stock, age, and size. A virgin female
sexually mature was given to these two males. The normal male
mated 72 times before the other, the clipped male 53 times. It might
appear that the female selected the normal male in preference to the
clipped one, or possibly that the male with normal wings drove the
other male away. That the operation on the wings may have an
influence on the male himself is shown in McEwen's results. He found
that clipped males lost their heliotropism. It was also possible that
the courtship of the normal male might make the female ready to
copulate and then she would mate with either male. Sturtevant
tested the last supposition by placing single pairs in vials, testing each
day an equal number of normal and clipped males. The length of
time before copulation was noted. The clipped male began to court
as soon as the normal, but a larger number of normal males mated in
the first 12 minutes than clipped males (59 to 25). Had the females
discriminated against the clipped males to an equal extent we would
have expected a much greater excess than 72 to 53 when they were in
competition. It appears, then, that the wings are useful in shortening
the time between the meeting of the individuals and copulation. The
display acts, however, almost as favorably for the other male as for
the exhibitor himself. The results showr, therefore, that here an est hetic
sense of the female need not be postulated, for she actually shows little
preference when she has been brought to the point of mating between
the male that aroused her and the other male that did not. This
critical test puts the problem in a different relation from that which
Darwin's theory of female choice was meant to throw light upon.

The reverse experiment — a clipped and a normal female of the same
age, size, etc. — showed that the male did not discriminate between
them, for in 52 first trials the normal female was paired with 25 tunes,
the clipped 27 times.

'A. H. Sturtevant. Experiments on Sex Recognition and (he Problem of Santa! Bateetta in
Drosophila. Journ. Animal Behavior, Sept.-Oct. 1915, vol. 5, No. 5, pp. 352, 353.

62 THE GENETIC AND THE OPERATIVE EVIDENCE

PART III.
THE GENETIC AND THE OPERATIVE EVIDENCE.

The genetic and operative evidence shows that there has been
included under the general term "secondary sexual characters" a
complex of cases that are the outcome of diverse physiological processes.
Sex-linked and sex-limited characters have often been confused; some
characters depend on the gonad; some of these involve the ovary,
others the testes. Still other characters fall under none of these groups,
but are the direct product of the male or female genetic constitution.
It is not surprising, therefore, that theories proposed on the informa-
tion derived from certain of these data are controverted by information
derived from other data. The theory of sexual selection, in its attempt
to bring all the facts under one point of view, has not escaped these
difficulties, even although it may be said that neither natural selection
nor sexual selection is concerned with the origin or even the kind of
variations with which it works. Nevertheless, the latter theory, by
ignoring the origin or the physiological process concerned in the
production of secondary sexual characters, may make assumptions
that are difficult to harmonize with the facts in the case, and we shall
find several instances of this sort. For example, if the hen had selected
the cock for his fine plumage (which, as we have seen, depends in part
on autosomal genes producing their effect without the cooperation of
the testes), she would be expected to endow herself with the same
adornments (if her selection worked), unless her ovary were already
producing some substance inimical to those that she is " calling forth"
by selection of the male. The problem is evidently, then, more com-
plex than appears on the surface, and is not so simple as it seemed
when these essential facts were unknown or ignored.

In the case of other theories, such as those of Wallace and of Cun-
ningham (that appeal more directly to the causes that are producing
the variation out of which the secondary sexual characters are built
up), the absence of information, physiological or genetic, has only too
often given these writers the opportunity to speculate without the re-
straints which a more recent knowledge of the facts has imposed on us.

It is obvious from what we have learned that we shall have to proceed
with more caution in disentangling the evidence before we can hope to
" explain" it. Despite the meagerness of our present information,
enough has been found out to indicate that we must be content for
a while with tentative and partial explanations even in the best-
known cases, and we must, I think, be prepared to admit that no one
theory may be able to account for all of the secondary sexual differences
that exist between the sexes.

The genetic evidence shows, in the case of cock-feathering versus
hen-feathering in birds, that only one or two Mendelian factor differ-

RELATING TO SECONDARY SEXUAL CHARACTERS. G3

ences are involved. The result may seem to mean that the secondary
sexual characters themselves have been acquired historically by a single
evolutionary step, and that in consequence the opportunity for selec-
tion to have accomplished such a result has been enormously facilitated.
Such an argument rests, however, as we know to-day, on a false inter-
pretation of Mendelian heredity. What the evidence really shows is
that one or two genes if present cause the testes to produce some
substance that prevents the cock-feathering from developing. The
genetic complex may require a hundred or a thousand or more special
factors that are directly and indirectly concerned with the development
of the cock-feathering, but one or two other factors may suffice to block
this machinery; or, to change the metaphor, these dominant factors
may be no more than so much sand poured into the clock. The clock
may have been slowly built up historically by many contributory
"factors," but a little sand may spoil its activity. Similarly in the hen
something produced by the ovary prevents the fullest possible genetic
action from taking place. Here at present we do not know whether a
single factor or a hundred "special" factors are necessary to produce
such an inhibition, but if, as one would like to suppose, it is the same or
partly the same genes involved in the ovary, and in the testes of hen-
feathered males, then a relatively few, one or two, factors will suffice
to bar cock-feathering from the female.

In a case like the clover butterfly, where the genetic relations work
out on the theory of one pair of factors that produce two types of
females and one type of male, it seems more reasonable to infer that
such a difference has not been slowly acquired by many smaller muta-
tional changes, because the two types are not adapted to live under
two different environments for which their differences fit them respec-
tively, but to live in the same environment. It has never been claimed,
so far as I know, that these two types of females have arisen through
some males preferring one, some another kind of female, so that even
although it may seem probable that the genetic situation is simple,
the simplicity can not be turned to the advantage of the theory of
sexual selection. It is unnecessary to discuss further the origin of the
factor or factors suppressing the development of one type in the male
or the probability of the multiplicity of such factors. In the rase (if
such species as Papilio memnon and P. polytes, with three types <>t
females, the situation is the same as above, with the addition of the
theory of mimicry, that "explains" some advantage accruing t.» each
type of female. Since the latter is only a form of natural selection,
we are not further concerned with the change here. Punnet t 's excellent
treatment of the problems involved in his recent book on mimicry
brings the subject down to date.

Meager as is the genetic and surgical evidence at present, it is enough
to show that only by further work along these lines can we hope to lay

64 THE GENETIC AND THE OPERATIVE EVIDENCE

a firm foundation for a scientific study of the subject. It is equally
important that critical evidence be obtained in regard to the effect on
the female of males of different types in competition. The instinctive
reactions of animals in these respects, their first reaction, the asso-
ciations that may or may not result, are practically an open field for
investigation. The entire equipment of human psychology of the in-
trospective school, that has been appealed to for help in a situation
itself little understood, reads often more like fiction than like science.

So far as one branch of the subject goes — the possible interpretation
of ornamentation in the male — there seem to be two ways at least in
which the subject calls for immediate investigation: First, if it can be
shown that, other things being equal, a more adorned male rouses the
female to prompter mating, it may be inferred with some probability
that in the long run such conduct would lead to the establishment of
the more effective individual, but this would not be true unless the
males mate, as a rule, more than once, for any advantage that might
accrue to a more ornamented male would not affect the course of evolu-
tion of the species if every other male found a mate too. Second, if it
could be shown that the special ornamentation of the male is only one
of several effects of a gene whose main effect is in some other direction,
then the advantage gained through natural selection in this other
direction would carry in its wake the advance in ornamentation, and if
the change affects one sex more than the other, owing to the difference
in the genetic complex of the two sexes, it would be called a secondary
sexual character.

A. Evidence from Mammals.

Owing to the differences in the secondary sexual characters of dif-
ferent breeds of sheep, we have more genetic information about such
characters in this group than in other groups of mammals. For-
tunately, also, in some of the breeds both castration and ovariotomy
have been performed, and consequencely we are in position to utilize
both sources of information for interpreting the situation. In certain
breeds both males and females have horns (Dorsets), in which case the
horns of the male are larger than those of the female. In other breeds
neither males nor females have horns (Suffolks). In still other breeds
the males have horns and the females are hornless (Merinos and Herd-
wicks). The clearest evidence that we have, both genetic and opera-
tive, is that obtained by Woods, as reported by Bateson, in which
horned (Dorsets) and hornless (Suffolks) breeds were crossed. In the
Dorsets, where both sexes have horns, those of the male are larger than
those in the female. When the young male is castrated the horns
develop, but only as far as in the female. It appears, therefore, that
the presence of the testis, probably through some secretion from it,

RELATING TO SECONDARY SEXUAL CHARACTERS. 65

contributes to the development of the horns. The other race, the
Suffolks, have no horns in either sex. Castration produces no change
in their hornless condition.

When a Dorset ram is crossed to a Suffolk ewe the sons have horns,
the daughters lack them. The reciprocal cross gives the same results.
The factor or factors involved are therefore not sex-linked. When
the Fi's from the cross or from its reciprocal are inbred, four classes
of offspring are produced, namely: Horned male, 3; hornless male, 1 ;
horned female, 1 ; hornless female, 3. The ratios, as above, are approxi-
mately 3:1:1:3.

A simple Mendelian explanation covers the results. If we assume
that the Dorsets, both male and female, are homozygous in a factor
for horns, H, that is not in the sex chromosome, and that the Suffolks
"lack this factor," i. e., that they have an allelemorphic factor for
hornlessness, the germ-cells are H-H and h-h, respectively. Only
one kind of individual, Hh, results in Fi. Since the male with this
formula develops horns, we must conclude that the presence of the
testis (through its secretions) causes horns to develop, while in the
female of this same composition horns are not produced because of the
absence of the testes. The sex-cells in these Fi individuals are H-h
and H-h. Chance meeting of these gametes will give 3 classes of
individuals, irrespective of sex, namely, (1) HH, (2) Hh, (1) hh. The
expectation for the males is that those of the composition (1) HH and
(2) Hh will develop horns, while those of the composition hh will not
develop horns. There should be 3 horned to 1 hornless male. In the
females we expect those with the composition (1) HH to develop
horns, since they have the same formula as the pure Dorset ; those with
the formula Hh are not expected to develop horns, because the Fi females
of this composition do not have horns; those with the formula hh are
not expected to develop horns, because they have the same composit Loo
as have the pure Suffolk. There should be 3 hornless to 1 horned
female. Combining both sexes, the expectation for F2 is 4 horned to
4 hornless. Arranged according to sex, these give the classes realized.
Horned male, 3; hornless male, 1 ; horned female, 1 ; hornless female, 3.
That this is the correct explanation is borne out by back-crossing the
hornless Fi female to a hornless Suffolk ram. The former has two
kinds of gametes, H and h, the latter only gametes that bear the h
factor. Half the sons should be horned, half hornless, because half of
them are Hh and half hh. But none of the daughters should have
horns, because neither the Hh nor the hh females produce horns. This
is the result realized, viz, 3 hornless offspring to 1 horned.

The preceding account of the inheritance of the factor for horns is
based on the combination of Dorsets and Suffolks used by Wood.
That other conditions may exist in other breeds and even in races of

66 THE GENETIC AND THE OPERATIVE EVIDENCE

the same breed is claimed by Arkell as a result of a large number of
crosses that he has carried out. He states, for instance, that in the
great Merino class, with its various sub-breeds, there are flocks in
which the males only are horned, but even here there may be individual
males that are hornless "and at times the females may also show some
signs of horn growth." In America, Arkell states, there are three
types of Merinos — the American, the Delaine, and the Rambouillet.
He quotes Plumb (Types and Breeds of Farm Animals, Boston, 1906)
as stating that "the American Merino ram carries heavy, spirally
twisted horns, but the ewes are hornless ; . . . . that the rams of the
National Standard or Victor-Beald Delaines may or may not have
horns ; that the Dickinson Delaines may have small horns, but a polled
head is pref erred,' ' etc. These conditions suggest that there may be
more than a single factor for horns in sheep or that there may be
modifying factors in certain breeds. In fact, Arkell and Davenport
attempt to cover the results of Arkell's experiments by assuming that
there is an inhibiting factor for horns that is carried by the sex chromo-
some. Such an inhibitor (I) would be double in the XX female and
single in the X male. It is assumed to be incapable of preventing the
development of horns in the heterozygous Hh male, the inhibitor being
there simplex (i.e., one I), while the double inhibitor is capable of prevent-
ing the horns in the heterozygous (Hh) condition, but not of preventing
the development of horns when the homozygous (HH) condition
occurs. There are several objections to this scheme: first, that there is
no evidence that a sex-linked inhibitor is present that affects the horn-
less breeds, for the evidence indicates rather that there is no factor for
horns present in them, at least in the Suffolks; second, the peculiar
balance between the factors for horns and the inhibitor seems an
extremely artificial statement. Arkell and Davenport intimate that
races with horned males and hornless females do not exist in a pure
state. That breeds impure in these respects may exist need not be
denied, but that pure races for such a dimorphic condition do exist
seems probable. Castle states, for instance, that he knows at first
hand of such races of Merinos. Castle also states that castrated Merino
rams in this race do not develop horns, and this result is in accordance
with statements made by Marshall for Herdwicks (a race with horned
males and hornless females). Under the circumstances it is certain
that the presence of the testes is one of the factors in determining
whether horns develop at all (as in Merinos), or in determining the
extent to which they develop (as in the Dorsets), rather than that the
difference between the sexes is due only to an inhibiting genetic factor.
Nevertheless, it may be well to keep open the possibility that there may
be different factors for horns in different races (allelomorphs or others),
or conversely, that the genetic composition of the races is different, the
factor for horns remaining the same, but producing a different effect.

RELATING TO SECONDARY SEXUAL CHARACTERS. 67

It may be pointed out in passing that if, as Arkell assumes, the
hornless races are due to the presence in them of an inhibitor for horns,
the results can be worked out without postulating that the inhibitor
is sex-linked. For example, if the hornless male and female be H II 1 1
and the horned male and female HHii, the F! horned males and
hornless females will be HHIi. The germ-cells will be HI and Hi in
each sex, which, by chance meeting, as shown below, gives the results
obtained by Wood. Thus:

HI v/ Hi female.

HIAHi male.

1H1HI+2, HIHi+1, HiHi.

These formula? give 3 horned males, 1 hornless male, 1 horned
female, 3 hornless females. This formulation, while appealing appar-
ently to a different set of factors from those used by Arkell, is in reality
the same in principle, since the heterozygous condition is here repre-
sented by Ii (instead of Hh) and sex determines that the heterozygous
male is horned and the female hornless.

The genetic relations of the Merino with horned males and hornless
females to the Dorsets, in which both sexes are horned (but in the
male the horns are larger), must be different from the genetic relation
in the other cross. There are two theoretical possibilities, viz, that a
different factor for horns is present that is either an allelomorph or
another different factor; or second, that a modifier is present in the
Merino that keeps down the development of the horns in the female.
An answer could be obtained by breeding Merinos to horned and to
hornless and getting F2 from both crosses. Arkell's data is not sufficient
to settle the question, because his numbers are often too small, but
chiefly because it appears that there were two genetic types present
in his flock of Merinos, one of which is characterized by scurs (very
short horns) in the females, the other by hornlessness in the female.
He found in a cross between a hornless father and Merino mother
(that had knobs or scab-like growths) that the daughters had horns
or scurs and carried a determiner for horns (as subsequent generations
showed). On the other hand, in other cases where the Merino mother
was without horns, her Fi daughters had no horns. In both easts the
Fi sons had horns. Arkell cites this cross as "proving" that the knobs
of Merino ewes depend for their development upon two horn deter-
miners (H'H'). It is not at all evident that the results lead to such
a conclusion, as other explanations will cover the ease as well.

