CONTRIBUTIONS TO EMBRYOLOGY

VOLUME XIII, Nos. 57-64

PUBLISHED BY THE CAKNEGIE INSTITUTION OF WASHINGTON
WASHINGTON, 1921

CARNEGIE INSTITUTION OF WASHINGTON
PUBLICATION No. 276

PRESS OF GIDSON BROTHERS
WASHINGTON

CONTENTS.

PAGE

No. 57. On the development of the lymphatics in the stomach of the embryo-pig. By JAMES R. CASH.

(3 plates, 3 text-figures) 1-15

58. On the fate of the primary lymph-sacs in the abdominal region of the pig, and the development of

lymph-channels in the abdominal anil pelvic regions. By F. L. REICHERT. (5 text-figures) .... 17-39

59. Relative weight and volume of the component parts of the brain of the human embryo at different

stages of development. By GEORGE B. JENKINS. (12 text-figures, 1 chart) 41-00

60. Abnormalities of the mammalian embryo occurring before implantation. By GEORGE \V. CORNER.

(2 plates, 1 figure) 61-66

61. The development of the external genitalia in the human embryo. By MILO HERRICK SPADLDING.

(4 plates, 2 text-figures) 67-88

62. Further experimental studies on fetal absorption. By GEORGE B. WISLOCKI. (1 plate) 89-101

III. The behavior of the fetal membranes and placenta of the guinea-pig toward trypau
blue injected into the maternal blood-stream.

IV. The behavior of the placenta and fetal membranes of the rabbit toward trypan blue
injected into the maternal blood-stream.

63. The distribution of mitochondria in the placenta. By G. B. WISLOCKI and J. A. KEY. (1 plate) . . 103-115

64. Cyclic changes in the ovaries and uterus of swine, and their relations to the mechanism of

implantation. By GEORGE W. CORNER. (4 plates, 2 text-figures) 117-146

HI

5 I

CONTRIBUTIONS TO EMBRYOLOGY, No. 57.

ON THE DEVELOPMENT OF THE LYMPHATICS IN THE
STOMACH OF THE EMBRYO PIG,

BY JAMES R. CASH,
From the Anatomical Laboratory, Johns Hopkins University.

With three plates and three text-figures.

ON THE DEVELOPMENT OF THE LYMPHATICS IN THE
STOMACH OF THE EMBBYO PIG,

The work of Dr. Mall (1896) on the anatomy of the lymphatics of the wall
of the stomach has established the plexiform arrangement of these vessels and their
relation to the other structures of this organ. By dissecting the stomachs of dogs in
which the lymphatics had been injected, and by reconstructions of such dissections,
the entire organ was shown to be supplied with four definite, homogeneous plexuses.
The most delicate of these, and the one most intimately related to the lining of the
stomach, is the mucosal plexus. It lies in the mucosa at the base of the gastric
glands just external to the muscularis mucosse and receives branches which lie
between the glands and extend out as far as the gastric pits. Owing to the delicacy
of these vessels and the fact that valves prevent their injection from the main
plexus of the submucosa, any attempt at complete injection is attended with much
difficulty. As a matter of fact, the mucosal vessels can be injected in the adult
only by direct puncture, which means that the needle must actually pass through the
muscularis mucosa? and enter the narrow zone between it and the base of the gland ;
it can be readily seen that such an exact placing of the needle occurs only by chance.
However, th( best specimens obtainable indicate clearly that the lymphatics
between the glands communicate directly with this delicate, homogeneous plexus
lying at the base of the glands of the mucosa.

Proceeding through the wall of the stomach toward its serous surface, the first
plexus of the submucosa is encountered. This consists of a single layer of lym-
phatics, just outside the muscularis mucosse, and forms the connecting link between
the mucosal plexus, just described, and the main plexus of the submucosa, the
principal plexus of the wall of the stomach. This main plexus is a dense mass of
vessels which practically fills the submucosa and is composed of several layers of
large, tortuous, thickly arranged lymphatics. Springing from it are numerous
slender vessels which traverse the muscular coats of the stomach, forming a coarse
plexus between them, and are gathered together on the surface of the stomach into
the subserous plexus, a dense, single layer of lymphatics situated just beneath the
peritoneal covering of the organ.

The work of Dr. Mall extended that of Sappey, Teichmann, and Loven. It
may be seen, therefore, that the anatomy of the lymphatic vessels in the wall of
the adult stomach is well known. The pathways of lymphatic drainage from this
organ, however, are not so clearly understood. All investigations along this line
have consisted in partial injections of the adult stomach. In the present experi-
ments the direction of flow of the injection mass from different points was noted and
from the information so gained the direction of normal lymphatic drainage, in vivo,
is inferred.

The most extensive work of this kind has been done by Cuneo. By injecting
colored substances directly into the main plexus of the submucosa, and watching its

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

the stomach. After Cunoo, 1902.

1. The coronary or principal current.

2. Right gastro-epiploic current.

3. Splenic current.

flow over the wall of the stomach and its appearance in regional lymphatic glands, he
has obtained a very good general idea of this question. As a result, he felt justified
in dividing the stomach into three main lymphatic zones (Poirier and Cuneo, 1902).
These arbitrary zones are shown in text-figure 1. The arrows show the direction
of the flow of lymph from the different parts. Zones 1
and 2 drain the glands of the lesser curvature, while the
flow from zone 3 goes to the hilum of the spleen and
then on to the pre-aortic glands. It is to be noted that
no drainage is indicated by way of the duodenum.
Cuneo states that most of the lymph flows directly
toward the lesser curvature and passes through glands
situated in this region, whose efferent ducts drain into
the pre-aortic lymph-nodes just anterior to the cy sterna
chyli. Zone 1 represents this area. In the area desig-
nated as zone 2 the flow of lymph is toward the pylorus, TEXT.FIQ. 1._Lymphatic tcrritoricn 0,
as indicated, but passes posterior to it to the glands
of the lesser curvature. From zone 3 lymph passes
through vessels in related folds of peritoneum, to the
hilum of the spleen and from there to the pre-aortic lymph-nodes.

As one looks at a cross-section of a stomach wall in which the lymphatics have
been injected, the extreme vascularity of the organ becomes immediately apparent.
The plexuses are so dense, permeate the entire organ so thoroughly, and are so rich in
their anastomoses that definite lymphatic zones and fixed points of drainage for the
different regions are anatomically difficult to establish. The more conservative
and apparently more accurate view is that the wall of the stomach contains four
complete lymphatic plexuses, in open communication with each other and through
which lymph may flow, peripherally, in any direction. It is quite difficult to con-
ceive of such dense, homogeneous plexuses being divided into any zone-like arrange-
ment. Certainly, the lymph will leave the stomach by way of the nearest and most
accessible avenue of exit, and injections made around the regions toward which
lymphatic flow occurs will apparently find zones in which most of the flow takes
place in a given direction; but in view of the complex nature of the vessels within
the wall of the stomach the accuracy of areas found in this manner becomes very
doubtful. Thus it is evident that our knowledge of the gastric lymphatics is not yet
complete. Such methods as those mentioned are sufficient only for the formation
of general ideas and beyond this point any statement is speculative. As yet, no
work has been done on the embryology of these vessels. To resolve such dense
lymphatic plexuses into their component parts by injection after they are fully
developed is obviously impractical. By a study of their origin and development it
was hoped that more accurate views might be adopted for the lymphatic drainage
of this organ, and it was with this idea that the present investigations in embryo
pigs were undertaken.

By the method of injection, embryos can be studied in various stages before
the lymphatic development is complete. In this way the origin of the vessels, the

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG. 5

points at which they reach the stomach, their manner of growth, their anastomoses,
and finally the relation of the gastric lymphatics to the lymphatics of other organs
can be observed. Embryos 40, 60, 80, and 120 mm. in length were used. Those
up to 90 mm. were injected through the retroperitoneal sac, or its remains, the
large lymphatic plexus of vessels between the Wolffian bodies. In the embryos
measuring over 90 mm. the injections were made directly into the submucosal
plexus at the lesser curvature, since valves and nodes, forming at this age, prevent
retrograde injection of the organ from the retroperitoneal sac. In such a study it
would be very helpful to have complete injections of the lymphatics of the adult
stomach and its surrounding organs, but the valves and nodes along the course of
the vessels act as effective barriers to extensive injections. For a successful
injection the embryo must be fresh, preferably with the heart still beating. Hypo-
dermic syringes of 1 c. c. capacity with fine needles (No 28), were used for inject-
ing. The best specimens were obtained with a saturated aqueous solution of
Berlin blue as the injection mass. Most of the material was cleared in oil of winter-
green by the Spalteholz method. Thick microscopic sections were cut to show the
progressive development of the plexuses within the stomach wall and the course of
the vessels which traverse the ligaments of the spleen. The actual technique of
injecting the retroperitoneal sac has been described by Baetjer (1908) in his work
on the morphology of that structure and need not be repeated here.

In their development the lymphatics of the stomach pass to it by way of its re-
lated folds of peritoneum. Therefore, a clear idea of the relation of these folds to
the stomach, as well as to the points of origin of the lymphatic trunks which invade
it, is imperative for an understanding of the development of these vessels. As the
relations of this part of the peritoneum in the pig are slightly different from those in
the human body, a brief description of them is necessary. The dorsal mesogastrium
which later forms the great omentum, contains the spleen between its layers. The
spleen lies ventralward, close along the greater curvature of the stomach, and has
two ligaments. The first of these is the gastro-splenic, that part of the omentum
connecting the spleen with the greater curvature of the stomach. The second is
the splenic ligament, that part of the omentum which forms the posterior wall of
the omental bursa and from the hilum of the spleen is continuous with the perito-
neum covering the Wolffian body, the general mesentery of the intestine and the
mesentery of the duodenum. The transverse colon, with its mesentery, lies free from
the omentum. The duodenum has a broad, fan-shaped mesentery which continues
cephalad to that part of the omentum forming the posterior wall of the omental
bursa and caudad to the common mesentery of the small intestine. The ventral
mesentery remains as the gastro-hepatic ligament, connecting in the usual manner
the lesser curvature of the stomach with the hilum of the liver.

The retroperitoneal sae, discovered by F. T. Lewis, has been well described by
Sabin, Baetjer, Heuer, and other workers in this laboratory. It is a triangular
structure which lies at the base of the mesentery at the level of the coeliac axis.
There are two small indentations along its sides, made by the adrenal bodies, which
thus divide the sac into two lobes. From the posterior lobe, which is the larger,

6

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

arise the lymphatics of the intestine and of the abdominal structures, while from
the anterior lobe spring vessels to the stomach, spleen, duodenum, diaphragm, and
lungs (text-figure 2). It is this anterior portion of the sac with which we are pri-
marily concerned in this study. The entire sac is obliterated during development,
being the primordium for the chain of lymph-glands which reach from the cceliac
axis to the bifurcation of the aorta.

The gastric vessels arise by two main trunks from the anterior lobe of the retro-
peritoneal sac. The right trunk is much larger than the left and passes behind the
stomach to the lesser curvature, where it invades the stomach at three different

D.V.

TEXT-FIO. 2. — Diagram of the retroperitoneal sac showing
ing the relation of the gastric trunks as> they leave its
anterior end. The early lymphatics extending to the
pillars of the diaphragm are also illustrated.

A., Adrenal; D. P., pillar of the diaphragm; D. V.,
diaphragmatic vessels; H. V., hepatic vessels; K., kidney;
R. G. T., right gastric trunk; R. P. S., retroperitoneal sac;
V. S. M., vessels of splenic mesentery, W. B., Wolffian
body.

points: (1) the opening of the esophagus, (2) the center of the lesser curvature, and
(3) the pyloric extremity of the lesser curvature. These three sets of vessels to the
stomach are well shown by a comparison of figures 6 and 7. It can be seen (fig. 6)
that another vessel of considerable size is given off from the right trunk, or arises
as a separate trunk from the anterior part of the retroperitoneal sac, and passes
through the mesentery to the duodenum (A. Duo. V.). It then extends along the
duodenum to the pylorus, where it anastomoses freely with the pyloric vessels,
reaching the stomach from its posterior surface (fig. 4) .

The left gastric trunk, or splenic trunk, as it may be called from its distri-
bution, divides into two branches. One passes directly to the cardiac pouch and
is not shown in the figures; the other passes through the splenic ligament to the
hilum of the spleen, the interior of which it has never been seen to enter, and then

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG. 7

traverses the gastro-splenic ligament to the center of the greater curvature of the
stomach where it ramifies to right and left (fig. 4). These lymphatics then anasto-
mose, over both the anterior and the posterior walls, with those from the lesser
curvature, where connections are formed with the lymphatics of the esophagus and
duodenum (cf. figures 4 and 5, plate 1).