Arkell's mating between Dorsets and Merinos (tables EX and \vi)
corroborates his view "that the knob of the Merino female is repre-
sented in the germ-plasm by the double determiner." The 5 I\ bods
had long horns, 3 Fi daughters had horns present, and 2 had them
absent (table xvi). If some of the Merino mothers used were homo-
zygous for a factor that inhibits the development of horns in the female

68 THE GENETIC AND THE OPERATIVE EVIDENCE

we can account for the hornless daughters, and if other mothers did
not have this factor (or were heterozygous for it) we can account for
the horned daughters. Evidently more evidence is needed. Arkell
himself assigns a corresponding difference to the mothers in these cases,
based on the observed fact that the mother that had knobs or scurs
were the ones that gave birth to the horned daughters. If the above
suggestion proves true, it shows that the Merino condition dominates
the Dorset condition. The result is in harmony with the view that
both have a common factor for horns, but that in addition the Merinos
have a non-sex-linked modifier that holds down the development of the
horns in the ewe.

What bearing have these results on the theory of sexual selection?
Clearly the Merino male, as constituted at present, develops horns
because he is a male, but only in the sense that his testes secrete some
substance that makes his horns grow. That maleness does not in itself
necessarily produce horn is shown by the absence of horns in the
Suffolk breed. Is it the same factor, present in the Merino, that
produces horns in both sexes of Dorsets when homozygous and in the
male only when heterozygous? If originally the ancestral race had no
horns, the appearance of factors for horns would, even in a heterozygous
condition, have sufficed in the males for the development of horns.
If this gave them any advantage either over the enemies of the race or
in the eyes of the female, such factors might be perpetuated, and
through transferrence to the females ultimately become homozygous
in both sexes. Both would then have horns, whether horns were or
were not of any advantage to the female, which would have them
because they have an advantage to the other sex.

Because the genetic evidence shows that a single factor difference
between the breeds with and without horns accounts for the horned
condition in one of them, it by no means follows that horns as they
exist arose as a single mutant factor change. True, they may have
arisen as a new single factor difference, but the Mendelian evidence
can not be claimed as evidence for this view. The a priori argument
based on the relation of horns in an adaptive sense to the rest of the
body would appear rather to indicate that they could not have arisen
at a single mutational step.

Concerning the still broader bearing of this evidence on the theory
of sexual selection, two distinct questions are involved: first, how has
the present racial difference in horns arisen in domesticated sheep, and
secondly, what was the original condition of sheep. Reversing the
order of these questions, we find that sheep were domesticated in Asia
and Europe before the dawn of history. "Whether our well-known
and useful animal is derived from any one of the existing wild species,
or from the crossing of several, or from some now extinct species, is
quite a matter of conjecture" (Flower and Lydekker's "Mammals").

RELATING TO SECONDARY SEXUAL CHARACTERS. 60

Most of the wild species of the genus (of which about 12 are recognized i
have horns in both sexes, but larger in the male. There are 'A wild
species in which the horns are lacking in the female, according to
Flower and Lydekker. If these have been crossed into the domesti-
cated breeds the condition shown by the Merino may go back to the
wild state. The third condition found in domesticated races, viz,
hornlessness, may have appeared under domestication. Such a change
might have arisen in either of the two other types and would be com-
parable to well-known losses of characters shown by domesticated
animals and plants. These losses of characters are usually ascribed
to actual losses of genes; any lost gene in the complex of factors neces-
sary for the production of horns might cause such a change. But there
is no advantage, in fact, in ascribing the loss in the character to a lost
in one of the factors producing that character, for any change of any
kind in the factor complex might bring about the same result and the
evidence from multiple allelomorphs should put us on our guard
against the all too easy assumption that a loss in a character involves
necessarily loss of a factor in the real sense in which loss is used in
ordinary speech.

The operative and genetic evidence for sheep shows that if the horns
in the male were developed through natural or sexual selection we
should expect them to develop also in the female. The greater develop-
ment in the male seems to be due to secretions from the testes which
probably are due to special factors that call them forth, but whether
such factors were also acquired to reinforce the effects being produced
through selection or were already present (reinforcement for horns
being only a by-product of their activity) can not of course be known.
We can suppose that special factors that suppress the development of
horns in the female may have arisen in the wild or in the domesticated
races and have been perpetuated because of some imagined benefit
conferred; or that in certain races factors were already present that
kept down the development of horns in the female. In any case such
factors do not cause their effects through secretions from the ovary,
because after ovariotomy horns do not develop; nor are they sex-
linked factors. Any speculation as to how natural or sexual selection
has brought about the evolution of the horns in sheep must reckon
with the conditions imposed on such speculation by the preceding
information. So far as I can see, it leaves the situation in this reaped
neither better nor worse off than before.

In deer the effects of castration are well known, but there is no
genetic evidence to show the kind of factors involved, since no crosses
have been made between species with differences in their horns. If t he
young male deer is castrated before the antlers have appeared, no horn-
develop. If castrated at the time when the antlers have begun to
develop, incomplete or imperfect development follows. The antlen

70 THE GENETIC AND THE OPERATIVE EVIDENCE

remain covered with the velvet, and are said not to be thrown off
periodically as in the normal male. If the adult stag with antlers is
castrated, the horns are precociously dropped, and, if replaced at all,
the new antlers are imperfect and are not renewed. I do not know of
any cases in which females have been spayed, but no doubt the ovaries
must sometimes become diseased. There are, however, a few records
of horns developing in this sex in old age, or presumably after disease
of the ovaries. Both male and female reindeer are horned. Castration
produces no effect on the development of the horns.

In the case of deer it is evident that the presence of the testes in the
male causes the horns to develop. The genetic factor, or factors, for
horns may be supposed to be carried by both sexes, but the effects of
the factor can be seen only when the testes are present. In the reindeer
and eland, on the other hand, the genetic factor for sex can produce
horns without the need of the environment produced by the testes.1
Whether we are dealing here with the same factor or whether the rest
of the hereditary complex makes the result different can not be known
without breeding experiments.

There is apparently a connection between the stage of development
of the horns and the age of the animal, as the following statement by
Yarrell2 (1858) indicates:

"The fallow-buck is at his best in his sixth, or at most in his seventh year;
after which, though the carcass may increase, the horns become smaller, and
irregularly going back annually through something like their former stages of
increase, a very old buck has from the state of his horns been mistaken for a
young one. In the osteological department of the Museum at Paris there
was, and may be now, the skeleton of a female reindeer in which the horns were
reduced to little more than a rudiment of the beam and the brow-antler; this
animal was so old that the molar teeth were worn down to the edges of the
alveolar cavities."

At first sight these results in the fallow deer appear to be only an age
condition, but since in old age a reverse process sets in, it may appear
more probable that the amount of secretion by the testes or other
glands may be the conditioning agent. In the case of the reindeer one
may hesitate to ascribe the change to the ovary without further
evidence.

In cattle the effects of castration as seen in oxen have been studied.
There is little here that is useful for our present purpose. The horns
are not inhibited and may even be larger than in the bull. The absence
of horns in certain races of cattle is apparently a dominant character,
but as the character is neither sex-limited nor sex-linked, the evidence
has no further bearing on the present topic.

1 In the eland as well as in the reindeer, in which both sexes have horns that begin in the latter
at least to develop before the gonads ripen, it is stated that castration does not prevent the devel-
opment of the horns in the male, but whether they are as large as in the normal male is apparently
not definitely stated.

2 Yarrell also states that after the fallow buck has reached the height of its maturity and
has 6 prongs in its antler, removal of one testis causes the next antler to have but 5 prongs.

RELATING TO SECONDARY SEXUAL CHARACTERS. 71

The effect of removal of the ovary from female calves has Keen
studied by Tandler and Keller. The height of the ovariotamiied female

is less than that of the cow. The same difference is found bet ween bull
and ox. Tandler and Keller call attention to the similarity of t he head
in male and female lacking the gonads. They conclude that the
ovariotomized female does not come to resemble the male, but that
removal of the gonad causes both sexes to converge to a common type.

Castration is frequently performed in horses, dogs, and cats, but as
the secondary sexual differences, aside from size and behavior, are not
very well marked in these animals, the results need not be here con-
sidered.

Steinach's experiments with rats are important, because by grafting
ovarian tissue into the castrated male, the male was caused to assume
certain characteristics peculiar to the female. The mammary glands
that are rudimentary in the male became much enlarged — not only
the glandular tissue increased in amount, but the mammae themselves
were greatly developed. The hair of the male is coarser than that of
the female. In the feminized male the hair was soft like that of the
female. The size was smaller than that of the male. The skeleton also
was affected, and Steinach thinks that it changed in the direction of a
female skeleton. Even more striking was the sexual behavior of the
feminized rat. The individual no longer reacted as male, but showed
some of the reflexes peculiar to the female. These results, that stand
almost alone, appear to show that several of the secondary sexual
characters of the female rat are due directly to the presence of the
ovary.

One of the most striking and definite results shown by castrated
rats (Steinach), guinea-pigs (Pirsche, Steinach), rabbits (Pauneet),
hedgehog (Marshall), and man is to be seen in the effect on the M»
sory glands connected with the male ducts as well as on the penis.
These remain small and infantile. Some substances produced by the
testes are essential for the development of these parts. Natural
selection rather than sexual selection would be the agency that here
comes into play.

In man the effects of castration have been often described. Kunuchs
have had a commercial value in some countries, as in Turkey :uul
China, and castration has been deliberately practiced on young
children. Certain religious sects, such as the Skops of Russia, have
advocated and carried out the operation. Disease has also at times
necessitated the removal of the testis, more often in adults than in the
young. The full effects are shown only when the operation has been
carried out before the secondary sexual characters have developed
The more striking difference between the sexes involve the heard, and
the hair on other parts of the body, the voice, the shape of bh* pelvis.
and the mammary glands. For a detailed account of the results.
the publications of Tandler and Grosz and Marshall's book on the
"Physiology of Reproduction" should be consulted.

72 THE GENETIC AND THE OPERATIVE EVIDENCE

The two most obvious changes in the eunuch are the absence oi the
beard and mustache and the small larynx, which produces a high-
pitched voice. In both these respects man differs from woman; in
both, however, the eunuch is like the boy as much as he is like the
woman. It is not evident, therefore, whether the eunuch has retained
the juvenile condition or has become more like the female. Moreover,
there is the possibility that there is no difference in the present case
between these two conditions. The distribution of hair on the pubis
of the eunuch is often said to be more like that in the woman than that
in the man, but there is apparently no sufficient evidence to show that
this is more than the juvenile condition or an undeveloped condition
of the male. As to the voice, there is no way of determining whether
the voice of the eunuch is feminine or juvenile. The development of
the mammae in the eunuch would be a better test, but it does not appear
from theliterature on the subject that the mammary glands and the
nipples of the eunuch are changed toward the female type. On the
contary, it appears rather that there is no such change. It is true that
the tendency toward the accumulation of fat may give the eunuch a
somewhat feminine appearance (since one of the foci of fat accumula-
tion is in the region of the breasts), but this in itself can scarcely be
claimed to be feminization, but due rather to the more slothful habit of
the eunuch that tends to obesity.

A more suggestive resemblance is found in the narrowness of the
shoulder girdle and broadness of the hips in the eunuch, but even these
rsemblances to the female should be regarded skeptically, since other
changes in the bones that result from castration are certainly not a
development toward the female type, but a peculiar specific effect of
the absence of testes on the growth of the bones. For instance, the
bones of the arms and legs are much longer in the eunuch than in
either the normal man or woman, in fact, more in the direction of the
male, who has longer legs than the female. The explanation usually
given is that the ossification at the ends of the bones and of the epiphyses
does not take place so soon as in normal men and women. The con-
dition here is that characteristic of the juvenile state that is carried
over into the adult, but whether the narrowness of the chest and
shoulder girdle of the eunuch is correlated in some way with the more
prolonged growth of the other bones has not, so far as I know, been
determined. That there is no apparent connection between the
shortness of the one and the greater length of the other does not
necessarily lead to the conclusion that there is no such connection.
For the present I think we must hold this point in reserve.

Steinach's evidence for the feminized rats, if it may be extended to
man, indicates that some of the female characteristics are due to the
presence of the ovary holding in check the genetic possibilities of the
female, as well as leading to the development of such characteristic

RELATING TO SECONDARY SEXUAL CHARACTERS. 7^

traits as the mamma?, etc. In the case of the pelvis the female departs
from the juvenile type of both sexes, and here one might look for
a better criterion. It is stated that the pelvis of the ox is more like
that of the female than it is like that of the male, and it lias been said
that this is true for the castrated rat and guinea-pig, but whether a
simple enlargement of the juvenile pelvis would make it resemble the
female type more than that of the male has not, so far as I know, been
carefully examined. Should it prove here that this is the case, the
evidence on this point would be no stronger than that for other charac-
ter differences. As has been stated, Tandler and Grosz think that the
changes in the skeleton of the ox, as well as those in the castrated cow
(skull, pelvis, and limb bones), are due directly to loss of the gonads
and are much the same in both. But their resemblance may possibly
be due more to an enlarged juvenile condition rather than that either
of them changes toward the normal skeleton of the other sex.

The statements that have been published concerning the effects of
removal of the ovaries in woman are, on the whole, unsatisfactory and
often contradictory. That the uterus and oviducts become smaller is
expected from what is known to occur in other mammals, and is
definitely recorded in the human female. That the breasts become
smaller is stated to be the case, but whether because of an actual
decrease in the glandular portion has not, so far as I know, been
shown. That hair is likely to develop on the upper lip of woman
without ovaries is also claimed as likely to occur, and this, too, is
sometimes seen in old women, but if it is interpreted to mean an
approach to the bearded condition of man it should be admitted that
the development is hardly sufficient to invite such a comparison.
Finally, it has been stated that the voice becomes deeper, more, there-
fore, like the male, but this has also been denied. If it could be
established that the voice changes and that it was brought about by
an enlargement of the larynx, similar to that which takes place when
the larynx of the boy changes to that of the man, it might seem not
improbable that the change was toward that of the opposite Bex
This would mean that the ovary produces some substance that pre-
vents the enlargement of the larynx in the female. But since it has
been shown that the enlargement in the male is caused by the develop-
ment of the testes, and that this enlargement is prevented by east rat i< m,
a paradoxical situation would present itself, viz, that the testes cause
the larynx to enlarge in the male and the ovary prevents the enlai
ment in the female. Until convincing evidence is forthcoming, the
question is better left undecided.

B. Evidence from Birds.

Probably a greater difference in the secondary sexual characters is
shown in birds than in any other group. It is true that there are

74 THE GENETIC AND THE OPERATIVE EVIDENCE

species, such as the doves and pigeons, in which the plumage of the
male is much like that of the female, but this is the exception rather
than the rule. At the other extreme are species like birds of paradise,
hummingbirds, fowls, pheasants, ducks, and many passerines, in
which the plumage of the two sexes is entirely different. Our knowledge
as to the relation between the nuptial plumage of the male and the
condition of the sex-organs rests largely on information gained by
castration in poultry and ducks and on the assumption of the nuptial
plumage in several species only at the mating season.

John Hunter in 1780 described a pheasant with male plumage. His
account of a similar change in a pea fowl is so complete that I venture
to quote it in full:

"Lady Tynte had a favorite pyed pea-hen, which had produced chickens
eight several times; having moulted when she was about eleven years old, she
astonished the lady and her family by showing the feathers peculiar to the
other sex, and appearing like a pyed peacock. In this process the tail, which
was similar to that of a cock, first appeared after moulting. In the following
year she moulted again, and produced the same feathers. In the third year
she did the same; at the same time she had spurs similar to those of a cock.
She died in the following winter during the hard frost, namely, in the winter
1775-6. She never bred after this change in her plumage. This bird is now
preserved in the Museum of Sir Ashton Lever." l

"From what has been related of these two birds, may it not reasonably be
inferred that it seems probable that all those wild pheasants of the female sex,
which are found with the feathers of the cock, had changed the nature of their
feathers, particularly at a certain age?

" If this idea be just, it shews that there is a disposition in the female to come
nearer and nearer to the male, at least in the secondary properties; or it may
rather be said that the female is later in producing this change than the male is;
for it has already been observed that both sexes when young differ not from
each other in these respects, but that the male appears to be the one that by
degrees separates from the female in its secondary properties."

Statements in regard to the effect of castration on poultry go back,
it appears, to Aristotle. Yarrel in 1811 and again in 1850 has given an
excellent account of many of the effects produced. His account of the
effects on the cock seem to be based partly on hearsay, and while they
contain much accurate information, yet the statement that the
plumage of the capon is intermediate between that of the cock and hen
is incorrect. The further statement that by cutting the oviduct the
hen assumes the plumage of the capon has been shown by Sellheim to
be erroneous. The operation referred to by Yarrel must have been
one in which the ovary was removed.

1 It might be supposed that this bird was really a cock which had been changed for a hen; but
the following facts put this matter beyond a doubt: First, there was no other pyed pea-fowl in
the country. Secondly, the hen had knobs on her toes, which were the same after her change.
Thirdly, she was as small after the change as before, therefore too small for a cock. Fourthly, she
was a favorite bird, and was generally fed by the lady, and used to come for her meat, which
she still continued to do after the change in the feathers.

RELATING TO SECONDARY SEXUAL CHARACTERS. 75

Yarrel described a female pheasant that had assumed some of tho
characteristic colors of the male. On dissection he found that the
ovary was diseased as well as the oviduct. He correctly assigns the
change in plumage to the condition of the ovary. He states further-
more that most of the female pheasants that he had examined t hat had
male plumage had not assumed the complete coloration of the male
In one case, however, a complete change had taken place. The change
in pheasants he thought was due to old age accompanied by partial
or complete loss of function of the ovary. For poultry he stat

"In the imperfect female the comb increases; a short spur or spurs appear;
the plumage undergoes an alteration, getting what is usually called 'foul-
feathered;' she ceases to produce any eggs, and makes an imperfect attempt to
imitate the crow of the cock. Being profitless in this state, she is usually made
away with. The proverb says :

A whistling woman and a crowing hen
Are neither good for gods nor men.

Our neighbors and allies the French, who seem to take a wider range in their
prejudice against habits which they consider irregular, have the following
proverb, which says:

Poule qui ehante, Pretre qui danse
Et Femme qui parle latin,
N'arrivent jamais a belle fin.

"I have seen two instances in which females of the wild duck have assumed
to a considerable extent the appearance of the plumage of the mallard, even to
the curled feathers of the tail. One of these birds, in my own collection, was
given me when alive by my kind friend the late John Morgan, esq. When this
bird was examined after death, the sexual organs were found to be diseased, as
in the case of the hen pheasants referred to, and figured in the 2d volume of the
History of our British Birds. In the published illustrations to his Fauna of
Scandinavia, M. Nilsson has given a colored figure of a duck in this state of
plumage (plate 163), which is called a barren female, and in which the curled
tail-feathers are made very conspicuous.

"From the general similarity in these females to the appearance assumed
for a time by healthy males in July, I am disposed to refer this seasonal change
in males, in this and in other species of ducks, to a temporary exhausted state
of the male generative organs, and their consequent diminished constitutional
influence on the plumage.

"A male shut up by himself from early spring to the end of July undergoes
no change in his plumage; but if he is allowed to associate with females till
their season of incubation commecnes, he then goes through the change, and
this appears to indicate the cause of the partial summer moulting.

"The appearance is somewhat different, but yet very interesting in insects
and Crustacea. In these classes the sexual organs are double and distinct,
arranged one on each side of the elongated mesial line. It sometimes happens,
that a species in which the sexes are of a different color, or markings, or form
has one sexual organ of each sort, male and female, in which case each half
of the same insect is developed under the exclusive influence of the sexual organ
on its own side. Instances are preserved among our collect ionfl of buttert!
mothes and beetles; and I have seen it twice in the common lobster.

76 THE GENETIC AND THE OPERATIVE EVIDENCE

"Nor is the human race exempt from the operation of the law which prevails
in the Mammalia. In women, at an advanced age, hair appears on the chin
and upper lip, and the voice alters, becoming deep in its tone. The beard in
old men becomes thin and soft, and our own inimitable Shakespeare has told us,

* * * his big manly voice
Turning again toward childish treble, pipes
And whistles in his sound.

Gurney (1888) has recorded several cases in which female birds
have assumed male plumage. For instance, he describes a female
merganser, Mergus serrator, assuming male plumage that showed no
signs of disease in the ovary. Mr. Cecil Smith had a female widgeon
(Mareca penelope) on his ponds near Trenton, which assumed the male
plumage some years ago, and which, so far as he knew, had not had
young nor laid eggs.

"On May 16th, 1887, a chaffinch (Fringilla Calebs) in full male plumage was
shot at Chapel Town, near Leeds, in Yorkshire, by the son of Mr. W. L. Jack-
son, M. P.; it was skinned by G. R. Grassham, assistant to Mr. W. E. Clarke
at the Museum, who, much to his surprise, found that it was a female, and
contained an egg, ready for laying, of a pale blue, without markings, and an-
other egg in a less forward state. This chaffinch is in every way in perfect
male plumage, and I am indebted to Mr. Clarke for his kindness in sending
these particulars with the specimen, which he received from Grassham a few
hours after the latter had dissected the bird.

"In the 'Norwich Nat. Trans.,' an enumeration was given of female Red-
starts (Ruticilla phoenicurus) assuming male plumage (I. c.) to which the fol-
lowing may be added: a hen R. phoenicurus assuming male plumage, and very
like Mr. Millais' described in the 'Norwich Nat. Trans.' iv., p. 182, was caught
by Mr. W. E. Clarke sitting upon her eggs, at Wike, near Leeds, in June, 1886;
at the same time Mr. Clarke saw the cock close by, which appeared to be in the
ordinary male plumage. The late Mr. Henry Doubleday's collection con-
tained a hen Redstart (R. phoenicurus) in male plumage, which had the ovaries
'quite perfect and full of eggs' (cf. B. of Norf., i, p. 370, note), probably one of
those alluded to by Yarrell (Brit. B. 1st ed. i, p. 240) in the remarks made by
him on the plumage of this species. I have some recollection of this Redstart
at the dispersal of Mr. Doubleday's collection, but do not know who was
the purchaser of it. There can be no doubt that more would soon turn
up if looked for; and now that attention has been drawn to the subject,
and the practice of dissection is getting more general among bird stuffers, it is
certain to be the case, not only in Ruticilla, but in other genera besides. Why
it should happen in Ruticilla phoenicurus oftener than in other Passerine birds
is hard to explain, but such is evidently the case."

"The same is recorded to have happened five or six times with the female
Red-backed Shrike (Lanius colluria); see 'the Field,' June 17, 1871, and April
25, 1885; Mag. N. H., iv, p. 344; 'B. of Suffolk,' p. 45; 'Ibis,' 1863, p. 292; but
the number of hen Redstarts which have donned masculine attire is greater.

"The following is a list of the species in which one or more instances of
females assuming male plumage are ascertained to have occurred :

Falco aesalon, fide Scully. (Cf. Sharpe, 'Cat. Birds Brit. Mus./ i, p. 407).

Tinnunculus alaudarius, fide Sharpe; col. fig. P. Z. S., 1874, p. 580.

Lanius collurio, fide Hoy.

Lanius vittatus, fide Blyth.

Ruticilla phoenicurus, fide Millais, Clarke and others.

RELATING TO SECONDARY SEXUAL CHARACTERS. 77

Fringilla coolebs, fide Clarke.

Linota cannabina, fide Blyth.

Linota rufescens, fide Blyth.

Nectarinia asiatica, fide Blyth.

Gallus (domestic fowl), fide Yarrell and others; col. fig. "B. of Sherwood," p. 183.

Pavo (peahen), fide Latham; fig. "Synopsis," ii, pi. 60.

Meleagris (Turkey), fide Bechstein.

Phasianus colchicus, fide Edwards and others. Of common occurrence in a semi-
domesticated state.

Thaumalea picta, fide Edwards.

Euplocamus nycthemerus, fide Yarrell.

Pucrasia nipalensis, fide Blyth.

Tetrao tetrix, fide Bond; col. fig. Dresser, "B. of Eur.," vi, 205.

Tetrao urogallus, fide Nilsson; col. fig. "Unser Auer-, Rackel- und Birkwild und
seine Abarten," by A. B. Meyer.

Otis tarda, fide Tiedmann.

Anas (domestic duck), fide Rowley; col. fig. "Orn. Misc.," i, p. 118.

Anas boschas, fide Hancock; fig. col. " Scandinavisk Fauna," pi. 163.

Fuligula marila, fide Blyth; see also P. Z. S., 1885, p. 246.

Mergus serrator, fide Gurney.

Mareca penelope, fide Cecil Smith.

"Perhaps the Kestrel (Tinnunculus alaudarius) ought not to be included
in this catalogue, for so many have been seen with the lower part of the back
blue or bluish, as to leave little doubt that the female generally becomes so if
she lives long enough.

"It is said that the females in Oriolus generally become as bright as males in
time ('Ibis/ 1864, p. 412; 'Field,' June 24th and July 8th, 1871)."

"P. S. — Mr. W. Tegetmeier tells me he has known a barnyard cock moult
into hen's plumage, which is the converse of the instances narrated in this
paper, and rather resembles the annual change which takes place in Anas
boschas and others of that tribe."

In a later notice Gurney makes the following statement :

"The bearded tit {Panurus biarmicus) may be added to the list of female
birds which are known to occasionally assume male plumage. In the summer
of 1882 a bearded tit, two years old, in Mr. J. G. Keulemans' aviary, hatched
five eggs and moulted, during which operation she suffered much from cold
and stiffness, and when she recovered her plumage it was partly that of the
male (c/. 'The Field,' Sept. 14, 1872)."

Brandt, who has reviewed the literature very thoroughly, cites the
following cases:

"Galeinacei: Gallus bankiva domest., Phasianus pictus, torquatus, col-
chicus, mongolicus and nycthemerus, Pavo cristatus domost., Meleaeris gallo-
pave domest., Perdix einerea, Tetrao urogallus, tetrix und bonasia.

"Passeres: Fringilla coelebs, Pyrrhula vulgaris, coccinea, Loxia chlons,
Turdus merula, Ruticilla phoenicurus, ochrura, chryaogastra, ("vallecula
Wolfii, Sturnus vulgaris, Ampelis cotinga.

"Scansores: Cuculus canorus, Edolius glandarius.

" Grallatores : Machetes pugnax.

"Natatores: Anas boschas domest.

"Es ware denkbar, dass die Hahnenfedrigkeit, wenn audi in verkapptem
Grade, alien Vogeln, selbst denjenigen zukomme, deren Gefieder una gea-
chlechtlich uniform zu sein scheint, Wie dem auch sei. cinzelne Genera und
Species scheinen mehr, andere weniger zur Arrhenoidie pnidisponirt.

78 THE GENETIC AND THE OPERATIVE EVIDENCE

bemerkt J. Geoffroy St. Hilaire (p. 511), dass Fasanen haufiger selbst als die
Hiihner hahnenfedrig werden, wahrend fur den Pfau, den man doch stets eines
naturlichen Todes sterben lasst, ihm nur ein einziger Fall (der von Hunter)
bekannt geworden. Wahrend Lorenz (vide Tichomirow) auf dem Moskauer
Markt haufiger hahnenfedrige Weibchen von Phasianus colchicus und mongol-
icus aufgefunden, ist ihm dieses fur Ph. chrysomelas bisher kein einziges Mai
gelungen, obgleich die Zahl der jahrlich in Moskau feilgebotenen Exemplare
dieser Art sich auf 8000 Stuck belaufen mochte."

The preceding cases relate to exceptional changes in the plumage as
observed in nature, or in birds kept under domestication. We may next
examine the cases where the ovary or the testis has been removed.

The earlier observations of Berthold, Wagner, Hanau, Samuel,
Sellheim, Pirsche, Foges, Shattock, and Seligman are sufficiently
covered by later work quoted below. Sellheim's work, however, is
especially to be noted, since he gives some measurements covering the
weight of the brain, heart, and body of the cock and capon, as well
as observations on the skull and skeleton. The weight of the brain
is slightly less in the capon, but the body-weight is greater. He ques-
tions whether the ovary has ever been successfully removed, and he
shows that the operation of resecting the oviduct does not, as was
supposed, lead to the degeneration of the ovary. On the contrary, he
found that after the effects of the operation had been removed the
ovary began again its functions.

From Goodale's careful summing up of the effects of castration only
the following points need be recalled: The feathers are little changed;
some of them, the hackles especially, become longer. The lowermost
tier of wing coverts are elongated as compared with those of the cock.
The spurs are practically the same in the capon and cock. The capon
is disinclined to give voice, but at times he crows. The molting is not
affected. The size of the capon is larger. He pays little attention to
the hens. He is not pugnacious, and if attacked will not often fight.
As a rule he does not pursue the hens, but if a hen squats down as the
capon approaches he will mount and go through the characteristic
mating reaction. The comb is extremely small, much smaller than
that of the female of the same race ; it is infantile rather than feminine.

Comparing these results with those that I have observed in the
castrated Sebright, we find that aside from the assumption of the full
plumage of the cock-feathered bird the Sebright shows all of the
characteristic features of the capon. The spurs develop, perhaps even
more fully than in the normal Sebright cock. He seldom crows, and
then weakly. The birds appear large, but the excessive development
of the feathers produces the effect. I have not weighed them to show
whether an actual increase in size takes place. Two of my birds are
notably large for Sebrights, but the others are smaller. Both large and
small cocks occur in the strain that I have used. My Sebright and
other capons neglect the hens, but I have seen them tread the hens

RELATING TO SECONDARY SEXUAL CHARACTER 79

on occasion. They will fight each other, if two strangers meet . 1 >ui the
attacks are not violent or prolonged. A normal male beata them
easily, and afterwards they run away from such birds. The combs and
wattles are very small and pale. If a piece of the testis is left in, the
comb is a fair index of its size. In the birds that changed back toward
a Sebright the comb slowly enlarged. After the second operation it
decreased again as the plumage once more changed to that of the cock.

Goodale's results with ovariotomized females are especially note-
worthy, since here for the first time we have definite information as to
the effects of the operation. By using a well-established breed, the
brown Leghorn, in which the dimorphism of the sexes is very striking,
the results are made all the more convincing. Goodale found that it was
possible to completely remove the ovary of young birds, for at an early
age the ovary is sufficiently compact to make its entire removal ]
sible. Later the ovary becomes more diffuse, and complete removal
is almost impossible. In a few successful cases, in which the ovary
had been completely removed, the bird assumed the full plumage of
the Leghorn cock, with red back, black breast, and long, pointed hackle
and saddle feathers. Spurs developed in all the operated females, even
when the ovary was not entirely removed. There can be little doubt
that the ovary holds back the development of the spurs, but as some
hens sometimes develop spurs, especially in certain breeds, it is not
entirely certain that in these cases the loss of the ovary is the cause of
the appearance. The comb (and wattles) developed to different
degrees; in some birds it was as large as in the cocks, in others no
larger than in the normal hen, but in all cases it was larger than in the
capon. What to conclude is doubtful. Tentatively it may be suggest ed
that the genetic complex that gives the female (ZW) produces a comb
as large as that shown by the female independently of the ovary, but
beyond this point the ovary inhibits the further development of the
comb, presumably by means of the same internal secretion that holds
down the cock plumage in the hen. In the male, on the other hand,
the genetic complex (ZZ) produces a comb much smaller than that of
the female (no more than that of the capon), and the testes produce a
substance that causes this comb to grow to the size of that of the eock.
Possibly, however, other internal secretions are involved.