After this brief description of the general pathways along which lymphatic
invasion of the stomach takes place, the characteristics of each invading set of
vessels and their part in forming the general lymphatic plexuses remain to be
considered.

jr^^^~^

TEXT-FIO. 3. — Pig embryo,
100 mm. long. Dia-
grammatic transverse sec-
tion of the stomach and
spleen with their perito-
neal ligaments. The ar-
rows which begin from the
retroperitoneal sac show
the directions in which
lymphatic invasion of the
stomach takes place. In
this specimen the stomach
is collapsed. The hilum
of the spleen is marked by
vessels of the splenic mes-
entery (V. S. M.); also
the greater curvature of
the stomach may be
located both by its vessels
(V. G. C.) cut in cross-
section and by the attach-
ment of the gastro-splenic
ligament.

The stomach may be regarded as being invaded from the posterior side at the
lesser curvature by the right trunk, and anteriorly along the greater curvature
by the left trunk. The posterior invasion of the lesser curvature takes place
along the gastro-hepatic ligament (the primitive ventral mesentery), while the
anterior invasion at the greater curvature is by way of the splenic and gastro-
splenic ligaments (the primitive dorsal mesentery of the stomach) and the mesentery
of the duodenum (text-figure 3). This arrangement offers a striking contrast to
the manner in which the lymphatics reach the intestine. Here, straight trunks
grow out from the retroperitoneal sac through the mesentery and, on reaching the
wall of the intestine, divide into right and left branches to surround the gut. In the
case of the stomach the right and left branches leave the retroperitoneal sac directly
and pass to the stomach by way of two separate folds of related peritoneum.

The earliest stage injected was an embryo 28 mm. in length (fig. 7). The
vessels arising from the retroperitoneal sac can be seen just reaching the stomach.
This is a very early stage and the injected vessels are those arising from the right
gastric trunk of the retroperitoneal sac. In this specimen the spleen and liver have
been removed. It has never been possible to inject the splenic vessels (left gastric
trunk) at this stage, but the mode of entrance of the right gastric trunk is well
illustrated. Figure 7 shows the primitive mass of lymphatics which later becomes
grouped into the right gastric trunk and distributed to the stomach in three parts :

j LI8!

8 LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

at the esophageal opening, at the center of the lesser curvature, and at the pylorus.
It is from this mass of lymphatics that most of the gastric vessels develop. It will
be noted in the figure that the lymphatics to the esophageal opening already suggest
the ring which is later to become so characteristic. The central mass to the lesser
curvature can be seen emerging from behind the hepatic vessels which are cut off
in the drawing. These vessels to the lesser curvature are shown spreading out over
the anterior surface of the stomach ; they also extend over the corresponding area
on the posterior wall. Two or three small vessels (P. V.) pass forward to the
pylorus and it is important to note that they always reach this structure at its
posterior wall. It is these vessels which will later anastomose with both the ascend-
ing duodenal vessels (A. Duo. V.) and the vessels of the greater curvature (V. G. C.),
as shown in figure 4.

Growth is very rapid and in slightly older specimens (35 to 45 mm.), in which
the splenic vessels have also appeared, the rapid invasion from the lesser cuvature is
apparent. The vessels reaching the esophagus have by this time formed a complete
subserous, periesophageal ring. This ring continues to develop and in older speci-
mens forms a dense, circular plexus around the esophagus, from which branches pass
in all directions to ramify over the stomach wall (fig. 8). The annular structure of
this plexus is not only manifest as a subserous set of vessels, but is maintained in
the depths of the stomach wall in this region as development proceeds.

The vessels which reach the stomach at the center of the lesser curvature are
by far the most numerous of any of the groups. As they traverse the gastro-
hepatic ligament they are in direct communication with the lymphatics of the liver,
which pass through this structure and also take origin from the retroperitoneal sac
(fig. 2) . On reaching the stomach they branch over both the anterior and posterior
walls and form numerous anastomoses with the vessels from the esophageal ring,
as well as with the lymphatics reaching the lesser curvature at the pylorus. These
pyloric vessels form the smallest group of lymphatics invading the stomach. They
spring from the right trunk just lateral to the large group at the center of the lesser
curvature and pass through the gastro-hepatic ligament to the posterior surface of
the pylorus, over which they anastomose freely with the set of vessels reaching the
pylorus by way of the mesentery of the duodenum.

This latter set of lymphatics (fig. 4), to which we may refer as the ascending
duodenal vessels, forms a trunk of considerable size. They can be injected only in
the younger embryos, due to the fact that one or more lymph glands are later
formed in the mesentery of the duodenum at the pyloro-duodenal junction. Their
rich anastomosis with the other gastric vessels presents a subject of unusual interest,
inasmuch as it is not generally stated that lymphatic drainage of the stomach takes
place partly by way of the mesentery of the duodenum. The magnitude of early
vessels invading the stomach from this source and the extent of their anastomoses
with the gastric vessels proper give proof of a well-defined pathway by which
lymphatic drainage from the stomach may take place. Figures 4 and 6 show the
extent of the anastomosis of these ascending duodenal vessels with the other vessels
of the pylorus.

LYMPHATICS IN THE STOMACH OP THE EMBRYO PIG. 9

The origin of the ascending duodenal group is somewhat varied in the different
specimens. In some, these vessels spring from the right gastric trunk as it leaves
the retroperitoneal sac; in others, they arise from the sac itself just posterior to the
right gastric trunk. In their course through the duodenal mesentery they present
no unusual features, but the lymph-glands, later formed along their course at the
pyloro-duodenal junction, offer a striking point of difference from the glands of the
other vessels leaving the stomach. These glands prevent the injection of the
duodenal lymphatics from points on the stomach and this fact most likely accounts
for the failure to recognize such a connection between gastric and duodenal lym-
phatics by methods such as have been hitherto employed to demonstrate them.

The left gastric trunk, arising from the retroperitoneal sac, immediately
divides into two branches. The lesser branch passes by way of the extreme left
portion of the omental bursa to the anterior surface of the stomach at the base of
the cardiac pouch. The cardiac pouch is a transient, embryonic structure. In the
pig, during embryonic life, there is a slight annular constriction of the stomach wall
near the cardiac end, which gives to that portion of the stomach a pouch-like
appearance (fig. 5). This structure becomes less evident as development proceeds
and in very late stages is scarcely perceptible. The lesser branch of the left cardiac
trunk is the principal source of invasion of the pouch and forms rich anastomoses
with the vessels reaching the stomach at the center of the lesser curvature and
esophageal opening. The greater branch of the left gastric trunk enters the omental
bursa just medial to the lesser branch. It passes by way of the posterior wall of the
omental bursa (which in this region constitutes the splenic ligament) to the hilum of
the spleen; from this point these vessels pass by way of the gastro-splenic ligament
to the center of the greater curvature of the stomach (figs. 4 and 5). The number
of vessels comprising this set varies from two to four. It is always difficult to be
certain of the exact number, as they are situated very close together in a densely
entwined mass. They have never been observed to enter the spleen, and sections
taken through the point where they pass the hilum of that organ have consistently
failed to demonstrate a trace of injection mass within its interior. So it may be
said that these vessels appear in no way related to the spleen itself but merely take
advantage of its folds of peritoneum to make their way to the stomach. They
reach that organ on the greater curvature, at a point about midway between the
cardia and pylorus, and from this point their growth proceeds in opposite directions.
Approximately half of them pass to the right, along the greater curvature, to meet
the pyloric vessels from the duodenum which grow along this course ; the other half
pass along the greater curvature, to the left, where they meet anastomosing vessels
from the cardia, esophageal ring, and lesser curvature (fig. 4). All along their
course on the greater curvature these vessels give off branches at right angles to
themselves, which extend over the anterior and posterior walls of the stomach to
meet similar branches given off from the great mass of lymphatics of the lesser
curvature.

Concomitant with the growth of the subserous trunks, growth of the other
plexuses takes place. Sprouting branches from the early trunks of the subserosa

10 LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

dip down into the stomach wall and, as the muscular layers are traversed, the plexus
between the muscular layers is formed . This plexus extends throughout the stomach
wall, but its vessels are of smaller caliber and its meshes much less dense than those
of the other plexuses. It should be borne in mind, however, that vessels lying in
dense, muscular tissue fill less readily with injection mass than those situated in the
loose tissues of the submucosa and subserosa; therefore, the appearance of the
muscular plexus may be less dense than it really is. This point considered, the small
size of the vessels composing the plexus of the muscularis and their striking constancy
of arrangement in all well injected specimens nevertheless are facts which tend to
show that this plexus is a relatively scanty one.

On reaching the submucosa these vessels form the most extensive plexus of
the stomach wall, the main plexus of the submucosa. Here lymphatic growth
takes place very rapidly. At an early stage (20 to 40 mm.) an extremely dense
plexus, several layers in thickness, composed of large, tortuous vessels, has been
formed. In embryos 70 to 100 mm. in length the entire submucosa is found to be
filled by this plexus. Its presence can be easily demonstrated throughout the
entire stomach by injecting directly into the submucosa. As it nears the pylorus
its vessels become smaller and many of them appear to end blindly in this region ;
however, a few of them have been seen to anastomose with those of the deep plexus
of the duodenum. In view of the general tendency toward anastomosis exhibited
by the lymphatics of contiguous tissue, which is especially marked in the lymphatics
of the subserous plexuses of the stomach and duodenum, this scanty connection
between the submucosal plexuses of these two organs can scarcely be considered as
representing the true morphology of the lymphatics in this region. It is evident
that the layers of the stomach wall at the pylorus are compressed by the tone of the
pyloric musculature, as are also the lymphatic channels of the submucosa. Such a
barrier proves quite effective in preventing complete injection of the vessels of this
region. The character of the apparent endings of the vessels of the submucosa at
the pylorus favors such a view. Instead of forming a complete plexus at this
point, most of the vessels end blindly and at different points along the pylorus, thus
giving the exact picture of an incomplete lymphatic injection elsewhere in the body.
From the injections we are justified in saying only that the pathway between the
submucous plexuses of the stomach and duodenum is not easily traversable during
tonic contraction of the pylorus. A sufficient number of anastomoses between
these two plexuses has been demonstrated to show that they are connected and, by
analogy to the lymphatics of the rest of the gastro-intestinal tract, one would expect
ready communication between them during relaxation of the pyloric sphincter.

The further growth of the lymphatics in the wall of the stomach takes place
from this dense submucosal plexus. Many smaller vessels grow still farther inward
and form the scanty plexus adjacent to the muscularis mucosse. This plexus lies
at the base of the gastric glands and sends branches between them.

Figure 8, plate 3, is from a complete injection of the stomach of a pig measuring
150 mm. It has been cut along the greater curvature and spread out to show
especially well the marked esophageal ring of lymphatics. This ring constitutes

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG. 11

one of the most striking characteristics of all injections of the stomach made before
the appearance of glands in this region. Its formation by vessels arising from the
retroperitoneal sac is well shown; branches may be observed passing in every
direction from all points of it ; those going to the cardiac pouch are well seen over
the anterior wall, while in the depths on the posterior wall are seen the vessels of
the left gastric trunk with which they anastomose. Those vessels passing to the
anterior wall are shown as a series of parallel trunks in the serosa which lead into
the deeper, much more extensive plexus of the submucosa within the depths of the
stomach wall. These serosal trunks, as well as the plexus of the submucosa, anasto-
mose with the corresponding trunks and plexus arising from the vessels of the greater
curvature ; they remind one of similar parallel trunks shown by Heuer in the develop-
ment of the lymphatics of the intestine. The posterior surface of the stomach is
seen on the left side of the figure. Here the posterior position of the pyloric vessels
arising from the lesser curvature is well shown, as is also the characteristic arrange-
ment of the serosal trunks and plexuses of the submucosa.

The development of lymphatics of the diaphragm is likewise well illustrated by
these injections. Large lymphatic trunks, accompanied by the lymphatics of the
lower lobes of the lungs which arise from this source, can be seen passing from the
anterior part of the retroperitoneal sac to the pillars of the diaphragm. The
pulmonary vessels penetrate the diaphragm, while the others spread out over its
surface. Also from the gastric plexus, at the lesser curvature of the stomach, many
large vessels pass directly to the diaphragm along the ligaments of the liver and
anastomose freely with those from the retroperitoneal sac, arriving by way of the
pillars (fig. 6) . Such an invasion of the diaphragm is readily explained by the double
origin of this structure, the pillars arising from the pleuro-peritoneal membrane,
while the body of the diaphragm arises from the ventral mesentery, septum trans-
versum, and body-wall. The embryonic lymphatics of the diaphragm are enormous
and soon cover the entire structure with a dense plexus. The active function of the
diaphragm as an agent of absorption is at once apparent by a glance at its rich
supply of lymphatics.