The operated hens are quiet and nearly voiceless. None of Goodale B
birds were heard to crow, yet this seems to be a well-known peculiarity
of old hens that have become cock-feathered. The operated hens are
not larger than the normal hens of the same breed. Their legs reman)
short, as in the normal hen; and in this respect and in size the ovarioto-
mized bird is externally a female. The poullards "never visit the
nests, never sing or cackle, show none of the normal female reactions,
and few or none of the male."

80 THE GENETIC AND THE OPERATIVE EVIDENCE

The influence of the ovary in suppressing the cock plumage has been
convincingly shown in an experiment of Goodale's, in which, after
removal of both testes from the young Leghorn cock, pieces of ovaries
were inserted into the body-cavity. As dissection showed later, several
of these implanted pieces grew onto the wall of the body-cavity. The
birds developed the plumage of a hen, although some traces of the male
plumage were at times present. The difference between the sexes is so
great in Brown Leghorns that the hen-feathering of the feminized
cockerels leaves no doubt that the presence of the ovary had produced
the female coloration.

Geoffrey Smith and Mrs. Haig Thomas (1913) have examined a
number of hybrid pheasants, some of which were sterile. They found
that the ovary (and oviduct) was often small and degenerate. There
was a more or less corresponding tendency for such female hybrids to
show male feathering, at least in a part of the plumage. The degenera-
tion of the sex element, however, does not take place until after the
time of synapsis, so that the younger germ-cells may be normal. The
later degeneration of these cells is not likely to influence the secondary
sexual characters, but may be an index of changes in other parts of
the ovary.

Geoffrey Smith had a breed of White Leghorns with cocks of two
classes — those that assumed cock plumage at 6 months, and those that
are like the hens for 8 months, after which they slowly assume the
cock-feathering. The difference is hereditary and appears to segregate.
Possibly this breed had one factor at least for hen-feathering that is
more effective for young birds than for older ones.

Smith states that birds and crabs (see infra) appear to give opposite
results, since removal of the ovary in the former leads to development
of secondary male characters and removal of testes in the latter to
secondary female characters. But he adds that he thinks the results
are really the same, because in the crab it is not the suppression of the
testis but the feminization of the male by the Sacculina that causes the
change.

There are a number of observations on ducks. Several cases have
been recorded where in old age the female assumed the male plumage
(Darwin, Shattock, and Sellheim). Also a few cases in which the testes
were removed. Those of Goodale are the most complete and striking.
The male duck has two characteristic plumages, one called the nuptial
also called the summer or breeding plumage that is assumed at
the molt in the autumn, and the other the eclipse plumage, which is
not identical with but much like that of the female. Here, then, we
find a new situation, and one that invites comparison with the con-
dition in Sebrights, in so far as the male becomes hen-feathered at
certain seasons.

RELATING TO SECONDARY SEXUAL CHARACTERS. 81

Throughout the greater part of the year the Rouen drake lias the
nuptial plumage. The head is green and the breast is claret. Two
median tail feathers are strongly curved; the next two are also often
curved. These four are called the sex feathers. At the close of the
breeding-season (July) both sexes molt. The male now has the same
coat as the female, or nearly so. The green head becomes brown to buff ;
the sex feathers are straight. The change back again to the nuptial
plumage beginsat the end of summer and is completed early in October.
Thus in the race of Rouens the eclipse plumage lasts only a very short
time. In the mallard it lasts longer. The eclipse plumage develops,
therefore, only when the testes are active, or, as Goodale puts it, "the
presence of the active testis is necessary for the drake to assume this
plumage." Conversely, the nuptial plumage comes on in the late
summer, when mating is over, and when the testes have shrunken and
are not active, at least as far as the sex-cells are concerned. In some
respects the situation is like that in the fowls, for in both the testes are
not necessary for the development of the full plumage, but in other
respects the situation is different, because at the time in the ducks
when the testes are active the eclipse plumage develops. Are we to
suppose that at the time of sexual activity a substance is produced
analogous to that produced by the ovary of the female? This seems
the most plausible assumption, for we know that if the testis is removed
the eclipse plumage does not appear. Such a situation suggests a
comparison with the Sebright, where it has been shown that the testis
must actively produce some substance which, like that in the ovary,
keeps down cock-feathering. It is plausible, even if it can not be
established, that the substance in the duck and the inhibitory substance
in the male Sebright are the same as that produced in the female.

Goodale's results with females (ducks) are not so clear cut, because
the ovariotomized females turned out to be of two sorts. One sort is
almost identical with the male, the other is more intermediate. There
are sufficient reasons for thinking, he says, that these differences are
not due to defective operations. Goodale suggests a genetic difference
in the females used, but this is apparently even to Goodale himself
not a very satisfactory solution. For our present purpose the impor-
tant fact is that the ovariotomized female may assume the perfect male
plumage. Evidently the ovary produces some substance which, as in
the hen, suppresses the potential plumage of the male. One such female
known to have had all the ovary removed never assumed the summer
(eclipse) plumage of the drake. On the other hand, another female
developed first the nuptial plumage, but this was replaced by the
summer coat "of the male of this variety." Again, in the summers *A
1914 and 1915 the change to the eclipse plumage was followed in the
autumn by a return to the nuptial plumage.

82 THE GENETIC AND THE OPERATIVE EVIDENCE

How can we explain the apparent discrepancy of Goodale's results?
In one case, the nuptial plumage was molted to nuptial plumage; in
the other case an eclipse plumage appeared at the breeding-season.
Goodale regards the latter case as a more perfect approach to the male
than the former, but this view undoubtedly offers serious theoretical
difficulties. It seems to me possible to suppose that in those cases
where the summer plumage appeared there was in reality enough
ovarian tissue (or related tissue) left after the operation to produce an
effect at the normal season for such ovarian tissue to become most
active. It might then suffice to eclipse the male plumage sufficiently
to make it very similar to the eclipse of the normal male. At any rate,
on this basis we have a consistent explanation of the entire complex
of phenomena.

What bearing have these results relating to castration and trans-
plantation on the theory of sexual selection? Granting, of course, that
selection takes the materials as it finds them, there may still be restric-
tions imposed on the theory by the kind of material offered. For
instance, the development of the plumage of the cock is independent
of the condition of his testes. Hence, if the female selected the more
vigorous male, she would not necessarily obtain one more ornate than
his less vigorous rivals. If the taste of the hen has built up the plumage
of the cock, it has been carried out then independently of the vigor
resulting from the greater activity of the testis. In a word, the more
vigorous male is not necessarily the most highly colored one. Darwin
concedes that these two conditions, high color and vigor, must go
together to insure success, or at least that the most vigorous and there-
fore the most highly colored male will have more offspring. Wallace's
contention that the greater vigor of the male accounts for his greater
development of plumage gets scant support from the facts of castra-
tion. One might rather contend that the female must be more vigorous,
since she is obliged to suppress plumage that is allowed to run riot in
the male.

Wallace's argument in favor of natural selection holding down the
plumage in the female as a protection to her while nesting might
appear to fit the facts better were it not that the quest for an explana-
tion of the male's plumage is thereby abandoned. It should not be
forgotten in this connection that the nest is generally only partly
concealed, that bright color at rest need not be conspicuous, and that
the male, exposed as he is through a considerable part of the year, still
manages to maintain himself in about equal numbers with the female.
Suppose, however, for the sake of argument, that natural selection has
kept under the full possibilities of the female. The modus operandi
would be competition between the least adorned females, suppression
being brought about by the activity of the ovary; while the male is
left therefore to exhibit the full possibilities of the genetic complex of

RELATING TO SECONDARY SEXUAL CHARACTERS. 83

his race without restraint. The facts in the case are that the plumage
of the male is the direct result of his genetic composition; th<' female
has the same genetic composition (the sex-linked characters are duplex >,
but the ovary produces a substance that holds them in restraint. Put
in this way, there is nothing further to be explained, unless we insist
on finding an explanation as to how the species came to have its
genetic constitution. In other words, if we are not satisfied with the
statement as to the actual situation, we must explain it by a utili-
tarian appeal to a relation between the plumage and the world outside
of the individual or the species. To those who feel unsatisfied to leave
the case as it stands on a physiological basis, there is another hypo-
thetical means of escape. It may be assumed that the genetic factors
that are instrumental in producing the secondary sexual characters
have also other but unknown influences in the economy of the species,
color and ornamentation being by-products of these factors whose
utility in other directions accounts for their presence. Such a philos-
ophy has perhaps one redeeming feature, since it suggests the possibil-
ity of searching for other influences — influences that only incidentally
give the striking coloration and ornamentation of the males.

At first sight the absence of cock-feathering in the Sebright may
seem to furnish the occasion for such a quest. It might appear that
since only one or two genetic factor differences are responsible for the
"nuptial" plumage of the male, that this plumage may have originated
in one or two genetic changes. Such an argument is fallacious, how-
ever, for very many genetic factors may historically have been neces-
sary to build up the nuptial plumage of the male. The breeding
experiment shows no more than that one or two other factors have
appeared that counteract the effect of all that the others are capable of
producing; the experiment throws no light upon how many or how-
few these other factors may be. That the nuptial complex is still
present in the Sebright is evident after castration. Castration shows
only that the testes in the Sebright produce some material that keeps
down the effects of all the other factors combined. This conclusion,
it is true, somewhat simplifies the problem for those who appeal to
natural selection as suppressing in the female the feathering of the
cock, because it shows that this could have been accomplished by one
or two Mendelian factors that appeared of such a kind that they caused
the ovary to produce a substance antagonistic to the influences Doming
from the genetic complex of the species.

With this by way of provisional exposition, let us return to the ques-
tion as to whether the Sebright-game cross throws any other light on
the possibly useful character of the genetic factor or factors that
produce cock-feathering. It is obvious that the evidence givee us no
clue at all, for with the exception of the normal allelomorphs of the
dominant factor for hen-feathering, all the other factors are still

84 THE GENETIC AND THE OPERATIVE EVIDENCE

present in the Sebright. The normal allelomorph in question need
not have had any relation to the other complex; in fact, it seems not
to have any, because the castrated Sebright (with both normal allelo-
morphs replaced by genes for hen-feathering) still develops the charac-
teristic cock-feathering.

The outcome in the duck with its double male plumage is still more
puzzling when we attempt to analyze the situation in the light of the
selection theory. At the height of the breeding-season, when his testes
are enlarged and functioning actively, a substance is being produced
that leads to the eclipse of the nuptial plumage. If the male were
selected by his partner for his plumage, he would be chosen for a
plumage that develops in the absence of the functioning testes. If the
male is chosen because of his greater aggressiveness or "activity" or
"vitality" due to the development of his testes, the result would be to
select males that would probably develop a better eclipse plumage.
The case is interesting because it gives an opportunity to distinguish
between a plumage that develops under the influence of the sexual
organs and one that does not; and the latter is paradoxically the
nuptial plumage. It is true that the male might be selected for his
nuptial suit, and, theoretically at least, female choice might still be
made responsible for this plumage, but this merely shifts the problem,
for it leaves "unexplained" the appearance historically of the effect of
the activity of the testes in suppressing this plumage for a short time
after maturity. No doubt an attempt might be made to show that
natural selection comes in at this time of the year in giving a protective
color to the male, but so long as any evidence is lacking as to the need
of this protection the argument serves rather to further complicate an
already difficult situation.

Goodale has written to me that there is an account, in the Agricul-
tural Journal, Union of South Africa, iv, 1912, of the effects of the
removal of the ovary of the female ostrich. I have not been able to see
the account, but according to my informant such female individuals
assume the male secondary characters.

Of unusual interest in connection with the seasonal change of plum-
age in males of dimorphic species are Beebe's experiments with scarlet
tanagers and bobolinks. In both species the males in their nuptial
plumage are very different from the females. Full-plumaged males of
both species, at the height of their "vocal and physical condition,"
were confined in small cages. The supply of light was gradually cut
off and a slight increase of the amount of food was allowed them. The
birds became less active in consequence and increased in weight. ' ' The
time for the fall molt came and passed and not a single feather was
shed." The birds had skipped the autumn molt and remained in their
nuptial plumage. The song soon died away ; " the birds seldom uttered
even a chirp." From time to time a bird was gradually brought into

RELATING TO SECONDARY SEXUAL CHARACTERS. S.j

the light for a week or two and meal-worms were added to the diet.
This invariably resulted in a full resumption of song.

"I found that a sudden alteration in temperature — either lower or higher —
wrought a radical change in the physical metabolism of the bird*. They
would stop feeding almost altogether, and one tanager lost weight rapidly.
A few feathers on the neck fell out, and in the course of some two weeks this
bird moulted almost every feather and came strongly into his normal winter
plumage of olive green. The metabolism set up by the change in temperature,
in its intent and rapidity, seems comparable only to the growth of a deer's
antlers.

"Early in the following spring individual tanagers and bobolinks were grad-
ually brought under normal conditions and activities, with quick result ; just
as the wild birds in their winter haunts in South America were at that time
shedding their winter garb and assuming the most brilliant hues of summer, so
the birds under my observation also moulted into the colors appropriate to the
season. The old scarlet and black feathers fell from the tanagers and were
replaced by others of the same color; from buff, cream, and black, the bobo-
links moulted into buff, cream, and black! There was no exception; the moult
was from nuptial to nuptial, not from nuptial to winter plumage. The dull
colors of the winter season had been skipped."

How are these results to be interpreted? Obviously the environ-
ment prevented the autumn molting; hence the birds necessarily
retained their nuptial plumage. But is this the whole story? Did they
not also remain sexually active with their testes producing sperm as in
the mating season? In other words, if feathers had been plucked from
them, would not the new feathers have been like those already present?
Despite the author's statement that not a single feather was molted,
is it not likely that occasionally a feather must have been accidentally
lost. If even one had been lost and an eclipse feather had replaced it,
the effect would not have escaped so keen an observer as Dr. Beebe.
It seems to me not unlikely that an occasional feather may have been
lost and replaced by a nuptial one. If so, then the results are most
probably interpreted as due to the birds having remained sexually
active. This condition suppressed the autumn molt, and at the same
time would cause any single feather lost to be like those still present
In support of such a conclusion I can appeal to Beebe's statement that
after a week in the light a full resumption of the song took place. It
is unlikely that sexual maturity would be attained in so short a time
unless the birds were already in the condition of sexual vigor. Perhaps
one can appeal also to Beebe's other statement, viz, that after a sudden
change in temperature, followed by a changed metabolism and km of
weight, the birds molted and assumed the eclipse (winter) plumage.
Here I should interpret the facts cited possibly to mean that the males
lost their sexual activity and in consequence developed the eclipse

plumage.

Until further information is obtained judgment must be suspended.
If, as Beebe's statements strongly suggest, the external conditions.

86 THE GENETIC AND THE OPERATIVE EVIDENCE

acting directly on the "metabolism," cause the changes observed, then
the experiments mean that environmental conditions affect directly
the development of the nuptial and the eclipse plumage; but if, as I
suggest here, the effects observed are due directly to the environmental
action through its effects on the testes, then the results fall more nearly
into line with those of Goodale on ducks, etc.

C. Evidence from Amphibia.

The thumbs of frogs enlarge at the breeding-season and shrink
afterwards. The enlarged thumb is used by the male in clasping the
female during copulation, and the rough papillae that appear over its
surface at this time may also help to anchor the male in his precarious
position on the back of the female. Since the pads and their papillae
are used in copulation, they belong rather in the class of accessory
organs of reproduction than in the class of secondary sexual characters.
Smith and Schuster state for Rana fusca that the testes are at their
smallest size in March and April after the breeding-season. From that
time until August they steadily increase in size and reach their maxi-
mum size in September. From September to March they are inactive
and full size, until the shedding of the sperm in March brings them
soon afterward to their lowest point again. It is to be noted that the
increase after March is associated with the increase in division rate of
the spermatogonia. The ripening of the sperm is finished in October.

The thumb-pads with their pigmented papillae are "cast off"
immediately after the breeding-season, the thumb remaining smooth
from May to September. The reduction of the pad is usually due to
the reduction of the glands and the disappearance of the papillae.
Smith and Schuster state: "During the months when the most active
growth of the testis is taking place the thumb-pads remain inactive
and smooth." The implication, apparently, is that one ought to expect
the growth in the thumb to take place when the germ-cells are most
actively dividing, if its growth is connected with their activity; but
there are no grounds for such expectations, because the influence of the
gonad may have nothing to do with the division rate of the germ-cells,
but rather with interstitial or other cells, and even here less with their
division rate than with their period of greater secretive activity.

"In August and September the epidermal papillae begin to be obvious, and
from this time onwards until about February a continuous increase of the
epidermal papillae and pigmentation occurs. During the greater part of this
time, when the thumb-pads are attaining their characteristic rough and pig-
mented appearance, the testes remain inactive and unchanged — a fact which
has been too readily overlooked by writers on the correlation of the primary
and secondary sexual characters."

Nussbaum (1909) and later Meisenheimer (1911) found that after
castration the thumb-pads disappear. Smith confirms this report in
all essential respects, although in certain details concerning the papillae

RELATING TO SECONDARY SEXUAL CHARACTERS. S?

he does not agree with the two former observers. His result s inon • t hat
castration at the breeding-season is rapidly followed by the loet of the
outer papillated layer of the thumb-pads, but castration at any other
season does not have "any marked effect," the papillae remaining for
5 months and more in the same condition as at the time of castration.
The essential point here, however, is that the excessive and even
special development at the breeding-season does not take place ner is
again assumed (apparently), if castration has taken place at boom
other time of the year.

Smith and Schuster's attempts to transplant the testes into other
males or females were unsuccessful, as the testes degenerate after a
time. Auto-transplantation of the testes were more successful.

Removal of the ovary had no effect on the thumbs of the female,
and even the injection of testes extracts into such females did not
cause them to develop pads. Nussbaum and Meisenheimer had
found that transplantation of pieces of the testes, and even injection
of testes extract, into castrated frogs caused an enlargement of the
thumb-pads. Smith shows that this conclusion rests on uncritical
evidence. At any rate, his own more carefully planned experiments
extending over the year show that the results obtained by Nussbaum
and by Meisenheimer may be accounted for on other grounds than the
effect of the injection or implantation.

The following statement by Smith is not without interest, since it
bears directly on an important question as to how internal secretions
may produce their effects.

"The deduction, therefore, which has been unduly based on Nussl mum's
experiments, that the testis of the frog contains an internal secretion, which,
on being circulated in the blood, calls for the development of the secondary
sexual characters, either with or without the mediation of the nervous system,

is without experimental foundation The fact that the developmental

cycle of the thumb depends for its normal course on the presence of normal
living testicular tissue can be equally well explained on the theory that the
testicular cells enter into a chain of metabolic processes in the body which do
not pursue their normal course in the absence of the testicular cells. This
disturbance of the normal metabolic processes of the body, resulting in the
failure of the metabolic organs of the body to give rise to their normal prod-
ucts in normal quantities, may have the result of inhibiting the further devel-
opment of the secondary sexual characters. The development of these latter
characters may depend, therefore, not directly on the action of an internal
secretion or hormone derived from the gonad, but on the elaboration of other
products in other organs of the body in their due proportions. These sub-
stances may be tentatively called 'sexual formative substances/ but we have
no reason for supposing that they are entirely devoted to sexual or reproduc-
tive purposes, and that they take no part in the ordinary metabolic processes
of the body."

The arbitrary distinctions that Smith here sets up do not seem to
me to contribute anything to the situation, and in fact in the end it
amounts to practically the same thing whether the hormone m

88 THE GENETIC AND THE OPERATIVE EVIDENCE

directly on some specific part of the body or whether in doing so it
acts on other parts as well. While it is more or less customary to
limit the term " hormone" to substances that do produce specific
effects in a particular organ, no one would, I suppose, deny that
a substance was acting as a hormone if at the same time it acted on
other parts of the body also, or even if its immediate action were on
some part and its ultimate action on another part of the animal.
Moreover, there is nothing in the evidence appealed to by Smith that
supports one rather than the other contention. It is not apparent that
the simpler idea of hormone action may not still apply. Failure to
implant the testes in castrated male or female, and failure of injections
to produce the results sought for, may mean no more than that the
experimenter failed to fulfill some one of the conditions present in the
normal frog at the breeding-season. Granting that the results recorded
by Nussbaum and Meisenheimer are open to the serious objections,
pointed out by Smith and Schuster, the facts recorded by all three
writers indicate that the maximum development of the pad takes
place when the testes are at their greatest development and that the
pad suddenly decreases if at this time the testes are removed. It
would seem to follow that since the swelling is connected with the
presence of a certain condition of the testes, its enlargement is to be
referred directly to the latter, and the case comes under the general
category of "secondary sexual differences," depending on the gonad.

The secondary sexual characters of Triton cristatus can not, as can
those of the frog, be supposed to be mechanically useful in mating, but
seem to be comparable in every respect with the secondary sexual
ornaments of higher animals. The work of Bresca has shown that their
development is under the influence of the testes. The most important
secondary sexual characters of the male are the dorsal comb and the
white stripes of the tail. The comb extends along the dorsal surface
of the body and of the tail (with a slight dip in the pelvic region). It
is fully developed during the breeding-season, when it reaches a height
of 1.5 cm. In winter it is only 0.66 mm. high, or even less. The white
stripes also are fully developed in the breeding-season. They extend
on each side from the cloaca to the end of the tail. In the female the
white stripe is sometimes faintly seen. The angles of the tail and of the
cloaca thickening are black-brown or black. The belly of the male is
bright orange or "Ziegel rot"; that of the female sulphur-yellow or
orange, but the difference is not constant. The upper surface of the
head of the male is marbled, especially during the breeding-season
almost disappearing during the rest of the year. Bresca found, when
the testes were removed from sexually mature males, that in the course
of a year all the important secondary sexual characters disappeared,
including the comb, the white tail stripes, and the marbling of the
upper surface. Removal of the ovaries did not affect the characters of

RELATING TO SECONDARY SEXUAL CHARACTERS. 89

the female. The black lower corner of the tail in the male is not
changed by castration.

When the skin along the middle line of the back of the female is
transplanted upon the back of a normal male (in place of his own cob
the transplanted tissue develops into a comb. In other words, under
the influence of the testis, the dorsal mid-line tissues of the female
change into those characteristic of the male. When pieces of skin of
a male with the white tail stripes are grafted on the side of the tail of
another male, the stripe remains, but when grafted similarly on a
female the stripe slowly disappears. The result shows that its presence
depends on the testis.

A remarkably clear case of hermaphroditism in amphibians was
found by V. la Vallette St. George. He found an individual of Triton
tceniatus that was outwardly a male with well-formed dorsal comb.
In the interior were two large testes in normal position and just
lateral to these on each side a large ovary. Sections showed ripe sperm
in the testes and typical ova in the ovary. Sperm-ducts were present,
but no oviducts. The presence of the testes will, of course, account for
the development of the secondary sexual characters of the male.

Other cases amongst the Anura have been recorded by Loisel and by
Marshall, Spengel, and Knappe. In the early stages of the gonad in
frogs there appears to be an hermaphroditic stage in which egg mother-
cells and sperm mother-cells are both present, at least in those individ-
uals that will later become males (Kusakowitsch).

The normal hermaphroditism of certain fish (Serranus) and its rare
occurrence in other species (recorded by Shattuck and Seligmann)
need not be recorded here.1

D. Evidence from Crustaceans.

In the Crustacea the secondary sexual characters are not marked,
except in a few cases. In the amphipods, Holmes has shown direct
contact plays the chief role in mating, and in the crayfish it has been
shown by Dearborn, Andrews, and Pearse that sex recognition is largely
tactile. Chidester also has shown this in crayfish. Even in crabs,
and especially those living on land which have well-developed eyes and
good vision, secondary sexual differences are as a rule slight and the
mating instincts simple. On the other hand, the enormous chela of the
male of the fiddler is supposed to be a secondary sexual difference
(mainly because no other use for it has been found). Pearse suggests
that the waving of this claw by the male is used as a sex signal . al t h. nigh
he is disinclined to accept Alcock's view that it has become "con-
spicuous and beautiful in order to attract the female."

The most remarkable case known of a change in the secondly
sexual characters of one sex into those of the other was discovered by

1 See the latter also for references to Lacrrtilia and Chclonia.

90 THE GENETIC AND THE OPERATIVE EVIDENCE

Giard in 1886. As a result of infection by parasitic Crustacea (e. g.,
Sacculina), the male crab develops the secondary sexual characters of
the female. It has been generally supposed, following Giard, that
this result is due to the destruction of the testes of the male by the roots
of the parasite that invades the spaces between the organs of the
host, and, in the case of the testis, ultimately brings about its partial
or complete destruction. Not unnaturally the results here were sup-
posed to be parallel to those of castration in vertebrates, and received
in fact the name of "parasitic castration." More recently Geoffrey
Smith has studied this phenomenon in the crab Inachus, infected by
the parasite Sacculina, and has reached the conclusion that the change
is not due to injury or to destruction of the testes, but to a change in
the metabolism of the crab brought about by the parasite.

Taking Geoffrey Smith's case of Inachus-Sacculina as typical, the
changes brought about are as follows : The parasites attach themselves
to the young crabs before the external secondary sexual differences have
appeared. In the females, the effect is to cause them to develop pre-
maturely the distinctively female characters. In the male, on the
other hand, the narrow abdomen of the male changes after a molt into
the broad abdomen of the female, which also develops ovigerous
appendages on its ventral surface like those of the female in every
detail. The larger claw of the male changes into that of the female,
which is different in form as well as in size. Some years ago I ven-
tured to raise the question as to whether these effects on the male
might not be interpreted as retention of the juvenile characters rather
than development of the female characters in the male. This might
appear more especially the case in the somewhat more juvenile shape
of the anterior abdominal appendages and possibly also in the shape
of the broader abdomen; but Smith has later shown that the results
can not be interpreted as juvenile, for when the changed organs are
examined in detail they are found to differ from the same organs in
the juvenile condition, and to be identical with those of the adult
female. I think, therefore, that we must accept this interpretation of
Giard and of Smith as correct. But Smith goes further and believes
that the effects may be carried so far that eggs develop in the old
testes; in other words, that the testis changes to an ovary. It seems
to me that the evidence to support this last point should be much
stronger than that advanced by Smith before we can accept this
interpretation, for we lack the essential control for this evidence. In
only a single case were eggs found — in the testis of a male that had been
infected, but from which the parasite had fallen off, and which was
presumably recovering from the effects of its presence. Now, it is
known that in the testes of some male animals a few eggs may occa-
sionally be found where there is no suspicion that the animal has
changed its sex. In some Crustacea, in scorpions, and in insects,
isolated instances of this kind have been found. Abnormal division

RELATING TO SECONDARY SEXUAL CHARACTERS. 91

of a spermatogonial cell, of such a kind that both sex BfaromOMn
(in the case of insects at least) got into the same cell might be expected
to cause such a cell to become, even in the male, an egg-cell rather
than a sperm-cell. The degenerative changes of the testes in the
hermit crab caused by the parasite might be imagined to favor such
abnormal division with its consequences. More significant, however,
is the fact that the parasite causes the absorption of the ovary when
it infects a young female, so that even all its eggs disappear. In other
words, the parasite is as injurious to the peculiarly female organ as it
is to the testis. Why, then, one can not but ask, should an influence
that causes such effects on the ovary first change a male into a female
so long as it is present and then when the parasite has disappeared leave
an influence behind of a kind that causes the ovary to develop— an
organ which the parasite destroys when the parasite is present? Is
it not more probable that only the secondary sexual organs were
changed, without change in sex, the single case of eggs observed being
caused in another way? This point can only be settled by direct exper-
imentation either by removal of the testis, by injuring it, or by inject i< m.
grafting, or feeding experiments. The extent of the testis and its posit ion
make it impossible to remove it by an operation, as I have found after
repeated attempts. It seemed easier to destroy it by radium. This I
have tried to do, using very powerful tubes, treating the crab (fiddler
crabs) for several hours. The crabs had had one claw removed — the
enormously large one — and were kept until the next molt, that occurred
from a week to six weeks later. In none of the cases was any change
produced. The large claw of the male regenerated, of course, not full
size after only one molt, but after several nearly full size and always
with the peculiarities of the male crab. The abdomen and the appen-
dages were not changed. Whether the significant cells of the testes,
if there are such cells apart from the germ-cells, were destroyed, can
not be told, for as yet the histological examination of the material has
not been made. Until a successful operation has been done. I think
we must hesitate to accept Smith's argument, although based as it is
on a series of interesting observations. His speculation is as follows.

"The reason why Sacculina causes the assumption of the adult feinal-
in Inachus is found in the facts: (1) that the roots of Sacculina elaboral
yolk-substance from the blood of Inachus of a similar nature to that which is
elaborated in the ovaries of an adult Inachus; (2) that in order to elaborate this
yolk-substance the roots take up from the blood of Inachus the female sexual
formation substance, which is the necessary material for forming tin- yolk
that the female sexual formative substance being absorbed by the Sacculina
roots is regenerated in excess; (4) that the presence of the female formative
substance continually circulating in large quantities in die body-fluids oJ the
infected crabs causes the production of adult female secondary sexual charac-
ters, and, when the parasite dies, of yolk-containmp eggs.

In brief, the evidence consists in showing that in the parasite a yolk-
substance appears, which Smith says comes from the blood of the crab

92 THE GENETIC AND THE OPERATIVE EVIDENCE

that produces it under the influence of the parasite. Incidentally, as
it were, this is said to be the same yolk-substance (but no sufficient
evidence that it is the same is given) that the egg stores up inside itself,
and it is assumed that it is a formative substance that causes the cell
that gets it (or contains it or secretes it — details are wanting) to
become an egg-cell. It is the excess of this substance produced by the
male crab, while still a male, under the influence of the parasite, that
affects the abdomen and its appendages in such a way that they assume
the female condition. There are too many assumptions in the argu-
ment, some of which are scarcely of a kind that our knowledge of
development, incomplete as it is, can allow us to accept without more
direct evidence in their support, to make this view very plausible.
Until better evidence is forthcoming, I fail to be convinced by Smith's
interpretation of his facts.

Into Smith's and Robson's interesting observations on the blood of
crabs, described in Smith's later paper (part 7, 1911), it is not necessary
to enter here, since the evidence taken as a whole offers little further in
support of his view than had been already assumed. The argument on
page 263 should not, however, pass unchallenged. Smith says:

"It is clear that the old and familiar idea of an internal secretion produced
by the gonad being the stimulus for the development of the secondary sexual
character could not be applied here, since at the time that the alterations in
the secondary sexual characters take place no ovary is present to give rise
to the required stimulus. It is suggested, therefore, that in some way the
stimulus must reside in the roots of the Sacculina," etc.