Here a word may be said about the lymphatics of the esophagus and their
relation to those of the stomach. Its vessels are derived from two sources. Those
of the lower end arise from the gastric vessels forming the periesophageal ring
at the cardia of the stomach, from which they spring as direct outgrowths and ex-
tend upward to a point on the esophagus about midway between the bifurcation
of the bronchi and the cardia of the stomach. Here they meet lymphatics which
have extended to the esophagus from the bronchial plexuses at the hilum of the
lungs, primarily arising from the thoracic duct. The lymphatic plexuses of the
esophagus are completed by the anastomosis of the vessels from these two sources.
In their general arrangement they are similar to those of the stomach wall, with
which they are continuous. Lymphatic drainage from the esophagus takes place
in two directions, the lower portion following the general path for gastric drainage
described for the cardia of the stomach, while the upper portion drains to the
bronchial plexuses.

12 LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

Information thus obtained from the injection of embryos did not necessarily
justify conclusions in regard to the lymphatic drainage in the fully developed and
adult stomach. Injections were therefore made of the stomachs of kittens, cats,
and adult pigs. In all specimens thus prepared lymphatic drainage was found
to be similar to that deduced from the study of the injected embryos. This
was clearly illustrated in the fresh stomach of a kitten 24 days old. Three
large glands were seen at the lesser curvature in the gastro-hepatic ligament, one
at the base of the splenic ligament, and one lying in the pyloro-duodenal flexure.
By injecting directly into the submucosa the paths of drainage were readily seen by
the spread of the injection mass through the lympathic plexuses of the stomach wall
and its appearance in the adjacent nodes. Injections were made at several sites:
(1) By injecting directly into the plexus of the submucosa, in the central region of
the anterior and posterior surfaces, the injection mass quickly appears in the nodes
of the lesser curvature; also, some of the vessels along the greater curvature are
filled by part of the injection flowing in that direction. Thus, the flow from the
anterior and posterior walls of the stomach was seen to pass in two principal
directions, toward the lesser and toward the greater curvatures, the major part
going to the former. (2) Injections near the pylorus and from some distance
along the adjacent greater curvature quickly flowed to the gland at the pyloro-
duodenal flexure, while, at the same time, an appreciable extent of the anterior and
posterior walls and the greater curvature was also filled. (3) When the plexus of
the submucosa near the attachment of the gastro-splenic ligament was filled, the
vessels passing from the greater curvature toward the spleen, as well as those from
the spleen to the large gland at the base of the splenic ligament, were injected.
Likewise, from this point of injection, flow occurred freely in all other directions, part
going toward the lesser curvature and part toward the pylorus.

Jamieson and Dobson made a careful study of the lymphatic drainage in the
fully developed human stomach by means of the method of injection and found the
main groups of glands to be associated with the larger arteries of the stomach and
spleen. Along the course of the coronary artery they found a group of glands
forming a periesophageal ring with another group in the gastro-hepatic ligament.
Associated with the splenic artery, in the gastro-splenic ligament, was another group
of glands. Along the course of the hepatic artery and in the bend of the duodenum,
along the gastro-duodenal artery, still other groups were observed. By injecting
directly into the submucous plexus of the stomach wall at various points, the course
of the flow of lymph to these glands was studied. From the results so obtained these
observers were led to believe that any points of division between areas drained by
any definite gland or group of glands were quite arbitrary. They found the plexus
of the muscularis to be a very scanty one. The lymphatics of all plexuses of the
stomach- wall near the cardia were seen to be continuous with those of the esophagus,
whose wall contains lymphatics arranged similarly to those of the stomach. Definite
communications were also found between both subserous and submucous plexuses
of the pyloric portion of the stomach and the duodenum. These connections of the

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG. 13

subserous plexuses were found to be rather scanty and mostly on the posterior walls
of these organs, while the submucous anastomoses were quite numerous.

Such observations are in strict accordance with the ideas one would be led to
form from a study of the development of the gastric lymphatics. The three points
of invasion from the lesser curvature correspond to the chains of glands along the
ascending and descending branches of the coronary artery and the hepatic artery.
The lymphatics reaching the stomach by way of the splenic ligaments, however,
appear to indicate that lymphatic invasion does not always take place along blood-
vessels, since the lymphatics traversing these ligaments are not accompanied by any
blood-vessels of appreciable size. More likely, it seems that related folds of perito-
neum determine the pathways by which the lymphatics reach the stomach; in many
cases the arteries take the same course. The greater number of anastomoses
between the vessels of the pylorus and duodenum than has commonly been observed,
and the fact that most of the connections between the subserous plexuses of these
organs are located on their posterior walls, are points which can be readily under-
stood from the development of the ascending duodenal group of vessels, which
are much more numerous in this region (figs. 4 and 6).

So, from the combined evidence offered by injections of developing lymphatics
in several embryonic stages, and the further injection of these vessels after the
lymphatic plexuses of the stomach have been completely formed, the lymphatic
supply of this organ is seen to be one of the richest in the entire body. Due to the
homogeneity of its plexuses, no strict division of its areas of drainage can be made
that will remain constant in any number of cases. So readily is any portion of the
stomach injected from any other part that, in all probability, the lymphatics of
the entire organ could be filled by a single injection into the submucosal plexus, if
performed slowly enough and with proper pressure. The greatest part of the
gastric drainage certainly takes place by way of the lesser curvature, but the other
two routes — by way of the splenic ligaments and duodenal vessels — are of quite ap-
preciable capacity. Many factors, such as peristalsis, muscle tone, venous engorge-
ment, pressure in the lymphatic vessels themselves, and (in pathological stomachs)
various degrees of obstruction must determine the relative amount of drainage by
these three outlets, the sources of whose output are in continuity and show no zonal
lines of demarcation.

14

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG.

CONCLUSIONS.

1. From a study of development of the lymphatics in the stomach of the embryo
pig, it is found that they invade that organ at two principal points, viz., the lesser
curvature, by way of the gastro-hepatic ligament, and the greater curvature, via
the ligaments of the spleen.

2. Arising in common with the gastric lymphatics is a large trunk which passes
to the duodenum, then up along the duodenum to the pylorus, anastomosing
freely with the lymphatics on the posterior surface of the stomach.

3. Part of the lymphatics invading the stomach at the lesser curvature, early
in their growth, form a dense periesophageal plexus of vessels in the subserosa.

4. From this periesophageal ring, as well as from the other vessels reaching the
stomach at both lesser and greater curvatures, branches are given off at right
angles on both the anterior and posterior walls of the stomach. These vessels
finally meet and encircle the stomach in a segmental manner similar to that' de-
scribed by Heuer for the intestine. There are a number of connections between
these segmental subserous trunks, forming them into a subserous plexus.

5. During the development of this plexus numerous branches from its trunks
pierce the layers of the stomach and form other gastric plexuses. All of these are
homogeneous layers of lymphatic vessels.

6. The richness of the gastric plexuses and their numerous anastomoses preclude
the theory of any sharply defined areas whose drainage would take place in constant
given directions.

7. There are four general pathways by which lymphatic drainage may take
place: (a) the lesser curvature, (b) the greater curvature, (c) the duodenum, and
(d) the esophagus.

I wish to thank Professor F. R. Sabin and Dr. R. S. Cunningham for the valua-
able assistance they have given me in this work, and to express my appreciation to
Mr. J. F. Didusch for the illustrations.

REFERENCES.

BAETJER, W. A., 1908. The origin of the mesenteric sac and
thoracic duct in the embryo pig. Amer. Jour. Anat.,
vol. 8.

CASH, J. R., 1917. On the development of the lymphatics in
the heart of the embryo pig. Anat. Rec., vol. 13.

, 1919. On the development of the lymphatics of the

btomach of the embryo pig. Anat. Rec., vol. 16, p. 145.

CUNEO, B., 1902. Systemelymphatique,(PoirierandCuneo).
Traite d'anatomie humaine, Poirier and Charpy.

CUNNINGHAM, R. S., 1916. On the development of the
lymphatics of the lungs in the embryo pig. Contri-
butions to Embryology, vol. 4, Carnegie Inst.
Wash. Pub. No. 224.

GRAY, H. Anatomy. Lewis ed.

HEUER, G., 1909. The development of the lymphatics in

the small intestine of the pig. Amer. Jour. Anat.,

vol. 9.
JAMIESON, J. R., and J. F. DOBSON, 1907. The lymphatic

system of the cjecum and appendix. Lancet.
KEIBEI, & MALL, 1912. Human Embryology.
MALL, F. P., 1896. The vessels and walls of the dog's

btomach. Johns Hopkins Hosp. Reports, vol. 1.
SABIN, F. R., 1913. The origin and development of the

lymphatic system. Johns Hopkins Hosp. Reports.

vol. 17.

LYMPHATICS IN THE STOMACH OF THE EMBRYO PIG. 15

DESCRIPTIONS OF PLATES.
PLATE 1.

Fia. 4. Pig embryo, 60 mm. Anterior view of stomach showing relation of the lymphatics of the lesser curvature to
those of the liver and diaphragm. The vessels of the splenic mesentery may be seen extending from
beneath the spleen to the greater curvature along which they have grown. Anastomosis has already
taken place between these vessels, growing toward the right along the greater curvature, and lym-
phatics from the lesser curvature, which may be seen coming from the posterior wall of the pylorus.
Anastomosis of the ascending duodenal vessels at this point is also shown. (X19.)
A. Duo. V., Ascending duodenal vessels; D., diaphragm; D. V., diaphragmatic vessels; H., liver; H. V.
hepatic vessels; S., spleen; V. G. C., vessels of the greater curvature; V. L. C., vessels of the lesser curva-
ture; V. S. M., vessels of the splenic mesentery.

FIG. 5. Pig embryo, 40 mm. Posterior view of injected stomach showing the main mass of lymphatics which reach
the stomach at the lesser curvature and their growth over the posterior stomach wall. The vessels
of the splenic mesentery are also here seen to anastomose with those extending from the lesser curvature.
These vessels, which reach the stomach by way of the gastrosplenic ligament, are also shown beginning
to extend to the right and left along the greater curvature. The spleen lies in its usual position. The
cardiac pouch is well marked in this specimen. (X 20.)

C. P., cardiac pouch; 0., esophagus; P., pylorus; S., spleen; V. G. C., vessels of the greater curvature;
V. L. C., vessels of the lesser curvature; V. S. M., vessels of the splenic mesentery.

PLATE 2.

FIG. 6. Pig embryo, 50 mm. View of stomach from right. The organ has been tilted forward to show the great mass
of lymphatics arising from the retropcritoneal sac behind it. The vessels to the duodenum and their
extension toward the stomach are shown. Lymphatics from the retroperitoneal sac, some of which
ramify over the diaphragm, others piercing it to reach the lungs, are also seen. The thoracic duct lies
at the left of the illustration. (X15.)

A. Duo. V., ascending duodenal vessels; D., diaphragm; Duo. M., mesentery of the duodenum; D. V.,
lymphatics of the diaphragm; H., liver; H. V., hepatic vessels; L., lung; Pul.. V., pulmonary
vessels, R. P. S., anterior end of retroperitoneal sac; T. D., thoracic duct; V. L. C., vessels of the lesser
curvature.

FIG. 7. Embryo pig, 30 mm. long. Anterior view of specimen from which the liver has been removed but the gastro-
hepatic ligament left intact. The dark line, along whose edges are seen cut ends of lymphatics, marks
the area from which the liver has been taken. These open vessels are those of the hepatic capsule. The
manner in which the lymphatics of the lesser curvature reach the stomach by way of the gastro-hepatic
ligament and their relations to the lymphatics of the liver are shown here. The branch of the left
gastric trunk passing to the cardiac pouch is also illustrated. The diaphragm may be seen posterior to
the cut edge of hepatic peritoneum. (X17.5.)

C. P., cardiac pouch; D., diaphragm; Duo., duodenum; D. V., diaphragmatic vessels; H. V., hepatic
vessels; L. G. T., left gastric trunk; 0., esophagus; P. V., pyloric vessels; V. L. C., vessels of the lesser
curvature.

PLATE 3.