The argument seems to imply that, since the secondary sexual
characters of the female can not be produced by an ovary in the infected
male, therefore the Sacculina must take the place of the ovary. But
why make such a supposition, for if the testes simply keep down the
development of the female characters, as Giard supposes, there is no
need either for an ovary or for a Sacculina to develop them. One might
as well argue that since the cock does not develop the secondary sexual
characters of the hen that an ovary is essential for their development —
which is true, but not in the sense implied.

Stamati (1888) states that he attempted to remove the testes of
adult crayfish and apparently succeeded, but since no effects are
expected until after a molt occurs (that may not take place for two
years or more), no results were obtained. Injections of the gonads
with an acid failed, since the animals died.

E. Evidence from Insects.

In 1899 Oudemans succeeded in finding a method of removing the
testes and ovaries from caterpillars, using a dimorphic species, Ocneria
dispar, the gipsy moth. The results were negative; none of the secon-
dary sexual characters of the male or female moths or the accessory
organs of copulation were in the least affected by the operation. The
castrated male copulated as readily with the female as did the normal

RELATING TO SECONDARY SEXUAL CHABACTBR8. '.).;

male, while the spayed females also behaved as normal individuals of
that sex behave. Kellogg, in 1904, repeated the same operation in the
silkworm moth on a small scale with the same results. Kopec and
Meisenheimer, in 1909, repeated in a more detailed way Oudemai
work. A further important addition was made by Kopec and by
Meisenheimer. They transplanted ovaries into a castrated male and
testes into a spayed female. Neither gonad produced any effect on
the characters of the other sex. It is interesting to note that the te
underwent their normal development in the body of a spayed female,
and even in one with the ovaries present, and that the ovary als< > under-
went normal development in the body of the male. In other words,
there is no intolerance of the tissue of one sex to the gonad of the other.
This result is all the more unexpected, because other observations have
shown that the color of the blood, and its chemical properties, is quite
different in the male and female moths of certain species.

In the case of moths, therefore, if these cases be regarded as typical,
the situation from the point of view of sexual selection is much simpler
than in birds in the sense that the secondary sexual characters are
directly the product of the genetic constituents of all the cells, and not
influenced indirectly by the secretions from the testes or the ovaries.
Sexual selection, therefore, if it is an agent in the evolution of the dif-
ferences between males and females, has acted on the genetic complex
to produce these effects on either sex without the result being involved
in the condition of the ovary or the testes.

Regen castrated crickets, Gryllus campestris, in the Larval stages and
found no effects on the adult structures. The castrated males chirped
like normal males and mated with the females. Spayed females v,
like normal females; they bored holes in the ground, but hid no eggs
in them, of course, as the ovary had been completely removed.

The only genetic evidence in the group of insects, outside of the
vinegar fly, relating to the secondary sexual inheritance of the second-
ary sexual characters is the following important experiments made by
Foot and Strobell :

The male of one of the bugs, Euchistas variolarius, lias a black spot
on the end of the abdomen — a spot that is not present in the female.
Foot and Strobell crossed a female of this species to another hug,
E. servus, that lacks the spot in both sexes. The daughters had no
spot, the sons a faint spot less developed than in voriolarius. Tfo
inbred gave (in F2) 249 females without a spot, 107 males with a spot,
and 84 males without a spot. The results are explieable on the view
that a single dominant Mendelian factor, not-sex-linked, causes the
spot in the males, but the presence of the gene in the female produ
no effect. The effect, therefore, is Bex-limited, I <.. its expression is
determined by the rest of the complex male or female.

The very important breeding experiments carried oul by Gold-
schmidt on varieties of the gipsy moth should be referred to in this

94 THE GENETIC AND THE OPERATIVE EVIDENCE

connection, but as I have recently reviewed these results in the paper
on gynandromorphs written in collaboration with C. B. Bridges,1 I
need only refer to that account here.

[Note added April 21, 1919.]

Shortly after the preceding paper was finished a theses by A. P6zard on the
secondary sexual characters of birds reached me. In it the author gives an
account of a number of experiments that he has made with poultry and with
pheasants. His description of the changes that take place after castration
are more exact and more detailed than any other so far recorded ; but in general
the results obtained by Pezard, through castration, are the same as those
that had been obtained by others. Castration of 4 male silver pheasants are
reported. No change in the plumage results, although the changes that take
place in the comb and wattles are the same in kind as those observed in fowls.
The sexual instincts and peculiarities of the voice and their belligerency are also
lost. Similarly 4 golden pheasants that were operated on gave the same results.

Three pheasants with mixed plumage (Phasianus colchicus) were examined.
Their testes proved, on histological examination, to be imperfectly developed.
It is not evident what relation existed between the facts and the mixed plum-
age. The suggestions made by Pezard seem inadequate to cover the cases.

Testicular tissue transplanted into castrated cocks whose comb, wattles,
etc., had undergone retrogressive changes brought about a return to the normal
conditions after an interval during which the implanted nodules had begun
to regenerate.

Testicular extract from the cryptorchid testes of swine was injected into
castrated cocks. In one case this resulted in a rapid growth in size of the
comb, which, after 2 months, had reached its full size. Cessation of the injec-
tions led immediately to a cessation of growth. Before injection the bird
exhibited the pacifistic characteristics of the capon, but the injections brought
out little by little the aggressive behavior of the normal male. The voice
reappeared and "nous assistons a une veritable crise de puberte."

A histological study of the testes of the fowl and of pheasants showed that
much connective tissue is characteristic of young birds. In the adult cock,
and during the mating season of the pheasant, the connective tissue becomes
largely crowded out by the enlargement of the tubules. Pezard concludes
that the "interstitial" cells in birds have nothing to do with the secondary
sexual characters, but that these come rather under the influence of the ger-
minal cycle of cells of the testes. The submergence of the connective-tissue
cells of pheasants during the breeding-season and their reappearance during
the rest of the year might appear to have some relation to the facts that I
have recently described in Sebrights, but as the nuptial plumage of the male
remains the same throughout the year we can not ascribe any direct influence
to this tissue. Nevertheless, the different tissues of the testes in birds that
show seasonal dimorphism of plumage should be carefully examined.

Pezard made a few observations on hens whose ovary had been removed.
His results are in accord with those of Goodale, except that he thinks that the
ovary has no influence on the erectile organs (comb, etc.) which acquire in the
spayed bird the same length as that of the normal female.

Two hens showing male characteristics and a pheasant similarly affected
are described. In all three cases an examination of the ovary was found to
be undeveloped or abnormal.

1 Carnegie Inst. Wash. Pub. No. 278, 1918.

RELATING TO SECONDARY SEXUAL CHARACTERS. 95

PART IV.

SUMMARY AND CONCLUSIONS.

1. The two principal results obtained were: (a) that cast ration of
hen-feathered Sebright males causes them to develop the full plum,
characteristic of the cock-bird; (6) that complete hen-feathering is due
to two dominant Mendelian genes.

2. A striking change takes place when the Sebright male is cad rated
(plate 1, figs. 3, 4; plate 3, fig. 1). The new feathers on the upper
surface of the head, neck, back, wings, rump, and tail-coverts assume
a different color and distribution of their pigment; they take on a new
shape, and in those regions where in the cock the barbules are absent
from a part of the margin of the feather, the same absence occurs in t he
castrated birds. Such feathers are present on the neck, back, wing-
bow, and rump. The transition is shown in the figures in plate 6,
where for comparison one of the old and one of the new feathers lie
side by side. The tail-coverts in the hen-feathered bird are short, and
like those in the hen do not cover the true tail. After castration they
become excessively long — longer, in fact, than in many cocks — and
cover the true tail feathers. The tail feathers themselves, moreover,
become increased in length, as do the posterior row of feathers of the
wing-coverts. On the breast and sides the change is less marked. The
castrated Sebright loses his erect carriage, but how far this is due to the
changes in his plumage and how far is real (as a result of a new balance
due possibly to the lengthening tail and its coverts) I can not decide.

3. While castration causes the hen-feathered male to make addi-
tions in color, length, and size of many feathers, it causes at the same
time the other retrogressive changes characteristic of the capon (a
castrated cock-feathered bird); the comb and wattles shrink and
become pale, the birds almost cease crowing, and become timid. They
do not make much effort to mate with the hens, but when they do they
show the usual copulatory reactions.

4. If feathers are removed at the time of castration, the new feathers
show the full effect of the removal of the testes, although they must
have begun to develop immediately afterward. It is suggested that
by means of this delicate test the time relations of the internal secre-
tion can be profitably studied.

5. Feathers that may have started their development at the time
of the operation show the old influence at the tip of the feat!
(plate 10) and the new one in the rest of the feather. The change is
abrupt, although the transition is perfect.

6. Incomplete castration of the hen-feathered male leads to smaller
changes in the same direction than those following complete east rat km.

Where such small pieces of the testis were left that complete ooek-
feathering followed, the bird slowly changed back to hen-feat hering

96 THE GENETIC AND THE OPERATIVE EVIDENCE

as the testes began to regenerate. When the regenerated pieces were
removed the bird became cock-feathered again.

7. One Sebright male whose testes appear to have been completely
removed did not change the character of the plumage. No testes were
found on autopsy. It is suggested that some other endocrine organs
have taken over the function of the testes, but as yet none such can be
indicated.

8. In one case an old hen-feathered (Fi) male began to change over
to cock-feathering. It was found that his testes had dwindled (prob-
ably through disease) to very small size (10 by 5 mm.).

9. The Fi male of the cross between the Sebright and game is also
hen-feathered (plate 2, fig. 1). After castration he becomes cock-
feathered (plate 2, fig. 4) and shows thereby the genetic type of the
heterozygous cock-feathered class in which his hen belongs. The
change in this male is even more strking than that in the Sebright.
The change in the individual feathers is shown in plate 7, figs. 1 and la.

10. Three types of F2 hen-feathered castrated males are shown in
plate 2, figure 3, and plate 3, figure 3 and figure 4. The first was
a dark bird that changed to a lighter red above. The third a gray
bird that became bright red ; the second was a light yellow that became
deep yellow, etc. The class of hens to which such males belong, as
cock-feathered birds, can thus be found out by castration. In this way
the F2, and back-cross, hen-feathered cocks can be classified with the
corresponding F2 cock-feathered males.

11. In the F2 generation, made up of birds from the direct and
reciprocal crosses taken together, there were 29 hen-feathered and
26 cock-feathered males. In the back-cross (Fi hen by game male)
the classes were 2 and 7. The results seem in better accord with
the assumption that two factors are present in the Sebright that stand
for hen-feathering; that either alone will give hen-feathered birds
(intermediate type?) , but that both together give the extreme type of
hen-feathering seen in the Sebright.

12. The difference in color in the two races (Sebright and Black
Breasted Game bantams) is very great. The former have almost
uniformly laced feathers, while the latter has the varied plumage of
the jungle-fowl. The game is strongly dimorphic in color and color-
pattern ; the Sebright has the same type of coloration and pattern both
in the male and female, but this is deceptive, as castration shows,
because the castrated male is as strikingly different from the normal
Sebright female as is the cock of other birds from the hen. The
resemblance of male and female in this race is due to the suppression
of the true male plumage by something produced in the testes. There-
fore the heredity of dimorphism resolves itself here into the problem
of the heredity of hen-feathering. That the female Sebright has the
same genetic factors as the male is shown by the fact that she trans-

RELATING TO SECONDARY SEXUAL CHAHAi I 1 .1 «.»7

mils hen-feathering in the same way as docs the male ami also by the
fact, as Darwin pointed out, that an old female Sebright whoee ovaries
had degenerated developed not the hen-feathered plumage of her own
cock, but cock-feathered plumage like that of most male poultry.

13. The color of the Fi birds is shown in plate 2, figs. 1 and _'. In
general, the feathers are stippled, black and light yellow being the two
most conspicuous ingredients. Since hen-feathering dominates, the
dimorphism is absent, or at least is so slight as to not attract attention
— little more, in fact, than in the Sebright race. Thecarriageof the male
is like that of the Sebright male. The Fi male and female are alike in
the direct cross and the reciprocal, or at least no conspicuous differ-
ence is found between the two classes of hens, indicating that no imp >r-
tant sex-linked factors are involved in the cross.

14. The F2 birds show a great variety of color and pattern, but those
obtained can be approximately grouped into 16 classes. The clasf
are, however, admittedly not uniform, indicating minor factors not
here reckoned with. The classification of the hens is easiest ; the F2
hen-feathered males can then in many cases be referred to the proper
classes; the F2 cock-feathered males can not be accurately classified
with their corresponding hens, except in the case of those that resemble
the two Pi males, the Fi male, and those that castration experiments
of the hen-feathered males have shown to belong to certain hen t y:

15. Despite the admitted difficulties of classification, it is suggested
that three factor-pairs of differences will cover the main color clac
seen in the F2 and in the back-cross. One or two of these seem to be
incompletely dominant, since the Fi birds are not like either parent in
any single character, nor are they like the wild type in so far as this is
represented by the game.

16. A histological examination of the testis of the male Sebright by
Boring and Morgan has shown that it contains cells like those present
in the ovary of all breeds of poultry. These cells are called luteal cells
by Pearl and Boring, from their resemblance to the cells of that name
found in the corpora lutea of mammals. In the mammals similar cells
are supposed to produce internal secretions that act as hormones.
Their function in the female bird is unknown, but the fact that after
the removal of the ovary the female develops the secondary sexual
plumage of the male suggests that some secretion from these sells
performs this function. Their occurrence in the male Sebright and their
complete absence, or paucity, in the males of other races supports
strongly the view that these cells are concerned with the suppression
of the secondary sexual plumage.

17. While in mammals the interstitial cells have been supposed to
produce an internal secretion that causes the development of some of
the secondary sexual characters of the male, and the fuller elaboration
of others, in birds no such connection exists, if we except th.

98 THE GENETIC AND THE OPERATIVE EVIDENCE

of the Sebright. Castration of ordinary males does not affect deleter-
iously the secondary sexual plumage (although it does the comb, be-
havior, etc.), in fact may even enhance their effects. But, while in
the mammal a secretion is necessary for the full development of the
secondary sexual characters, in the Sebright a secretion inhibits certain
of them. What element in the ordinary bird and in the Sebright
causes the full development of the comb, wattles, sexual behavior, etc.,
is not known. Possibly it is the sexual elements themselves, but
possibly it is a secondary influence of the luteal cells producing a con-
trary effect on these parts from its effects on the feathers ; but possibly
more than one kind of secretory cell is present in the testis of the cock.

18. The causes of the development of the secondary sexual characters
are seen to be of such diverse physiological kinds that one may well
hesitate to apply the same explanation as to their evolution. In fact,
it is pointed out that several of the theories that have been suggested
run counter to the conditions that bring about the development of the
secondary sexual characters.

19. An attempt is made to give a critical review of Darwin's theory
of sexual selection in the light of the modern genetic and operative
results on the secondary sexual characters of the vertebrates. It is
pointed out that far from extending the general theory in its applica-
tions, the modern work has shown in the first place that the underlying
conditions that call forth the development of the secondary sexual
differences are so diverse in the different groups of animals that it is
a priori very unlikely that this evolution can have been directed by the
same external agent, such as the choice of the female, for such an
assumption carries with it in several cases other implications concerning
the causes of the suppression of these same characters in the female
herself, etc. In the second place, it is pointed out that the problem of
the excessive development of certain characters in the male whose
genes are present in both sexes no longer oppresses us as it did Darwin,
for it has been shown both by the genetic and by the operative work
that a single factorial difference may be at the root of exceedingly
great differences in the individual. Such results, while they admittedly
do not in most cases tell us that the differences involved have arisen
at a single progressive step, show us nevertheless that such differences
may depend on very simple initial differences, and if so, the entire
problem becomes enormously simplified. To Darwin the excessive
development of color and ornamentation appeared due to a long, slow
process of evolution laboriously brought about by the female through
selection of those males a little more ornamented than their fellows.
To-day we have found out that in many cases the genetic composition
of a male with such ornamentation and of a female without it may be
almost identical, except that the genes in one chromosome are duplex
in one sex and simplex in the other. Owing to this initial difference, the

RELATING TO SECONDARY SEXUAL CHARACTER 99

female in birds produces an internal secretion that Buppn in her

the ornamentation shown by the male, and in the mammal an internal
secretion produced by the testes causes the full development in the male
of the secondary sexual characters. If, as seems probable, th< re-

turns are some particular kind of substance, the condition thai led to
their appearance historically need not have been very complex; and
if not, the problem appears simplified. It still remains to give Borne
reasonable explanation as to why such substances should continue to
be produced if their products — the secondary sexual characters
possess no "beauty" for the female. Here more work is necessary,
but the modern genetic point of view may possibly give an important
clue. We are coming to realize more fully that the hereditary genes
generally have more than a single effect on the characters of the animal.
The secondary sexual characters may, then, be only by-products of
genes whose important function lies in some other direction. If, for
example, the secretion produced by the cells of the male have an
important influence on his output of energy, or strength, or activity,
their secondary influence over certain parts of the body would not call
for any further explanation on the modern view of natural selection.
If the secretions of the ovary of the female bird have some direct rela-
tion to her physiological processes that are important in the develop-
ment of the oviduct, for instance, it would be a matter of no importance
from an evolutionary point of view if that same secretion suppree
in her the development of the high color shown by the male.

D. H. HILL LIBRARY
North Ore1!"--} State College

BIBLIOGRAPHY.

Alcock, A., 1892. On the habits of Galasimus annulipes. An. Mag. Nat. Hist., VI.
Ancel, P., et Bouin, 1906. Sur l'effet des injections d'extrait de glande interstitielle du

testicule sur la croissance. Compt. Rend. Acad. Sc. Paris, CXL1I.
Arkell, T. R., 1912. Some data on the inheritance of horns in sheep. N. H. Agr. Exp.

Sta. Bull., 160.

, 1912. The nature of the inheritance in sheep. Science, n. s., XXXV.

, and C. B. Davenport, 1912. Horns in sheep as a typical sex-limited char-
acter. Science, n. s., XXXV.
Bateson, W., and E. R. Saunders, 1902. Experiments with poultry. Reports Evol.

Committee, I, Part II.
, E. R. Saunders, and R. C. Punnett, 1905. Experimental studies in the

physiology of heredity. Reports Evol. Committee, II.

, 1906. Poultry. Reports Evol. Committee, III.

, 1908. Experimental studies in the physiology of heredity. Reports Evol.

Committee, IV.
, and R. C. Punnett, 1911. The inheritance of the peculiar pigmentation of the

silky fowl. Jour, of Genetics, HI.

,1913. Mendel's principles of heredity. Cambridge.

Baur, E., 1914. Einfuhrrung in die experimentelle Vererbungslehre, Berlin.

Beebe, C. W., 1908. Preliminary report on an investigation of the seasonal changes of

color in birds. Amer. Nat., XL1I.
Bergendal, D., 1888. Uber abnorme Formen der ersten abdominalen Anhange bei

einigen Krebsweibschen. Bihang till K. Svenska Vet.-Akad. Hand-

lingar, XIV.
Bertkau, P., 1891. Beschreibung eines Arthropodenzwitters. Arch. f. Nat. ges., LVII.
Bond, C. J., 1913. Some points of genetic interest in regeneration of the testis after

experimental orchectomy in birds. Jour, of Gen., III.
, 1914. On a case of unilateral development of secondary male characters in a

pheasant, with remarks on the influence of hormones in the production of

secondary sexual characters. Jour, of Gen., III.
Boring, Alice M., 1912. The interstitial cells and the supposed internal secretion of the

chicken testis. Biol. Bull., XXIII.
, and Raymond Pearl, 1917. Sex studies, IX. Interstitial cells in the repro-
ductive organs of the chicken. Anat. Record, XIII.

, 1918. Sex studies, X. Hermaphrodite birds. Jour. Exp. Zool., XXV.

, and T. H. Morgan, 1918. Lutear cells and hen-feathering. Jour, of Gen-
eral Physiology, I.
Bourne, A. G., 1884. On certain abnormalities of the common frog. Quart. Jour. Micro.

Science, XXIV.
Brandt, A., 1889. Anatomisches und Allgemeines uber die sogenannte Hahnenfedrigkeit

und uber anderweitige Geschlechtsanomalien bei Vogeln. Zeits. f.

wiss. Zool., XLVI11.
Bresca, G., 1910. Experimentelle Untersuchungen uber die sekundaren Sexualcharaktere

der Tritonen. Arch. Entw.-mech., XXIX.
Castle, W. E., 1912. Are horns in sheep a sex-limited character? Science, n. s., XXXV.
Chidester, F. E., 1911. The mating habits of four species of the Brachyura. Biol.

Bull., XXI.
Cholodowsky, N., 1908. Uber den Hermaphroditismus bei Chermes-Arten. Zool.

Anz., XXV.

100

BIBLIOGRAPHY. 101

Cilleuls des, J., 1912. A propos du determinisme dee oaractJ

les oiseaux. Compt. Rend. Soc. Biol., LXX11L
Cunningham, J. T., 1908. Tlie heredity of secondary sexual oharacten in relation t..

hormones, a theory of the heredity of BOmatOgenie character! \'

Ent. Med., XXVI.
Darwin, Charles, 1868. Animals and plants under domestical ion.
Davenport, C. B., 1906. Inheritance in poultry. Carnegie ln-t. \\ 1, Pub. No

1908.
,1909. Inheritance of characteristics in domestic fowl. Oarnegfr In t \\

Pub. No. 121, 1909.
,1911. Another case of sex-limited heredity in poultry. Proo. Boc Exp

Biol., Med., IX.

, 1912. Sex-limited inheritance in poultry. Jour. Exp. Zoo]., XIII. 1912.

Dittrich, V. R., 1888. Uber Andrena praecox 9 und Aslacus fluviatilis 9 mit teihi

mannlichen Kennzeichen. Zeit. f. Ent., XIII.
Durham, F. M., and D. C. E. Marryat., 1908. Note on the inheritano) of sex in can;:

Reports Evol. Committee, IV.
Foges, Arthur, 1903. Zur Lehre von den secundiiren Geschlechtscharaktcnn. Arch. f.

Phys., 93.
Fowler, C. H., 1894. Notes on some specimens of antlers of the fallow deer, showing

continuous variation and the effects of total or part ial oast i at ion. I 'r< »c.

Zool. Soc. London.
Geddes, A. C, 1910. Abnormal bone growth in the absence of functioning te>ticlcs.

Proc. Roy. Soc. of Edinburgh, XXXI.
Giard, A., 1904. Comme la castration agit-elle sur les caracteres sexuel- secondairesT

Compt. Rend. Soc. Biol., LVI.
Gissler, C. F., 1881. Description of a hermaphroditic phyllopod crustacean (Eulmnt-

chipus). Am. Nat., XV.
Goodale, H. D., 1909. Sex and its relation to the barring factor in poultry. Science, X XIX.

, 1910. Some results of castration in ducks. Biol. Bull., XX.

, 1910. Breeding experiments with poultry. Proc. Soc. Exp. Biol. Med., VII.

, 1911. Studies in hybrid ducks. Journ. Exp. Zool., XX.

, 1913. Castration in relation to the secondary sexual characters of Brown

Leghorns. Am. Nat., XLVII.

, 1916. A feminized cockerel. Journ. Exp. Zool., XX.

, 1916. Gonodectomy. Carnegie Inst. Wash. Pub. No. 243.

, 1917. Crossing-over in the sex chromosome of the male fowl. Science, X I A 1

Gurnev, J. H., 1888. On the occasional assumption of the male plumage by female birds.

Ibis, VI.
Guyer, M. F., 1909. The spermatogenesis of the domestic guinea (Numida meleagrit

dom.) Anat. Anz., XXXIV.
, 1909. The spermatogenesis of the domestic chicken (Callus gallu.s dom).

Anat. Anz., XXXIV.

, 1909. On the sex of hybrid birds. Biol. Bull., XVI.

, 1909. La livre" e du plumage chez les hybrids de pints de el <h- poule. Butt.

du Museum, d'hist. Nat. No. 1.

, 1910. Accessory chromosome in man. Biol. Bull., XIX.

Guthrie, C. C, 1911. On evidence of some influence in offspring from engrafted ovarian

tissue. Science, XXXIII.
Hadley, P. B., 1913. Studies on inheritance in poultry. 1. The constitution of the White

Leghorn breed. Bull. Agr. Exp. Station of the Rhode [aland 81

CoUege, No. 155.
,1914. 11. The factor for the black pigmentation in the White Leghorn bn

Bull. Agr. Exp. Station of the Rhode [aland State College, No 161,
Halban, J., 1903. DieEntstehungderGi'.schlcchtscharaktcrc. Aroh. f . GynHlrologbi, I W.

102 BIBLIOGRAPHY.

Hanau, A., 1897. Versuche liber den Einfluss der Geschlechtsdriisen und die secundaren

Sexualcharactere. Arch. f. ges. Phys., LXV.
Harris, J. A., A. F. Blakeslee, and W. E. Kirkpatrick, 1918. The correlation between

egg production during various periods of the year in domestic fowl.

Genetics, 111.
Hegar, A., 1903. Korrelation der Keimdriisen und Geschlechtsbestimmung. Beitrage

zur Geburtshilfe und Gynakologie, VII.
Heinroth, O. Von, 1909. Ein lateral hermaphroditisch gefarbter Gimpel (Pyrrhula

pyrrhula europaea). Ges. Nat. Freunde zu Berlin.
Herbst, C., 1901. Formative Reize in der tierischen Ontogenese. Leipzig.
Herlitzka, A., 1899. Sul trapiantamento dei testicoli. Arch. f. Ent.-mec, IX.
Holmes, S. J., 1908. Phototaxis in fiddler crabs and its relation to theories of orientation.

Journ. Comp. Neurol., XVIII.
Hudson, W. H, 1892. The naturalist in La Plata. London.
Hull, J. E., 1918. Gynandry in arachnids. Jour, of Genetics, VII.
Hunter, John, 1780. Account of an extraordinary pheasant. Phil. Trans. Royal Soc.

London., LXX.
Hurst, C. C., 1905. Experiments with poultry. Reports Evol. Committee, II.
Jones, E. L., 1914. The Campine Club, 1914, Year Book (see also Farm Poultry, June,

1914, page 113).
Kellogg, V. L., 1904. Influence of the primary reproductive organs on the secondary

sexual characters. Jour. Exp. Zool., I.
Knappe, E., 1886. Das biddersche Organ. Morph. Jahrb., XL
Kopec, S., 1913. Nochmals tiber die Unabhangigkeit der Ausbildung sekundarer Ge-

schlechtscharaktere von den Gonaden bei Lepidopteren. Zool. Anz.,

XLI11.
Kurz, W., 1874. tlber androgyne Missbildung bei Cladoceren. Sitz. d. math.-nat.-wiss.

Klasse. d. Akadem. Wissensch., Wien., LX1X.
Lefevre, G. Sex-limited inheritance in poultry. Anat. Record, XI, p. 499, 1916-17.
Lillie, F. R., 1917. The free-martin; a study of the action of sex-hormones in the fcetal

life of cattle. Jour Exp. Zool., XXIII.
Lippincott, W. A., 1918. The case of the blue Andalusian. Amer. Nat., L1I.
Loewy, A., 1903. Neuere Untersuchungen zur Physiologie der Geschlechtsorgane. Ergebn.

d. Phys. II. Jahrg., 1 Abt., Biochemie.
Loisel, 1901. Grenouille femelle presentent toutes les caracteres sexuels secondaires

du male. Comptes Rend. Soc. Biol.
McEwen, R. S., The reactions to light and to gravity in Drosophila and its mutants.

Jour. Exp. Zool., XXV.
Marshall, A. M., 1884. On certain abnormal conditions of the reproductive organs of the

frog. Jour. Anat. and Physiol., VIII.
Marshall, F. H. A., 1910. The phsyiology of reproduction. London.

, 1910. Physiology of Reproduction. New York.

de Meijere, J. C. H, 1911. Uber getrennte Vererbung der Geschlechter. Arch. Rass.

Gesell. VIII.
Mayer, A. G., 1897. A new hypothesis of seasonal-dimorphism in Lepidoptera. Psyche,

VIII.

, 1900. On the mating instinct in moths. Psyche, IX.

Mazzetti, L., 1916. I Careteri sessuali secondari e le cellule interstitiali del testicolo.

Anat. Anz. XXXVIII.
Meisenheimer, J., 1909. Experimentelle Studien zur Soma-und Geschlechtsdifferen-

zierung. Jena.
, 1911. tlber die Wirkung von Hoden- und Ovarial-substanz auf die sekun-

daren Geschlechtsmerkmale des Frosches. Zool. Anz., XXXVIII.
Mobius, P. J., 1906. Uber die Wirkungen der Kastration. Halle.
Montgomery, T. H, Jr., 1909. Further studies on the activities of Araneads, II. Proc.

Acad. Nat. Sci., Philadelphia.

BIBLIOGRAPHY.

UYA

Morgan, T. H.

1907.
1909.

1913.
1914.

1914.
1915.

1917.
1918.
1918.
1918.
1918.

Montgomery, T. H., Jr., 1910. The significance of the oourtehip Mid secondary Mnal

characters of Araneads. Amer. Nat., XI. IV.
1905. An alternative interpretation of the origin of gynandrOIUOtph

insects. Science, XXI.

Experimental Zoology. New York.

A biological and cytological study of BOX determination in phyll'

and aphids. Journ. Expr. Zool., VII.

Heredity and sex. New York.

Mosaics and gynandromorphs in Drosophila. Proc. Boo, BioL Mid

Med., XI.

Sex-limited and sex-linked inheritance. Amer. Natur., XLYI1I.

Demonstration of the appearance after castration of ooek-featherinf in

a hen-feathered cockerel. Proc. Soc. Exp. Biol, and Med., XIII.

The theory of the gene. Amer. Natur., LI.

A critique of the theory of evolution. New York.

Concerning the mutation theory. Scientific Monthly, May.

Changes in factors through selection. Scientific Monthly, June.

Evolution by mutation. Scientific Monthly, July.
Morgan, T. H., and H. D. Goodale, 1912. Sex-linked inheritance in poultry. Annals

New York Acad. Science, XXII.
MtiLLER, R., 1907. Sexualbiologie. Berlin.

1 1908. Das Problem der sekundaren Geschlechtsmerkmale und die Ti.rzucht.

Myers, B. D., 1916. Histological changes in testes following vasectomy. Anat. Beoord

X, 1916.
Nicholls, F., 1731. An account of the hermaphrodite lobster. Phil. Trans. London,

XXXVI.
Nussbaum, M., 1905. Einfluss des Hodensekrets auf die Entwicklung der BrunstorgMM

des Landfrosches. Verhandl. d. Naturhist. Vereins. f. Rheinl.-W •

falen, LX1I.

Innere Sekretion und Nerveneinfluss. Ergebn. Anat. und Ent.-ges. , XV.

Hoden und Brunstorgane des braunen Landfrosches (Rami fusca).

Arch. f. d. gesamte Phys., CXXVI.
Oudemans, H. Th., 1899. Falter aus kastrierten Raupen, wie sie aussehen und mk

sich benehmen. Zoolog. Jahrb., Abt. Syst., XII.
Packard, A. S., 1875. On gynandroniorphism in the Lepidoptera. Mem. Boston -

Nat. Hist., II.
Painter, T. S., 1913. On the dimorphism of the males of Mema vitlata, a jumping spider.