FIG. 8. Embryo pig, 150 mm. long. Dorsal view of stomach which has been cut along its anterior surface parallel
to the greater curvature and the specimen spread out to show the relation of the lymphatics of the
lesser curvature to those of the anterior and posterior walls of the stomach. The greater curvature is
marked by the vessels which have their course along its border (V. G. C.). It may be noticed that
the superficial plexus is composed of vessels whose general course is at right angles to the curvatures,
while the deep plexus consists in a much denser, homogeneous mass of lymphatics. (X12.)
A. W., anterior wall of stomach; D. G. P., deep gastric plexus (plexus of the submucosa) ; 0., esophageal
opening; P., pylorus; Pan., pancreas; P. W., posterior wall of stomach; R.P.S., retroperitoneal sac;
S., spleen; V. G. C., vessels of the greater curvature; V. L. C., vessels of the lesser curvature.

Ay

4V

•^

J. R. CASH

PLATE 1

.

D.V.

A.Duo.V.

FIG. 4.

0.

V.G.C:

V.S.M.-

V.L.C.

V.S.M.

V.G.C.

J. F. Diduseh fecit.

FIG. 5.

J. R. CASH

PLATE 2

H.V.

P.V.-

V.L.C.-

L.

T.D.

'

'>-/1-

C.P.

0.

J. F. Didusch fecit.

FlG. 7.

J. R. CASH

A.W. V.G.C.

PLATE 3

Pan.

R.P.S.

P.

0.

V.L.C.

5.G.P.

J. F. Didusch fecit.

CONTRIBUTIONS TO EMBRYOLOGY, No. 58.

ON THE FATE OF THE PRIMARY LYMPH-SACS IN THE ABDOMINAL
REGION OF THE PIG, AND THE DEVELOPMENT OF LYMPH-
CHANNELS IN THE ABDOMINAL AND PELVIC REGIONS,

BY F. L. REICHERT.

With five text-figures.

17

ON THE FATE OF THE PRIMARY LYMPH-SACS IN THE ABDOMINAL
REGION OF THE PIG, AND THE DEVELOPMENT OF LYMPH-
CHANNELS IN THE ABDOMINAL AND PELVIC REGIONS,

Since 1900 interest in lymphatics has been centered about the question of
their origin and development. The primary lymph-sacs have been worked out
and in general their fate determined. Sabin (1901-2) has shown that the earliest
lymphatic buds arise in association with the anterior cardinal veins and form the
primary lymphatics in the neck — the paired jugular sacs. The other primary sacs
originate indirectly from the vena cava through the veins of the Wolffian body.
These are the paired iliac sacs, the cisterna chyli (Sabin, 1912) and the retroperi-
toneal sac (Lewis, 1906; Baetjer, 1908). Quoting from Sabin (1913):

"In most general terms the jugular sacs drain the anterior half of the body; the iliac
sacs drain the posterior half of the body; while the retroperitoneal or prse-aortic sac drains
the viscera. The cisterna chyli with the thoracic duct connects the jugular and renal
lymphatics."

The recent work on the origin of blood-vessels by the differentiation of angio-
blasts, or vaso-formative cells from mesenchymal cells (Stockard, 1915; Sabin, 1920;
Streeter, 1920), may result in modifying our present conception of the origin of the
lymphatics to some such form 'as this: that they arise either directly from the endo-
thelium of the veins or in part by a differentiation of new cells analogous to angio-
blasts. The experiments of E. R. and E. L. Clark (1920) point in that direction.
It is clear, however, that lymphatics begin centrally, close to certain veins, and
spread peripherally.

Considering the sequence in the development of the primary lymphatic system,
it was at first thought by Sabin (1901) that the earliest lymphatics in the pig were
sacs which bud off from certain veins. F. T. Lewis (1906), going back a step
farther, showed that in the rabbit the formation of the primary lymph-sacs is
preceded by a plexus of blood-filled capillaries connected with the veins. This was
confirmed by Sabin (1912) for pig and human embryos, and by Huntington and
McClure (1910) for the cat. Recently E. R, and E. L. Clark (1920) have carried
the investigation back still farther in the chick, practically to the beginning, when
the first evidence of lymphatics consists of a few blood-filled vessels lined with
characteristic lymphatic endothelium and connected with veins. These vessels
later join to form a circumscribed continuous plexus of vessels which does not have
a continuous lumen and which is still connected with the venous system in a number
of places. This plexus is the primordium of the lymph-sacs or lymph-hearts, from
which the lymphatics of the body arise and grow peripherally. As stated by Sabin
(1913), these buds in connection with veins form "a new type of vessels, namely,
lymphatics. These buds unite to form plexuses which develop into sacs. These
sacs may become lymph-hearts. From these lymph-sacs or hearts lymphatic

19

20 FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

capillaries gradually invade the body in orderly sequence in definite and character-
istic zones and layers. The growth is always in the capillary bed — that is, all
lymphatics develop as capillaries, and the earliest ones to develop become the first
lymph trunks or ducts." To make the story more complete (at least in the pig), in
the next stage these lymph-sacs break up and are converted into a coarse plexus
of lymph-channels, which in turn are transformed into lymph-nodes or glands.

As to the secondary lymphatic system — that is, the peripheral vessels — the de-
velopment of those of the skin (Sabin, 1904), small intestine (Heuer, 1909), lung (Cun-
ningham, 1916), heart and stomach (Cash, 1917, 1921) has already been studied.

The work of A. H. Clark (1912-13) on the fate of the jugular lymph-sacs gives
a general survey of the lymphatics of the anterior region of the body in the pig and
shows how the lymphatic channels of the head, neck, and thorax are related to the
primary lymph-glands. In the present paper it is desired to give a similar survey of
the lymphatics in the abdominal and pelvic regions and to determine more
definitely the fate of the primary sacs of the abdominal region. No attempt will be
made to present in detail the development of the lymphatics of the different organs.

Throughout the investigation the encouragement and assistance of Dr.
Florence R. Sabin have been a constant help, and my sincere appreciation and
thanks are due to her.

METHOD.

The material consisted of living pig embryos varying in length from 4 to 20
cm. The lymphatics were injected and the specimens cleared by the Spalteholz
method as described by Sabin (1915). Silver nitrate (2 per cent) and india-ink
were used for injections. With the latter the injections were nearly complete.
Some were made through the retroperitoneal sac, but more thorough injections of
the abdominal and pelvic regions were secured by introducing the needle into the
thoracic duct on the left side, a short distance posterior to and behind the arch of
the aorta, as described by Heuer (1909).

At about the stage of 7 cm., valves are being formed in the lymphatic vessels
and the primary sacs no longer remain homogeneous, but begin to break up into
plexuses of vessels which are the forerunners of the primary lymph-glands. For this
reason, beyond the stage of 8 cm., when it was difficult to get the injection mass to
pass the valves, the different organs were injected directly and the lymphatic
drainage to the several glands was thus indicated.

In the pig the origin of the lymphatics of the abdominal and pelvic regions can
be traced to lymphatics arising ventral to the aorta (fig. 1) and dorso-lateral to the
aorta (fig. 4). With the growth of the embryo and the beginning formation of the
lymph-glands this distinction can not be made out so readily, since in both cases the
glands tend to lie lateral to the aorta. When glands are definitely formed, those
arising from lymphatics dorso-lateral to the aorta in the main lie with those from
the ventral side.

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 21

LYMPHATICS ARISING VENTRAL TO THE AORTA.

RETROPERITONEAL SAC.

In the smaller embryos, beginning at about 23 mm., there is a primary lym-
phatic sac — the mesenteric (Lewis, 1906) or retroperitoneal sac, as it was termed
later (Baetjer, 1908) — situated outside of the peritoneal cavity. This sac lies
between the paired Wolffian bodies and gonads, just ventral to the point of renal
anastomosis of the subcardinal veins, and extends from the region of the superior
mesenteric artery to a point near the bifurcation of the aorta (fig. 1). Up to about
7 to 8 cm. it is a complete sac spreading sheet-like between the Wolffian bodies
(Heuer, 1909, figs. 3 and 4), and later between the kidneys.

Lewis (1906) describes the sac in rabbit embryos of 21 mm. as situated at the
root of the mesentery, just ventral to the point of renal anastomosis of the sub-
cardinal veins. Baetjer (1908) has shown that it has its origin in numerous venous
capillaries which, by increase in number and by fusion, form a sac which has definite
connections with the subcardinal veins. There follows a gradual obliteration of
these venous connections and for a short period the sac is an isolated structure with
irregular margins. By the stage of 25 mm. it has gradually developed connections
with the rest of the lymphatic system by an upgrowth of small lymphatic capillaries
along the margin of the aorta. Heuer (1909) describes three main connections with
the lymphatics dorsal to the aorta :

(1) There are several afferent vessels which arise from the mesial trunk as it
leaves the anterior end of the sac and course ventro-dorsad and slightly anterior to
enter the thoracic duct (fig. 2, Anas. ant.).

(2) The main efferent connection is a stout trunk (fig. 2, Anas, maj.) opposite
the anterior pole of the kidney, between the cisterna chyli and the anterior end of
the sac.

(3) A third connection is made from its posterior portion to the caudal part
or fan of the iliac sacs. Also, along the renal portion of the aorta there are numerous
anastomoses with the lymphatics dorso-lateral to the aorta.

Anterior portion of sac. — This large sac ventral to the aorta may be divided
in the region of the hilum of the kidney into an anterior and a posterior portion
(fig. 1). The anterior portion gives rise to two main trunks, the large anterior or
coeliac trunk and the small posterior or superior mesenteric trunk. Figure 1 shows
the ventral or retroperitoneal sac in a pig embryo 7.8 cm., after it had begun to
split into a plexus of lymphatic vessels, the forerunners of its transformation into
lymph-glands. In the pig it is at this stage a very extensive plexus, in contrast with
its earlier sac-like form (Heuer's fig. 4), and its main portions are seen to lie on
either side of the aorta, where they form a coarse network of channels. It will be
noted that the fold of mesentery and the cceliac trunk have been cut and the
stomach turned aside to expose the lymphatics from the anterior end of the sac.
At this stage the kidneys have developed into large structures and the Wolffian
bodies have begun to degenerate and are pushed to the side. The artist has clearly
shown the relationship of the various parts of the reproductive system.

22

FATE OP PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

The large coeliac trunk (figs. 1 and 2, T. coel.) immediately divides into a right
and a left branch, the latter soon breaking into the mesial trunk and the left trunk.
The left trunk, which includes the gastro-splenic trunk, is not shown in figure 1,
being on the other side of the fold of mesentery, but it can be well seen in figure 3.
The mass of lymphatics at the anterior end of the retroperitoneal sac courses
cephalad for a short distance and then turns ventrally, breaking up into several
main lymphatic branches. This ventral curve is shown in figure 2. A little way
beyond the origin of the mesial trunk, lymphatic vessels pass laterally to spread (at

V I c vmin.
V I

Diaph

61. suprar.

V.I. hep.

T. coel .
Liq.diaph.

V.l.ren.

Vlqon

Meson

T. il.com

Flo. 1. — Ventral view of pip em-
bryo (No. 73) 7.8 cm. long,
showing ventral or retroperi-
toneal sac broken up into a
coarse plexus, a stage about
midway between the earlier
complete sac and later group
of lymph-glands. Lymphat-
ics injected with india-ink '
through the thoracic duct. '
Drawing made from speci—
men by J. F. Didusch.

Liq Geniroinq.
Or. uret.
in urog.

T. hypo.

FATE OP PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

23

first in the form of a film, but in older embryos as distinct vessels) over the adrenal
capsule and the anterior pole of the kidney (fig. 3, V. 1. cap. gl. suprar.). As the
trunk twines ventrally around the body of the pancreas, lymphatics are given off to
the middle portion of that organ. These pancreatic vessels are not shown in figure
3, but would come off in front of the lower portion of the trunk of the lesser curvature
(T. c. V. min.). Laterally, vessels pass from the mesial trunk to the diaphragm
(fig. 1, V. 1. diaph.), uniting with those from the anterior end of the iliac sacs.

From the mesial trunk, as previously stated, there are several efferent vessels
emptying into the thoracic duct (fig. 2, Anas, ant.) On either side of these vessels
several long lymphatic channels run cephalad with the pulmonary ligament to
supply the lower portions of the lung (fig. 2, V. 1. pulm.). The largest trunk of the

D.thor

V.I. hep.

Vena umb.

D.omph.mes,

V I me3

Ramus Retroren..
S.il.

5. retroperil.

5. il.pors caud.

T.post.

FIG. 2. — Right lateral view of pig em-
bryo (No. 24) 6 cm. long, showing
connections and relative positions
of the primary lymph-sacs — that
is, of the retroperitoneal sao and
the iliac sacs. Lymphatics in-
jected with india-ink through the
thoracic duct. Drawing made
from specimen by J. F. Didusch.

_Vena
omph mea,
A umb.

24

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

mesial mass is the gastric trunk of the lesser curvature, which is described by Cash
(1921). This is shown in part in figure 1, but still better in figure 3, which is a
view of the stomach from below, designed to show the main mass of lymphatics of
the lesser and greater curvatures and special lymphatics of the pylorus. It shows
also the relation of the lymphatics to the pancreas. It is to be noted that the
spleen, with the lymphatic vessels in its capsule and mesentery, has been removed
and that the gastro-splenic trunk has been cut just as it is dividing into its two
terminal branches. The intestines have been removed, leaving only the stump of
the superior mesenteric trunk. This gastric trunk terminates at the lesser curvature
by dividing into two branches: (1) an ascending or left branch, which supplies the
cardiac region of the stomach, forms a network of periesophageal lymphatics, and
sends out vessels to the medial portion of the diaphragm; (2) a descending or right
branch, which likewise supplies the cardiac region, diaphragm and lesser curvature
(fig. 2, V. 1. diaph.; T. c. v. min.), and anastomoses with other vessels supplying the
diaphragm and lesser curvature.

The left trunk, of which the main branch is the gastro-splenic trunk (fig. 3),
passes ventrally to the fundus of the stomach. Near its origin it sends lymphatics
to the tail of the pancreas. It continues on to the fundus as the gastro-splenic
trunk, then divides beneath the posterior third of the spleen, one branch im-

V.l.c.v.maj.

Vl.pyl

V.I. hep.

V.I. corp.
pan.

T.qast.dexf
V. I. asc. duo

V.l.cap.ql.suprar.

V.l.capulpan

Fio. 3. — Pig embryo (No. 74), 6 cm. long. Ventral view of anterior end of retroperitoneal sac, showing the main gastric
trunks and the lymphatics of the pancreas. The spleen has been removed. Lymphatics injected with
india-ink through the thoracic duct. Drawing made from specimen by J. F. Didusch.

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 25

mediately breaking up to supply the region of the fundus, the other coursing along
the length of the spleen in its ligament to a point near the pyloric extremity. Here
it again divides, the smaller branch to continue to the pole of the spleen, the larger
turning obliquely toward the greater curvature of the stomach, where it forms a
T-shaped termination (fig. 3, V. 1. c. v. maj.) One limb passes toward the cardiac
pouch, breaking into many small vessels on the curvature and finally anastomosing
with the vessels supplying the fundus; the other limb passes toward the pylorus
and likewise drains the greater curvature and anastomoses with the lymphatics of
the pylorus. These anastomoses are well shown in figure 3. In none of the
specimens could an injection be secured which involved the parenchyma of the
spleen itself.

The right branch arises from the right half of the cceliac trunk and courses
ventro-laterally along the gastro-pancreatic line with the hepatic artery, breaking
into two branches — the hepatic vessels and the right gastric trunk (fig. 3, T. gast.
dex.). From the right gastric trunk the pyloric vessels (fig. 3, V. 1. pyl.) pass along the
pancreas to the pyloro-duodenal junction, where they turn abruptly forward over
the pylorus to anastomose with the lymphatics of the greater curvature . This trunk
also sends lymphatics to the body of the pancreas (fig. 3, V. 1. corp. pan.).

There are two sets of lymphatics going to the liver. The main mass passes
along with the hepatic artery, going behind the pyloric end of the stomach (figs.
1, 2, and 3; V. 1. hep.). These vessels unite with the other lymphatics accompany-
ing the biliary ducts to form the hepatic trunk (fig. 2, V. 1. hep.) . These lymphatics
course along with the blood-vessels within the liver in the perivascular and peri-
lobular tissue, but not to the lobule of the organ itself (Mall, 1901). The second
hepatic set or trunk arises just posterior to the large anterior or cceliac trunk. This
lateral trunk splits off laterally from the right side of the retroperitoneal sac toward
the portal vein, which it accompanies for a short distance, then branches at right
angles, sending vessels in both directions along that vein. Those passing posteriorly
unite with those going to the head of the pancreas. As they approach the liver,
these vessels along the portal vein join with those of the hepatic branch from the
cceliac trunk to enter at the hilum of the liver, although some accompany the biliary
ducts and go to the gall-bladder.

The other main ventral branch from the anterior end of the retroperitoneal sac
is the small posterior or superior mesenteric trunk, the stump of which is shown in
figure 3. This trunk arises from the sac at about the origin of the superior mesenteric
artery and its branches become the satellites of the branches of that artery in the
mesentery (fig. 2, V. 1. mes.). The trunk appears in figure 2, which is a view of the
lymphatics from the side, and shows quite well the relation of the retroperitoneal sac
to the iliac sacs, both as to position and connections. It also brings out the enor-
mous mass of lymphatic vessels arising from the anterior end of the retroperitoneal
sac. The trunk divides shortly, the smaller (duodenal) branch (figs. 2 and 3, V. 1.
asc. duo.) supplying the lower portion of that section of the small intestine and
anastomosing with the vessels from above. Lymphatics are also given off from
this duodenal branch to the head of the pancreas (fig. 3). The other, the mesen-

26 FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

teric branch (fig. 2, V. 1. mes.), is large and supplies the rest of the small intestine as
well as the coils of the large intestine, the ascending, transverse, and descending
colon.

Posterior portion of sac. — Arbitrarily, the retroperitoneal sac was divided at the
hilum of the kidney into an anterior and a posterior portion, but even at the anterior
pole the sac is seen to broaden so that it occupies the interval between the mesial
borders of the two organs. In earlier stages it extends to the hilum of the Wolffian
body. In later stages (7 to 8 cm.) vessels from the lateral side of the sac enter the
kidney at the hilum in company with other lymphatics from the dorso-lateral side
of the aorta (fig. 1, V. 1. ren.) . In earlier stages, when the Wolffian bodies are large
and completely hide the kidneys in a ventral view, the whole posterior lateral edge
of the sac sends a sheet of parallel blunt processes, transverse to the axis of the
embryo, which extend a short distance on both the dorsal and ventral capsules of
the Wolffian bodies, as well as entering the mesial border in company with the
blood-vessels of the glomeruli. At the stage shown in figure 1 the sac does not
extend to the mesial border of the Wolffian body, but only a short distance over the
ventral surface of the kidney, where numerous stout vessels accompanj' the meso-
nephric veins. As these veins atrophy the satellitic lymph-vessels disappear. The
remnants of the blunt processes are here shown passing obliquely backward to the
posterior part of the Wolffian body and leading to a characteristic plexus along the
medial ridge of the Wolffian body — the gonadal plexus (fig. 1, V. 1. gon.). Thus, of
the numerous blunt processes going to the Wolffian body, only those related to the
gonads and other reproductive structures persist; the others disappear and this
gonadal plexus becomes the genital branch, appearing in the stage shown in figure 1
as a fine plexus arising from the retroperitoneal sac at its lateral border, just cephalad
to the bifurcation of the aorta, and passing laterally to the posterior medial end of
the Wolffian body. Here part of the vessels turn sharply and course cephalad along
the mesial border of that organ forming a fine plexus about the gonadal blood-
vessels. This anterior branch continues to the hilum of the gonad which it supplies.
The genital branch also forms a network over the caudal end of the Wolffian body,
as well as giving off branches to the Mtillerian and Wolffian ducts and the gonadal
ligament. In the earlier forms the sac sends lymphatics to the medial side of the
umbilical cord, where they anastomose with those from the posterior portions of the
iliac sacs. In the later stages they fail to inject.

Thus the peripheral spread of the lymphatic vessels from the posterior part of
the retroperitoneal sac has been described; it remains only to mention the third
efferent connection, which arises from the posterior end of the retroperitoneal sac,
passes laterally to the outer side of the umbilicus, and then turns dorsally to join
the caudal ends of the iliac sacs.

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 27

LYMPHATICS ARISING DORSO-LATERAL TO THE AORTA.

Besides the paired anterior or jugular sacs there is another pair, the iliacs,
which grow along the dorso-medial edge of the Wolffian body, dorso-lateral to the
aorta, and drain the posterior half of the body. Figure 4 shows these two iliac sacs
slightly dorso-lateral to the aorta and also their relation to the cisterna chyli and
thoracic duct. It will be noted that there is an extensive plexus between the
two sacs, while the paler vessels in the depth are from the retroperitoneal sac.
The remnants of the Wolffian bodies show at the posterior borders of the kidneys.
The iliacs are connected with the jugulars by the cisterna chyli and thoracic duct.
Sabin (1912) has shown that the cisterna chyli and iliac sacs arise from the mesone-
phric veins of the vena cava.

In the younger specimens, in the region of the diaphragm the thoracic duct is
seen as two vessels, one on either side of the aorta, lying dorsally. These anasto-
mose freely in front of and behind the aorta by lateral vessels. At about the same
level as the anterior end of the retroperitoneal sac the vessels of the thoracic duct
unite to form posteriorly an irregular dilatation, the cisterna chyli, extending to a
point near the posterior poles of the kidneys.

From the lymphatics lying dorso-lateral to the whole length of the aorta — that
is, from the thoracic ducts, the cisterna chyli and the anastomosing iliac sacs-
other lymphatics are given off dorsally as satellites of the segmental arteries which
accompany the anterior and posterior rami. No vessels were observed, however,
to enter the vertebral canal with the spinal branch of the posterior ramus, which is
in accord with previous embryological studies, from which it was evident that the
central nervous system contains no lymphatics.

ILIAC SACS.

Caudal to the cisterna chyli, and draining into it, are two large trunks, one on
either side of the aorta. These are the paired iliac sacs, which in earlier stages
appear to be united. Small at the cephalic end where they form the cisterna chyli,
they broaden out gradually to form a rectangular sac between the dorso-medial
surfaces of the kidneys and show two expansions. The lateral one is at the hilum
of the kidney to which it sends vessels and a small trunk extending laterally in the
body- wall to a retrorenal position, then turning abruptly caudad to unite with a
trunk from the posterior portion (fig. 2, Ramus retroren.). In older embryos the
latter gradually disappears. The other expansion occurs at the posterior end and
may be called the caudal portion of the iliac sac. It is roughly fan-shaped, the
expansion being latero-posterior, the outer margin curving with the posterior pole
of the kidney (fig. 4, S. II. pars caud.).

In later stages, even at 5 cm., a distinct change is noted in the lymphatics dorso-
lateral to the aorta (fig. 4). The iliac sacs are no longer fused, but come to lie on
either side of the midline between the aorta and the medial edge of the kidney.
They empty through one or two trunks into the cisterna chyli in the region of the
adrenals. Their median border, accompanying the aorta, is straight and may
sometimes, even in the earlier stages, send anastomosing vessels to the other side.

28

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

With the growth of the embryo these sacs come to lie entirely lateral to the aorta in
approximation to and above the retroperitoneal sac, which likewise has become
situated lateral to the aorta.

These three sacs are connected by small channels which show in figures 2 and 4,
and when they are transformed into lymph-glands the result is a rather confusing

V I 3iaph

D thor.

V.l.ql. suprar.

61 suprar.'

V.l.coli des.

Rect.

V. I. sacr. lot

Ao.abd

Nod iliomqA

FIG. 4. — Dorsal view of specimen seen in figure 1, showing primary lymphatics dorsal and dorso-
lateral to the aorta, namely, the thoracic duct, cisterna chyli, and the paired iliac sacs.
Lymphatics injected with india-ink through the thoracic duct. Drawing from specimen
by J. F. Didusch.

mass of right and left juxta-aortic lymph-nodes which is made up of glands aris-
ing from lymphatics both dorsal and ventral to the aorta, and which extends from
just above the hilum of the kidney to the bifurcation of the aorta, forming the
main mass of lumbar nodes of the adult. Since the glands arise not only from
the ventral and dorsal sacs, but probably also along some of the anastomosing
vessels between these sacs, the resulting lymph-glands in the adult are exceedingly

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 29

difficult to analyze in terms of dorsal and ventral lymphatics. However, there are
glands arising from a small dorsal trunk or plexus which form the post-aortic or
prevertebral nodes, but these are few in number (fig. 5, Lymgl. praevert.).