Zool. Jahrb. Abt., Syst., XXXV.
Pearl, R., and M. R. Curtis, 1909. Studies on the physiology of reproduction m the

domestic fowl. Biol. Bull., XVII.

, 1910. Studies on hybrid poultry. Rep*. Maine Agri. Exp. Station.

( and F. M. Surface, 1910. On the inheritance of the barred color pattern in

poultry. Arch. Ent.-mech., XXX.
; , 1910. Further data regarding the Bex-limited inheritance of the

barred colored pattern in poultry. Science, XX XII.

, and A. M. Boring, 1917. Sex studies. Anat. Rec, XII I.

, f 1917. The corpus luteum in the ovary of the domestic fowl. Am. Journ. of

Anat., XXIII.
Pearse, A. S., 1912. The habits of fiddler crabs. Philippine Journ. Science, VII.
f 1914. On the habits of Uca pugna and Uca pugilator. \\ i.-<on. MM.

XV1L , ,, l . ■

Petrunkewitch, A., 1911. Sense of sight, oourtahip, and mating in

theraphosid spider from Texas. Zool. Jahrb., Al . XXXT

Pocock, R. I., 1905. The effects of castration on the bornfl of the pfOOgbook. Pro.

Zool. Soc, London.

1905.
1909.

104 BIBLIOGRAPHY.

Poll, H., 1909. Zur Lehre von den sekundaren Sexualcharakteren. Sitz. Ges. Nat. Fr.

zu Berlin.
Poncet, A., 1903. De l'influence de la castration sur le developpement du squelette.

Compt. Rend. Soc. Biol., LV.
Porter, J. P., 1906. The habits, instincts and mental powers of spiders. Jour. Phys.,

XVII.
Potts, F. A., 1906. The modification of the sexual characters of the hermit crab caused by

the parasite Pellogaster. Q. J. M. Sc, L.
Punnett, R. C, 1915. Mimicry in butterflies. Cambridge.
Punnett, R. C, and P. G. Bailey, 1914. On inheritance of weight in poultry. Journ. of

Genetics, IV.
Rasmussen, A. T., 1917. Seasonal dhanges in the interstitial cells of the testis in the wood-
chuck, Marmota inonax. Am. Jour, of Anat., XXII.
Reagan, E. P., 1916. Some results and possibilities of early embryonic castration. Anat.

Record, II.
Reeves, T. P., 1915. On the presence of interstitial cells in the chicken's testis. Anat.

Record, IX, 1915.
Rogen, J., 1909. Kastration und ihre Folgeerscheinungen bei Gryllus campestris L. Zool.

Anz., XXXIV.
, 1910. Kastration und ihre Folgeerscheinungen bei Gryllus campestris L.

Zool. Anz., XXXV.
Ribbert, 1897. tJber Veranderungen transplantierter Gewebe. Arch. f. Ent.-mec, VI.
, 1898. liber Transplantation von Ovarium, Hoden und Mamma. Arch. f.

Ent. mech., VII.
Romanes, G. J., 1892. Darwin, and after Darwin. Chicago.
Rorig, A., 1899. Welche Beziehungen bestehen zwischen den Reproduktionsorgauen der

Cerviden und der Geweihbildung derselben? Arch. Entw.-mech., VIII.
, 1901. Korrelationen zwischen gewissen Organen der Cerviden und den

Geweihen derselben. Verhandl. internat. Zoolog. Kongress, Berlin.
Shattock, S. G., and C. G. Seligmann, 1906. An example of true hermaphroditism in the

domestic fowl with remarks on the phenomenon of allopterotism. Trans.

Pathol. Soc. of London.
Smith, Geoffrey, 1910-1912. Studies in the experimental analysis of sex. Q. J. Micro.

Sci., LIV-LVIII.
, 1913. On the effect of castration on the thumb of the frog, Rana fusca.

Zool. Anz., XLI.
Smith, Geoffrey and Mrs. Haig Thomas, 1913. On sterile and hybrid pheasants. Jour.

Gen., III.
Spengel, J. W., 1876. Das Urogenitalsystem der Amphibien. Arbeit. Zool. Zootom. in

Wiirzburg, HI.
Spengel, J. W., 1889. Zwitterbildung bei Amphibien. Biol. Centralbl., IV.
Spillman, W. J., 1908. Spurious allelomorphism. Am. Nat., XLII.

, 1909. Barring in barred Plymouth Rocks. Poultry, V.

Stamati, G., 1888. Sur l'opration de la castration chez l'6crevisse. Bull. Soc. Zool.

France, XIII.
Steinach, E., 1910. Geschtechtstrieb und echt sekundare Geschlechtsmerkmale als Folge

der innersekretorischen Funktion der Keimdriisen. Zeit. f. Phys.,

XXIV.
, 1912. Willkurliche Umwandlung von Saugetier-Miinnchen in Tieren mit

ausgepragt weiblichen Geschlechtscharakteren und weiblicher Psyche.

Archiv. ges Phys., CXXXXIIII.
Stephan, P., 1902. De 1'hermaphroditisme chez les vertebres. Annal. de la faculty des

sciences de Marseille, XII.

BIBLIOGRAPHY. 105

Stockard, C. R., 1911. The fate of the ovarian tissues when planted on different <>rv

Arch. f. Ent. Mech., XXXII.
, and G. N. Papanicolau, 1917. The existence of a typical CBsUrom oyek

in the guinea-pig. With a study of its histological and physiological

changes. Am. Jour, of Anat., XXII.
Stotsenburg, J. M., 1913. The effect of spaying and semi-spaying young albino rats

(Mus norvegicus albinus) on the growth in body weight and body liimth.

Anat. Rec, VII.
Sturtevant, A. H., 1911. Another sex-limited character in fowls. Science, XXXIII.
, 1912. An experiment dealing with sex-linkage in fowls. Journ. Exp. 'A<«A ,

XII.
, 1915. Experiments on sex recognition and the problem of sexual selection in

Drosophila. Journ. of Animal Bchav., V.
Swift, C. H., 1914. Origin and early history of the germ cells of the chick. Amor.

Jour, of Anat., XV.
Tandler, J., und K. Kellar. Die Korperform der weiblichen Fruehkastraten dflfl Ki;

Archiv Entw-mech., XXXI.
, und S. Grosz, 1909, 1910. Ueber den Einfluss der Castration auf den

Organismus. Archiv f. Entw-nech., XXVII, XXIX, XXX, 1909.
, und S. Grosz, 1913. Die biologischcn Grundlagen der Bekundaren < '■>-

schlechtscharaktcre. Berlin.
V. la Valette St. George, 1895. Zwitterbildung beim kleinen Wassermokh. Arch.

Mik. Anat., XLV.
Walker, C. E., 1908. The influence of the testis upon the secondary sexual characters of

fowls. Proc. Royal Soc. of Med. x

Wallace, A. R., 1891. Darwinism: An exposition of the theory of natural selection.

London.

, 1891. Natural selection and tropical nature. London.

Weber, M., 1890. tjber einen Fall von Hermaphroditismus bei Friugilla cothbs Zool.

Anz., XIII.
Yarrell, W., 1827. On the changes in the plumage of some hen-pheasants. Phil.

Trans., 117.
, 1857. On the influence of the sexual organ in modifying external character.

Journ. Proc. Linn. Soc, I.

DESCRIPTION OF PLATES.

Plate 1.

Fig. 1. Black -Breasted Game bantam cock. He is typically cock-feathered, but, as in all
games, his hackles and tail-coverts are shorter than in the cocks of other
breeds. The comb was dubbed by the breeder.

Fig. 2. Black-Breasted Game bantam hen. The great contrast in color between the cock
and hen is practically the same as that in the Brown Leghorn, in most races
of Tosa fowls, and in the wild type Gallus bankiva.

Fig. 3. Sebright cock, "hen-feathered." The short hackles, the rounded feathers of the
back and saddle, and the shortness of the tail-coverts are characteristic
features of these males. For details of individual feathers from different
regions see plates 6 and 8.

Fig. 4. A castrated Sebright male. The drawing was made about a year after the opera-
tion. This particular bird developed a lighter color than did other castrated
Sebrights (see plate 3, fig. 1). The entire dorsal region has changed its color,
and the feathers have also changed in shape, length, etc. Note especially
the very long hackle and saddle feathers (for details see plate 6, fig. la) and the
change in the wing-bow. The tail-coverts have also grown long.

Plate 2.

Fig. 1. Fi hen-feathered male out of Game by Sebright. The hen-feathering in this bird is

as complete as in the Sebright.
Fig. 2. Fi female out of Game by Sebright.
Fig. 3. Castrated male originally hen-feathered (292), nearly black in color, as shown by the

individual feathers of plate 7, figure 2. After castration the bird has become

red above, with black iridescent tail-coverts, and deeper yellow (or red) below.
Fig. 4. Castrated Fi male, originally like figure 1. Note especially the change in color of

the whole upper surface that has become red, like that of the jungle-fowl.

The tail-coverts have grown long and are now iridescent black. The breast

has changed least, but is a richer yellow. The comb and wattles and ear

lobes are shrunken, as in all capons.

Plate 3.

Fig. 1. A castrated Sebright male. The operation was performed on a juvenile bird; the
drawing was made a year later. The bird is typical as to the change in color
that takes place in the Sebright. He was darker red than the bird shown
in plate 1, figure 4. The red was more mahogany than the picture shows.
The original feathers were like those in plate 6, fig. 2 (there erroneously
referred to as those of light-colored Sebright) .

Fig. 2. An F2 hen-feathered very dark male. The condition of his plumage at the time of
the operation is shown in this figure. The change that took place after castra-
tion is shown in the next figure.

Fig. 3. The change that took place in the bird drawn in figure 2 is shown here. The whole
upper surface has become red, except the tail-coverts, which are iridescent
black. Note also the change in color on the wing-bow. For the details of the
feathers see plate 9, figures 1, la.

Fig. 4. A castrated F2 bird that had been hen-feathered and had changed over to cock-
feathering, as shown here. The color and the details of the original hen-
feathering are shown in plate 9, figures 2 and 2a.
106

DESCRIPTION OF PLATES. K)7

Plate 4.

Fig. 1. One of the original Black-Breasted Game males used in the breeding eanerimi

Compare with colored drawing, plate 1, figure 1.
Fia. 2. A Black-Breasted Game hen used in the breeding experiments. CompSTC with

colored drawing, plate 1, figure 2.
Fig. 3. A Sebright male. The bird was used in the later back-crosses and not in the origin.il

experiments. He is typical of his breed.
Fig. 4. A Sebright female. One of the birds used in the original experiments.
Fig. 5. An Fi male. This bird had just reached maturity and was younger than the one

drawn in plate 2, figure 1.
Fig. 6. An Fi hen of the same age as the last. The pattern changed a little as the hir-1

became older.

Plate 5.

Fig. 1. An adult Sebright male for comparison with the next figure.

Fig. 2. A castrated Sebright male. This photograph shows the same bird from v. hiell the

drawing, plate 1, figure 4, was made. It is the lighter colored bard refemd

to in the text.
Fig. 3. One of the two Fi castrated birds. For comparison see the colored drawing in

plate 2, figure 4.
Fig. 4. A castrated Sebright. This bird is darker, and in this sense more typical than

figure 2.
Fig. 5. One of the castrated Sebright males which at one time after castration was as

extremely cock-feathered as figure 2, but slowly "went back" towards !

feathering, as the figure shows especially in the hackle and saddle. The

details are much better shown in the feathers photographed in plate 8, Ggurefl

1, 2, 3, 4, la, 2a, 3a, 4a, 16, 26, 36, 46.
Fig. 6. The same bird was opened and the regenerated pieces of the testis removed. He

returned later, as shown here, to full cock-feathering.

Plate 6.

Figs. 1, 1a. Typical old (1) and new (la) feathers (after castration) of the same bird. This
is the "fighter" male drawn in plate 1, figure 4, and photographed in plate
5, figure 2.

Figs. 2, 2a. Typical old (2) and new (2a) (after castration) feathers of another Sebright.
This bird developed after castration darker feathers than did the last bird.
Its feathers were more like those that other castrated Sebrights develop
Legend on plate 6 erroneous as far as 2 and 2a are concerned.

Plate 7.

Figs. 1, 1a. Typical old (1) and new (la) (after castration) feathers of an Fi bird. (See

plate 2, figures 1 and 4.)
Figs. 2, 2a. Typical old (2) and new (2a) (after castration) feathers of bird shown in plate

3, figures 2 and 3 (No. 292).

Plate 8.

Typical feathers of "dark" Sebright (1, 2, 3, 4) that after incomplete castration ehonfed to
cock-feathering (la, 2a, 3a, 4a), then later, as pieces of the testei that had !
left behind in the old situs regenerated, began to go back toward- hen-
feathering (16,26, 36, 46). The bird was then opened .main, and the re^-n. r-
ated pieces removed, when it again became cock-feat bend (1«, 26, 3c, 4c),
and has so remained for more than a year.

108 DESCRIPTION OF PLATES.

Plate 9.

Figs. 1, 1a. Typical feathers of hackle and saddle from hen-feathered bird (No. 68)

plate 3, figure 2, that changed over to the cock-feathered bird of plate 3,
figure 3.

Figs. 2, 2a. Typical feathers of an Fi male (2) that changed over partly as a result of degen-
eration of his testes, into a cock-feathered bird (2a) . The change was not so
great as it is after castration.

Figs. 3, 3a. Typical feathers of Sebright male that slightly changed towards cock-feathering
(old hackle feather missing).

Plate 10.

Figs. 1, 1a. Old (1) and new (1a) wing-coverts of normal Sebright (1) and castrated (1a).
Figs. 2a, 2b. Upper row, to right, "Transitional" hackle feathers (2a), and a slightly

later changed-over feather from wing-bow (2a), and from back (2b). Second

row, to left, old (2), transitional {2a), and changed-over feather (26), from

saddle of Sebright.
Fig. 3. Three feathers (tail-covert, wing-bow, and saddle) of an F2 hen-feathered game-like

male.
Fig. 4. A series of breast feathers from an F2 bird. At one end of the series (the left) the

feather is spangled, at the other barred.
Fig. 5. A series of breast feathers from another F2 bird. At one end of the series (the

left) the feathers are penciled, at the other end they are barred.

T. H. MORGAN

1. Black Bre isted • lam< Banl m

2. Female.

T. H. MORGAN

1. lien- feathered Fj male.

2. Fj female.

T. H. MORGAN

PLATE 3

1. Castrated Sebright male.

2. F, Hen-feathered male.

■

MORGAN

PI A

1. Black-Breasted < lame male.
:>. Sebrighl male.
5. Hybrid male.

2. I
-
6. Hybrid i

MORGAN

2

3

.)

1. Adult Sebright male

3. Castrated F, male

5. Castrated Sebrigb.1 male with >

T. H. MORGAN

PLA^

Feathers of •light" colored E
after castration i I ■<. 2a).

T. H. MORGAN

E 7

Saddle

■

Feathers of F, hen- feathered male

Feathers of a darker hen-feal

T. H. MORGAN

A*

•

thers from hen- feathered in '
feathered male (I1. 2a, after

testes regenerated ( l\ 2b, 3b, l! I . then cl
castration (lc. 2C,

r. H. MORGAN

Hackle Bad

II, •

Feathers showing complete (1) 01
feathering to cock-feathering <

T. H. MORGAN

• 10

\

Covert

2a

Saddle

^^

Hackle

Br<

Normal. 1, 2; transitional, la, 2a, and changed
3, 4 and 5. Feathers from F. birds.

\

623 BE br 2515

02/17/97 41245 , .