From the cephalic end of the iliac sacs lymphatics pass forward, anastomosing
with those from the retroperitoneal sac to supply the diaphragm (fig. 4, V. 1. diaph.).
The adrenals receive several vessels (V. 1. gl. suprar.) which arise from this cephalic
end of the sacs in company with those which pass to the diaphragm. As mentioned
previously, the kidney receives at its hilum vessels from the iliac sac (fig. 4, V. 1. ren.).
There appears, therefore, to be a double supply to that organ, the vessels from the
iliac sacs and those from the posterior half of the retroperitoneal sac. These vessels
do not seem to penetrate very deeply into the kidney. Besfdes the vessels to the
hilum of the kidney, a rather stout trunk, broad at the base, arises at this point and
extends laterally in the dorsal wall. At times this is quite a large lateral expansion
from the iliacs and forms a retrorenal sac which gives off one or more branches
passing posteriorly at right angles to the body-wall, and anastomosing with similar
vessels from the caudal end of the iliacs (fig. 2, Ramus retroren.) . It also sends a few
vessels into the lateral body-wall, where a small plexus is formed which drains part
of the deeper layers of the wall. Very few if any superficial vessels drain into it.
As stated before, in older embryos these retrorenal branches could not be injected.

Caudal portion of sac. — The fan-shaped caudal portion of each iliac sac lies in
the body-wall between the posterior pole of the kidney and the bifurcation of the
aorta (fig. 4) . It extends laterally and dips ventrally on its mesial side. Later it
forms an inverted Y-shaped expansion, one branch passing laterally along the
posterior pole of the kidney, the other going directly caudal and then turning
ventrally to supply the lower portion of the body and legs.

From the lateral branch, in the earlier stages, a few vessels or a small trunk
course cephalad to anastomose with the vessels from the retroperitoneal sac.
Laterally, this caudal portion sends a group of lymphatics through the body-wall
to form a superficial node just above the crest of the ilium (the ilio-inguinal node)
which is characteristic for the pig, and drains the skin of the posterior half of the
body and tail (Sabin, 1904). From the medial side of the posterior branch of the
caudal portion one or two vessels pass caudally in the dorsal wall along with the
lateral sacral vein (fig. 4, V. 1. sacr. lat.). On the postero-medial side of the sac a
short trunk extends ventrally, medial to the umbilical artery, to anastomose with
another trunk from the retroperitoneal sac.

The posterior trunk (fig. 4, T. post.), which drains the lymphatics of the legs
and some of the pelvic organs, is the continuation of the medial branch of the caudal
portion, which courses ventrally and laterally, passing lateral to the umbilical
artery along the common iliac artery (fig. 1 , T. il. com.) . Only the proximal portion
of the posterior trunk is shown in figure 4, for as it turned ventrally it became less
and less distinct from the dorsal side of the embryo; but in figure 1, which is a
ventral view of the same specimen, the continuation of the posterior trunk is well
seen. It divides into several branches, two of which are proximal. From its medial
edge, near its origin, the trunk sends a vessel ventro-medially to the lateral side of

30 FATE OP PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

the umbilical cord, where it anastomoses with a similar vessel from the retroperi-
toneal sac. It also sends lymphatics ventrally, in company with the vessel to the
umbilical cord, to break up into smaller vessels over the posterior pole of the Wolffian
body, where it evidently supplies part of the Mullerian and Wolffian ducts (shown on
the left side of fig. 1) just at the inferior pole of the Wolffian body. This seems an
important point, for, as will be noticed in figure 1, the lymphatics of the gonad
itself are related primarily to one set — the ventral lymphatics, derived from those
of the Wolffian body — while the lymphatics of the accessory genital organs are shown
to be embryonically related to two sets of lymphatics — those from the ventral or
retroperitoneal sac, and those from the dorso-lateral or iliac sacs.

The posterior trunk then continues as a definite group of lymphatics, the com-
mon iliac trunk following the artery of that name. It terminates by dividing into
two smaller branches, the more important of which is the femoral or external iliac.
These terminal vessels of the posterior trunk are satellites of the main arteries of the
pelvis and hind-limb (fig. 1). In the stages at which the injections were made, the
first or primitive arrangement of the blood-vessels of the posterior part of the body
has changed, so that the blood comes from the aorta by way of the two common
iliac vessels, which arise dorso-laterally just anterior to the origin of the umbilical
arteries. Each passes caudally, lateral to the umbilical cord, sending a branch (the
ilio-inguinal) around the pole of the kidney. At or just beyond the passage into
the pelvic wall it divides into two main vessels. One of these, the internal iliac or
hypogastric, which runs dorso-medially, divides into a number of terminal branches.
The other, the femoral or external iliac, continues ventrally to the front of the hind-
limb. At the proximal end of the femoral the now a trophic sciatic artery arises.

The common iliac lymphatic trunk broadens considerably at the bifurcation of
the common iliac artery and then divides into a ventral and dorsal branch. The
dorsal trunk — the internal iliac or hypogastric (fig. 1, T. hypo.) — accompanies the
internal iliac artery and terminates by breaking into a number of vessels, satellites
of the lesser arterial branches. A short distance from its origin the ventral branch—
the femoral or external iliac (T. il. ext.) — sends a few lateral branches with the
sciatic artery. It divides shortly, many of the vessels turning medially to form the
inguinal trunk (fig. 1, V. 1. ing.) which drains a plexus of lymphatics along the
inguinal ligament. This plexus is converted into two large elongated nodes, one
situated more deeply than the other. These break up into a group of deep and super-
ficial nodes which drain the lower abdominal wall, external genitalia, perineum,
mesial side of the thigh, and practically the whole of the lower extremity. The
remainder of the vessels of the femoral trunk (fig. 1, V. 1. fern.) continue caudally,
following the femoral artery in its course in the leg.

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 31

Relation of the Peripheral Lymphatics to the Primary System.

I. Lymphatics arising ventral to the aorta.

A. Retroperitoneal sac.

1. Anterior portion.

(a) Cceliac trunk.

(I) Left branch.

(A) Mesial trunk.

(1) Adrenal (capsule) and kidney (ant. pole).

(2) Pancreas (body).

(3) Diaphragm.

(4) Anterior anastomosis to thoracic duct.

(5) Lung.

(6) Trunk of lesser gastric curvature.

(a) ascending (left) periesophagus, cardia, diaphragm.

(6) descending (right) cardia, diaphragm, lesser curvature.

(B) Left trunk.

(1) Pancreas (tail).

(2) G astro-splenic trunk.

(a) Fundus of stomach.

(6) Splenic capsule.

(c) Vessels of greater gastric curvature ; (1) fundus; (2) pylorus.

(II) Right branch.

(A) Hepatic vessels.

(B) Right gastric vessels; (1) pylorus; (2) pancreas (body).
(6) Lateral trunk.

(I) Portal vein.
(II) Biliary ducts.

(c) Superior mesenteric trunk.

(I) Duodenal branch; (A) pancreas (head).
(II) Mesenteric branch.

(d) Main anastomosis to cisterna chyli.

2. Posterior portion.

(a) Lateral branches.
(I) Kidney.
(II) Wolffian body.

(III) Genital branch.

(A) Gonad.

(B) Wolffian body (caudal).

(C) Mullerian and Wolffian ducts.

(D) Caudal gonadal ligament.

(IV) Umbilical cord (medial side).

(V) Posterior anastomosis to caudal end of iliac sacs.
(6) Medial branch.

(I) Descending colon.

II. Lymphatics arising dorso-lateral to the aorta.

A. Thoracic duct.

B. Cisterna chyh.

C. Iliac sacs.

1. Diaphragm.

2. Adrenal.

3. Kidney.

4. Retrorenal branch.

5. Caudal portion.

(a) Anastomosis with retrorenal branch.

(b) Ilio-inguinal node.

(c) Lateral sacral branch.

(d) Posterior anastomosis with retroperitoneal sac.

(e) Posterior trunk.

(I) Proximal branches.

(A) Umbilical cord (lateral side).

(B) Wolffian body (posterior pole). (1) Mullerian and Wolffian ducts.

(C) Rectum.

(II) Common iliac trunk.

(A) Hypogastric trunk.

(B) External iliac trunk; (1) sciatic vessels; (2) femoral vessels; (3) inguinal vessels.

32 FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

The schema given on the preceding page shows graphically the relation of the
peripheral lymph-vessels to the primary lymphatic system. In general the lym-
phatics arising ventral to the aorta (the retroperitoneal sac) drain the whole or part
of the lung, diaphragm, liver and biliary passages, stomach, small and large intes-
tine, capsule of spleen, pancreas, kidney, gonad, Mullerian and Wolffian ducts,
umbilical cord, and Wolffian body.

From the lymphatics arising dorso-lateral to the aorta (the iliac sacs) are
supplied, either wholly or in part, the diaphragm, body-wall, adrenals, kidney,
bladder, umbilical cord, Mullerian and Wolffian ducts, and the entirf posterior half
of the body.

The structures which drain into the primary lymphatics arising both ventral
and dorso-lateral to the aorta are the diaphragm, kidney, and Mullerian and Wolf-
fian ducts.

RELATION OF THE LYMPH-GLANDS TO THE PRIMARY AND SECONDARY LYMPHATIC

SYSTEMS.

Sabin (1913) has shown that the primary lymph-nodes develop from the pri-
mary lymph-sacs; that the pre-aortic or retroperitoneal glands develop from the
retroperitoneal sac; and that from the iliac sacs there arises a chain of small nodes
lateral to the aorta and a large group of glands on either side of the aorta opposite
its bifurcation. Dorsal to the aorta are the prevertebral nodes extending from the
lower end of the cisterna chyli to the aortic bifurcation.

Of the secondary lymphatic nodes developing along the lymphatic vessels, she
states that the mesenteric glands are secondary for the retroperitoneal sac. In the
pig the secondary glands from the iliac sacs are simple and two in number — the
ilio-inguinal gland (very characteristic for this animal) and the inguinal group of
glands.

In this study, injections, as nearly complete as possible, were made of pig
embryos 20 cm. or more in length to determine what glands receive the drainage
from the abdominal and pelvic viscera. At this stage the glands are easily seen in
the gross and, by injecting the organs with india-ink, their relations to the different
glands can be readily demonstrated. Figure 5 is a composite drawing of an embryo
pig 20 cm. long, viewed from the ventral side, to show the grouping and positions of
the various lymph-glands relative to each other and to the different abdominal
organs. Those glands formed from lymphatics arising ventral to the a.orta are
indicated in solid color; those from lymphatics dorso-lateral to the aorta are
stippled. The glands along the vessels posterior to the bifurcation of the aorta,
however, which are in solid black, belong to the group arising from lymphatics
dorso-lateral to the aorta and were not stippled because of the perspective.

In the abdomen are glands — the pre-aortic — extending from above the cceliac
axis to the bifurcation of the aorta. Anteriorly these are grouped into the cceliac
(fig. 5. Lymgl. ccel.) and main mesenteric glands (Lymgl. mes. sup.) and posteriorly
into the lumbar pre-aortics (Lymgl. praeaorticse) . Lateral to the aorta are the

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

33

juxta-aortic glands; on the right side some are prevenous, others retrovenous, in
relation to the vena cava (fig. 5). Posterior to the aorta are the prevertebral
glands. Glands are formed along the external iliac artery and smaller ones about
the internal iliac or hypogastric artery.

Lymgl. fundi

LymaJ lien.
Lymql cv. min

Lymql coel.

Lymql.mes.eup.
Lymql juxla-intes

Lymgl. praeven.
Vena Cava
Lymfjl.praevert.

Lyrrigl. i I. COTI.
Lyrrigl. ilioing.

Lymgl juxta-aorlicoe

Lymql prae-aorficoa
Ovorium

Colon "descendens

Uterus

Flo. 6. — Composite diagrammatic drawing viewed from the ventral side of a pig embryo 20 cm.
long. The glands arising from the retroperitoneal sac are in solid color, those
arising from the iliac sacs are stippled. However, the glands along the^vessels
posterior to the bifurcation of the aorta, which are in solid black, belong^to.the
group arising dorsolateral to the aorta, but were not stippled because of thejper-
epective. Drawing made by J. F. Didusch.

34 FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

The retroperitoneal sac is transformed into the glands of the coeliac axis and,
at the origin of the superior mesenteric artery, into the main mesenteric glands and
also into the preaortics, the prevenous, and part of the juxta-aortic glands. The
iliac sacs become the retrovenous, prevertebral, and the remainder of the juxtaaortic
glands. The glands constituting the large group opposite the bifurcation of the aorta,
as well as the proximal iliac glands, are formed from the caudal ends of the sacs.

DRAINAGE OF THE VISCERA.

It has been shown that in the earlier stages the stomach has three main systems
of drainage: (1) the trunk of the lesser gastric curvature, (2) the gastrolineal trunk,
and (3) the right gastric branch. This system is likewise indicated by the lymph-
glands, the three main channels of drainage being indicated by arrows in figure 5.

1. In a pig of 20 cm., when the injection is made about the periesophageal
region, the vessels shown on the anterior side of the fundus and about the cardiac
orifice drain around the neck and then posteriorly to several large nodes situated
slightly posterior and to the left of the orifice on the pillar of the diaphragm (fig. 5,
Lymgl. fundi.). Other lymphatic vessels, more from the middle portion of the
stomach, drain to a gland on the right of the cardiac orifice (Lymgl. c. v. min.).

2. Lymph from the inferior part of the fundus drains along the posterior margin
of the splenic ligament, following the lesser gastric vessels of the lienal artery, where
it empties into several glands lying in the splenic ligament. From these nodes it
drains posteriorly into several large glands near the left periesophageal glands (fig.
5, Lymgl. fundi). Part of the lymph from the greater curvature drains with that
from the gastro-epiploic vessels to the middle region of the spleen, where it enters
several glands. Connection is demonstrated between these glands and nodes situ-
ated along the splenic vessel. Not only these glands, but also the periesophageals,
follow the splenic artery to the cceliac axis (fig. 5, Lymgl. ccel.). From the superior
and inferior medial regions of the stomach, vessels run toward the lesser curvature,
along with the branches of the left gastric artery, to drain with a number of smaller
nodes which are located to the left of the esophageal ring (fig. 5, Lymgl. c. v. min.)
and drain, along with the right periesophageals, to the glands about the cceliac axis.

3. Lymphatics injected on the superior surface of the stomach near the pylorus
drain in the direction of the lesser curvature, but run in the lesser omentum,
turning with the vessels and emptying into very large glands situated along the
hepatic artery (fig. 5, Lymgl. hep.). These hepatic glands also have as afferents the
vessels of the greater curvature ; that is, the lymphatics of the greater curvature pass
toward the pylorus accompanying the right gastro-epiploic artery (fig. 5, Lymgl.
gastroepip.), following that vessel from its junction with the gastro-duodenal to
the hepatic artery, where they drain into these large nodes. Thus, about the
cceliac axis are a number of glands that in the earlier stages were in the form of
trunks from the anterior end of the retroperitoneal sac, and they, too, show three
main systems of drainage — a left, a mesial, and a right.

In the lesser omentum are small lymph-glands draining part of the liver. The
drainage from these nodes is seen to accompany the vessels from the pylorus which

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 35

follow the right gastric artery to empty also into the large hepatic glands, and from
these into the cceliac glands. When the gall-bladder was injected, some vessels
were seen to follow the course of the common duct to the duodenum, then to
accompany the portal vein to one of several large nodes situated about that vessel
in the angle of the pancreas (fig. 5, Lymgl. d. choled.). Smaller glands were also
seen along this vein peripheral to the larger ones. The efferents from these glands
go to one or two pre-aortic nodes lying between the coeliac axis and the superior
mesenteric nodes (fig. 5). Here, too, as in the earlier stages, there is a double
drainage of the liver.

In these later stages the lymphatics of the spleen failed to inject. Those of the
ligament, however, drain into glands at the hilum, from which the efferent flow is
through glands which are disposed about the splenic artery.

The small intestine, large intestine, and descending colon drain into the large
mesenteric glands, of which there are several groups. The largest and most im-
portant of these are the juxta-intestinal glands (fig. 5, lymgl. juxta-intes.) which
are the first to receive the lymph coming from the intestine. They are very large
prominent, elongated glands, lying close together, their long axes parallel to that
of the intestine. They form a distinct nodular rim or collar to the outer part of the
mesentery. Efferents from the juxta-intestinal nodes pass through intermediate
glands in the mesentery about the origin of the superior mesenteric artery (Heuer,
1909). These retromesenteric glands then receive afferents from the head of the
pancreas, duodenum, and intestine. This system of secondary lymph-glands of the
intestine, with the peripheral nodes very large and prominent, and the proximal
ones smaller and very limited in number, is quite a contrast to the system of the
stomach and neighboring organs, in which the glands increase in size and impor-
tance the nearer they approach to their termination about the coeliac axis.

The lymphatics of the pancreas could not be injected directly, but there are
indications that this organ has at least four collecting trunks: (1) The drainage
from its tail is to the splenic group of glands; (2) the posterior part of the body
sends some efferents to the periesophageal nodes; (3) the duodenal portion drains
along with the lymph from the pyloric portion of the stomach to the large hepatic
nodes; (4) the head of the pancreas sends efferents to the superior mesenteric
glands. Bartels (1906) was able to inject the lymphatics of the pancreas in new-
born animals, especially the dog, and in a later paper (1907) he described their
drainage into the regional glands, which, in general, corresponds to the drainage
indicated above.

It is evident from the earlier stages that the adrenals have a capsular drainage
to the cceliac group of glands, while the lymph from the organ itself drains into a
gland formed from the cephalic end of the iliac sac, i. e., a juxta-aortic gland which
is seen lying anterior to the renal glands (fig. 5, Lymgl. g. suprar.).

Although not definitely determined, it would seem from the double type of
earlier injections that the lymphatics from the kidney would drain into the pre-
aortic and perhaps post-aortic, but mainly into the juxta-aortic glands. Several
large nodes can be easily seen at the hilum lying behind the renal vessels, suggesting

36 FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

their origin from the iliac sacs, and a fairly large node anterior to the vessels suggests
its origin from the retroperitoneal sac. The separation of the juxta-aortic glands
by the vena cava, into a prevenous and a retrovenous group which were formed
from lymph-sacs ventral and dorso-lateral to the aorta, is shown in the right side
in figure 5, where the double drainage from the kidney to these two sets of glands
at the hilum is indicated.

In young embryos, vessels from the region of the inferior mesenteric artery pass
to the colon, and in later stages glands are seen about the inferior mesenteric artery,
evidently draining the descending colon.

The drainage from the ovaries and testes is by vessels that accompany the
correspondingly named arteries, first through one or two secondary nodes, then to
the lumbar region, where these lymphatic trunks turn medially and run toward
their terminal glands, which are situated ventral or ventro-lateral to the aorta and
vena cava, extending from the bifurcation of the aorta to about the level of the
posterior pole of the kidneys; that is, to pre-aortic, prevenous, and juxta-aortic
glands in the lumbar region. To these same glands drains part of the lymph from
the body of the uterus and Fallopian tubes in the female and from the epididymis
in the male. These glands have been formed from the root of the genital branch
as it comes off from the posterior end of the retroperitoneal sac (fig. 1, V. 1. gon.).
When the horns of the uterus are injected, the flow is both toward the cervix and
peripherally toward the tubes. The medial portion of the horn and part of the
cervix drain to a very large gland at the bifurcation of the external and internal
iliac arteries. Thus the lymph from the sexual glands has only one channel of flow,
i. e., to the glands that were formed from the posterior portion of the ventral or
retroperitoneal sac. The lymph from the excretory canals of the sex-glands, on
the other hand, has a double course of flow; that from the structures originating
proximal to the sex-glands goes to the glands formed from the ventral sac, while
the portions of the excretory canals distal to the sex-glands drain to glands arising
from the primary lymphatics dorso-lateral to the aorta.

Injections of the bladder showed lymphatics draining to the external iliac
glands. Drainage from certain regions of the bladder passed through intermediate
lateral vesicle glands peripheral to the iliac nodes. Occasionally a vessel could be
injected to the glands at the bifurcation of the aorta.

The lymphatics from the posterior trunk of the iliac sac become transformed
into a chain of lymph-glands which are disposed fairly regularly about the blood-
vessels (fig. 5, Lymgl. il. com.). In general, they are similar to those described by
Poirier and Cuneo (1904), except that there is a characteristic ilio-inguinal node in
the pig which is not seen in man.

The inguinal glands drain the superficial lymphatics over the lower abdomen,
external genitalia, ventral side of the tail, the leg, and the medial side of the thigh.
The efferents of the inguinal glands are distal external iliac nodes.

An ilio-inguinal node (fig. 5, Lymgl. ilio-ing.), characteristic of the pig, lies in
the course of the ilio-lumbar artery near the crest of the ilium. It drains the super-
ficial lymphatics of the posterior body-wall, down over the hip and the root of the
tail (Sabin, 1904).

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC. 37

SUMMARY.

In the posterior half of the body, the first stage in the formation of a connected,
well-developed primary lymphatic system in the pig is the presence of primary
lymph-sacs, lying ventral and dorsal to the aorta. The ventral sac is the large
retroperitoneal sac, which extends from the region of the cceliac axis to the bifurca-
tion of the aorta. Dorsal, or dorso-lateral, to the aorta is the remainder of the
primary system — the paired iliac sacs — which extend from the region of the adrenals
to the bifurcation of the aorta.

According to their origin from these primary sacs, lymphatic vessels t>f the
organs and structures in the abdominal and pelvic regions fall into two great
groups, dependent upon whether they arise from the primary lymphatics ventral
to the aorta, or from the primary lymphatics dorso-lateral to the aorta. As a
general statement, it may be said that the ventral sac supplies the abdominal
viscera, while the dorso-lateral sacs supply the retroperitoneal structures — the
viscera, the body-wall, and the lower extremities — although, on account of consider-
able overlapping, this generalization must be regarded as only approximate.

Thus, from the retroperitoneal sac are given off lymphatics which drain the
whole or part of the lung, diaphragm, liver and biliary passages, stomach, small and
large intestine, capsule of spleen, pancreas, kidneys, Wolffian bodies, gonads, Miil-
lerian and Wolffian ducts, and umbilical cord. The dorso-lateral lymphatics — that
is, the paired iliac sacs — drain wholly or in part the diaphragm, body-wall, adrenals,
kidneys, bladder, Miillerian and Wolffian ducts, umbilical cord, and the entire
posterior half of the body. Only a few structures, notably the diaphragm, the
kidneys, and the Miillerian and Wolffian ducts, which later become the excretory
canals of the sex-glands, have lymphatics arising from both the ventral and the
dorso-lateral sacs.

On following the development of the lymphatics in these posterior regions of
the body, this early system of drainage is found to continue into the adult stage of
lymphatic growth. The primary sacs change into lymph-glands which receive
afferent vessels from the same organs and structures as did their parent sac. To a
considerable extent the position of the primary lymph-glands, differentiated from
the primary sacs, falls into two groups — the glands ventral and ventro-lateral to the
aorta, and those dorsal and dorso-lateral to the aorta. Here, too, in the adult stage
of lymphatic growth, the general statement can be made that the organs and
structures situated within the abdominal cavity have their lymphatic drainage into
the group of glands whose position is ventral or ventro-lateral to the aorta. On the
other hand, those structures and organs lying outside of the abdominal cavity, as
well as the entire posterior half of the body, have their lymphatic drainage eventually
into the group of glands situated dorsal or dorso-lateral to the aorta.

38

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

ABBREVIATIONS.

Anas, ant.,

Anas, maj.,

Ao. abd.,

A. il. com.,

A. umb.,

Cist, chy.,

Diaph.,

D. meson.,

D. omph. mes.,

D. thor.,

Gl. suprar.,

Gon.,

Lig. caud. gon.,

Lig. cran. gon.,

Lig. diaph. meson.,

Lig. genitoing.
Lymgl. coel.,
Lymgl. c. v. min.,

Lymgl. d. choled.,

Lymgl. fundi,

Lymgl. gastroepip.,

Lymgl. gl. suprar.,

Lymgl. hep.,

Lymgl. il. com.,

Lymgl. ilioing.,

Lymgl. juxta infest.,

Lymgl. juxta aortica

Lymgl. lien.,

Lymgl. meg. sup.,

Lymgl. praeaorticse,

Lymgl. praeven.,

Lymgl. praevert.,

Lymgl. retroven.,

Meson.,

Nod. ilioing.,

Oe.,

Or. uret.,

Ovid.,

Pyl.,

Ramus retroren.,

Rect.,

Ren.,

anterior anastomosis.

major anastomosis.

abdominal aorta.

common iliac artery.

umbilical artery.

cisterna chyli.

diaphragm.

mesonephric duct.

omphalo-mesenteric duct.

thoracic duct.

suprarenal gland.

gonad.

caudal gonadal ligament.

cranial gonadal ligament.

diaphragmatic ligament of the

mesonephros.
genito-inguinal ligament,
cceliac lymph-gland,
lymph-gland of le&ser curvature of

stomach.

lymph-gland of biliary duct,
lymph-gland of fundus.
gastro-epiploic lymph-gland,
suprarenal lymph-gland,
hepatic lymph-gland,
common iliac lymph-gland,
ilio-inguinal lymph-gland,
juxta-intcstinal lymph-gland,
juxta-aortie lymph-gland,
splenic lymph-gland,
superior mesenteric lymph-gland,
pre-aortie lymph-glands,
prevenal lymph-glands,
prevertebral lymph-glands,
retrovenal lymph-glands,
mesonephros.
ilio-inguinal node,
oesophagus,
orifice of ureter,
oviduct,
pylorus.

retrorenal branch,
rectum,
kidney.

S. il.,

S. il. pars caud.,

Sin. urog.,

T. coel..

T. c. v. min.,

T. gast. dex.,
T. gastrolien.,
T. hypo.,
T. il. com.,
T. il. ext.,
T. mes. sup.,
T. post.,
Ur.,
V. 1. asc. duo.,

V. 1. cap. gl. suprar.,

V. 1. caput pan.,
V. 1. corp. pan.,
V. 1. coli des.,
V. 1. c. maj.,

V. 1. c. min.,

V. 1. diaph.,

V. 1. fern.,

V. 1. fundi,

V. 1. gl. suprar.,

V. 1. gon.,

V. 1. hep..

V. 1. h. mebon.,

V. 1. ing.,

V. 1. mes.,

V. I. pulrn.,

V. 1. pyl.,

V. 1. rect.,

V. 1. ren.,

V. 1. sacr. lat.,

Vena omph. mes.,

Vena umb.,

Venl.,

Ves. urin.,

iliac sac.

caudal part of iliac sac.

urogenital sinus.

coeliac trunk.

trunk to the lesser curvature of
stomach.

right gastric trunk.

gastro-splenic trunk.

hypogastric trunk.

common iliac trunk.

external iliac trunk.

superior mesenteric trunk.

posterior trunk.

ureter.

lymph-vessel of the ascending du-
odenum.

lymph-vessel of the suprarenal
capsule.

lymph-vessel of head of pancreas.

lymph-vessel of body of pancreas.

lymph- vessel of descending colon.

lymph-vessel to greater curvature of
stomach.

lymph-vessel to lesser curvature of
stomach.

lymph-vessel of diaphragm.

femoral lymph-vessel.

lymph-vessel of fundus.

lymph-vessel of suprarenal gland.

gonadal lymph-vessel.

hepatic lymph-vessel.

lymph-vessel at hilum of mesone-
phros.

inguinal lymph-vessel.

mesenteric lymph-vessel.

pulmonary lymph-vessel.

pyloric lymph-vessel.

lymph-vessel of rectum.

renal lymph-vessel.

lateral sacral lymph-vessel.

omphalo-mesenteric vein.

umbilical vein.

stomach.

urinary bladder.

FATE OF PRIMARY LYMPH-SACS IN ABDOMINAL REGION OF PIG, ETC.

39

BIBLIOGRAPHY.

BAETJEH, WALTER A., 1908. On the origin of the mesen-
teric sac and thoracic duct in the embryo pig.
Amer. Jour. Anat., vol. 8, 303.

BARTEI.S, PAUL, 1906. Ueher die Lymphgefasse des
Pankreas. II. Das feinere Verhalten der
lymphatischen Verbindungen zwischeu Pan-
kreas und Duodenum. Arch. f. Anat. u.
Physiol., Anat. Abth.

, 1907, Ueber die Lymphgefasse deg Pankreas.

III. Die regionaren Drilsen des Pankreaa beim
Menschen. Ibid., Anat. Abth.

CASH, J. R., 1917. On the development of lymphatics in
the heart of the embryo pig. Anat. Rec., vol.

13, 451.

, 1921. On the development of the lymphatics in

the stomach of the embryo pig. Contribu-
tions to Embryology, vol. 13, Carnegie InsL.
Wash. Pub. No. 276.

CLARK, A. H., 1912. On the fate of the jugular lymph-sacs,
and the development of the lymph-channels in
the neck of the pig. Amer. Jour. Anat., vol.

14, 47.

CLARK, E. R. and E. L., 1920. On the origin and early
development of the lymphatic system of the
chick. Contributions to Embryology, vol. 9,
Carnegie Inst. Wash. Pub. No. 272.

CUNNINGHAM, R. S., 1916. On the development of the
lymphatics of the lung in the embryo pig.
Contributions to Embryology, vol. 4, Carnegie
Inst. Wash. Pub. No. 224.

HEUER, GEORQE J., 1909. The development of the lym-
phatics in the small intrstine of the pig.
Amer. Jour. Anat., vol. 9. 91.

HUNTINQTON, G. S., and C. F. W. McCLURE, 1910. The
anatomy and development of the jugular
lymph-sacs in the domestic cat. Amer. Jour.
Anat., vol. 10, 177.

LEWIS, FREDERICK T., 1906. The development of the
lymphatic system in rabbits. Amer. Jour.
Anat., vol. 5, 95.

MALL, F. P., 1901. On the origin of the lymphatics of the
liver., Johns Hopkins Hosp. Bull., vol. 12, 146.

POIRIER, P., and B. CUNEO, 1904. The lymphatics. Trana.
by C. H. Leaf. Poirier and Charpy, Treatise of
Human Anatomy.

SABIN, F. R., 1901. On the origin of the lymphatic system
from the veins and the development of the
lymph-hearts and thoracic duct in the pig.
Amer. Jour. Anat., vol. 1, 367.

, 1904. On the development of the superficial

lymphatics in the skin of the pig. Ibid., vol.
3, 183.

, 1912. On thf origin of the abdominal lymphatics

in mammals from the vena cava and the renal
veins. Anat. Rec., vol. 6, 355.

, 1913. The origin and development of the lym-
phatic system. Johns Hopkins Hosp. Re-
ports, monograph, new series, vol. 5.

, 1915. On the fate ol the posterior cardinal veins

and their relation to the development of the
vena cava and azygos in the embryo pig.
Contributions to Embryology, vol. 3, Car-
negie Inst. Wash. Pub. No. 223.

, 1920. Studies on the origin of blood-vessels and

of red blood-corpuscles as seen in the living
blastoderm of chicks during the second day of
incubation. Contributions to Embryology,
vol. 9, Carnegie Inst. Wash. Pub. No. 272.

STOCKARD, C. R., 1915. The origin of blood and vascular
endothelium in embryos without a circulation
of the blood, and in the normal embryo.
Amer. Jour. Anat., vol. 18, 227.

STREETER, G. L., 1920. A human embryo (Mateer) of the
presomite period. Contributions to Embryol-
ogy, vol. 9, Carnegie Inst. Wash. Pub. No. 272.

CONTRIBUTIONS TO EMBRYOLOGY, No. 59.

RELATIVE WEIGHT AND VOLUME OF THE COMPONENT PARTS

OF THE BRAIN OF THE HUMAN EMBRYO AT DIFFERENT

STAGES OF DEVELOPMENT,

BY GEORGE B. JENKINS,
Department of Embryology, Carnegie Institution of Washington.

With twelve text-figures and one chart.

41

RELATIVE WEIGHT AND VOLUME OF THE COMPONENT PARTS

OF THE BRAIN OF THE HUMAN EMBRYO AT DIFFERENT

STAGES OF DEVELOPMENT,

INTRODUCTION.

The following volumetric study of the developing encephalon in the human
embryo was undertaken as a supplementary part of a program of investigations
in the morphology and differentiation of the central nervous system with which I
have been engaged for several years. The work was made possible through the
kindness of Dr. Streeter who, in furtherance of its purpose, has placed at my
disposal the collection of embryos of the Carnegie Laboratory of Embryology, to-
gether with models, notes, and such other related material as has been accumulated.
Under these conditions it was possible to secure a sufficient number of develop-
mental stages for the investigation, including the consideration of the relative rate
of growth of the various parts of the encephalon during the whole period of intra-
uterine life. It is with pleasure that I take this opportunity to acknowledge my
indebtedness for the hospitality, aid, and interest that have been extended to me on
the part of the staff of this laboratory.

The question of growth of the animal body as a whole and of its various com-
ponent parts has occupied the minds of many investigators over a long period of
time, and yet in the mass of literature which has resulted from these studies com-
paratively little is to be found that deals with the volumetric development of the
central nervous system in the human embryo. A study of the literature on the
subject of weight, size, degree of development, etc., of the brain reveals the fact
that practically all of the observations reported have dealt with postnatal material
and the great majority with adult brains, secured in the main from various institu-
tions, such as asylums for the insane, almshouses, and homes for the aged. The
great questions seem to have been those concerning body size, the attempted
correlations between brain size and intellectuality, and the controversy over the
differences in brain size in the two sexes. Not a few reports frankly deal with
grossly pathological material. Much additional work, of anthropological value
mainly, has been done along the lines of craniometry, notably by West (1894), who
made a number of observations upon school children of both sexes, recording the
head measurements as a part of the general growth conditions, Lewandowsky
(1910), and Hrdlicka (1919), who treat of the more pertinent question of the
correlation between skull size and cranial capacity.

Aside from these latter cases the bulk of the work reported may be roughly
grouped into four divisions :

(a) The study of brains of individuals of superior intellectual attainments
(Spitzka, 1907; Marshall, 1892; Manouvrier, 1885).

43

44 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

(&) The tabulation of the brain-weights and measurements obtained from
large series of cases, including material of various races and ages and representing
both sexes (Boyd, 1861; Donaldson, 1895; Ziehen, 1899, 1903; Vierordt, 1893).

(c) Statistical studies in which these tables are used to determine the correla-
tions existing between brain-size and age, brain-size and intellect, and the compar-
ison of brain, body, and intellect in the two sexes at various ages and in different
races, as has been so ably done by Pearl (1905a, 19056) who, in addition to the above
material, used the tables reported by Bischoff (1880), G. Retzius (1900), Marchand
(1902), Matiegka (1903), and Boyd-Marshall, all as reported by Donaldson (1895).
The employment of such biometric methods gives results of great value to the
investigator, since it enables him to make a great number of comparisons and to
tabulate and express his results with a clarity and an exactness not possible by any
other method.

(d) This last class would embrace the relatively limited number of observa-
tions dealing with material sufficiently young to come within the scope of the
present paper, and includes extracts from Major Boyd's (1861) monumental work
on approximately 2,600 cases autopsied at Marleybone Workhouse and the Somerset
Lunatic Asylum. These cases were grouped, according to age, into 18 periods, the
first 3 of which are of value in this connection. They comprise 48 premature still-
born, 83 term still-born, and 90 new-born infants. The various measurements as
given by Boyd were the total body-length from the vertex to the inner side of foot
behind the ball of the great toe, and the following head measurements : circumference
around the occipital protuberance to the space behind the eyebrows; transverse,
from the opening of one ear, over the vertex to the opening of the other. Doubtless
the method of subdividing the encephalon employed by Boyd was, as assumed by
Donaldson, for the purpose of making the cerebrum one unit, the cerebellum one
unit, and the medulla, pons, and midbrain one unit; though it is probable that a
part, at least, of the mesencephalon was included with the cerebrum, since, upon
removing the calvarium and dura, Boyd "removed the right and left hemispheres
of the cerebrum in slices to the tentorium." The cerebellum was isolated by cutting
the crura close to the pons.

A closer study of the group of specimens tabulated as premature still-born,
which are recorded as measuring 10 to 18 inches total length (CH), or approximately
250 to 460 mm., shows that they would correspond to an age range from the end of
the fifth to the end of the eighth month, according to Strata's tables as reported by
Martin (1914). According to Mall (1910), these two extremes would for the
younger correspond to a sitting-height (CR) of 167 mm., age 20 weeks; and for the
older a sitting-height of 310 mm., age 35 weeks. The weights of the brain parts in
these two extremes would be in the smaller fetus 41.25 grams for the cerebrum, 1.98
grams for the cerebellum, and 1.32 grams for the medulla-pons; and in the larger
fetus 297 grams for the cerebrum, 24.75 grams for the cerebellum, and 1.65 for the
medulla-pons. These figures must at best be accepted with reservations, since
the tables from which they are taken are undoubtedly inaccurate in some particulars.
For example, the weight of the encephalon in no case equals the added weights of

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 45

the various parts of the same brain. These results are probably due to the use of
crude apparatus and to a tendency to give results in round numbers.

Ziehen (1903) includes 9 embryos in a large series of cases reported. The age
of 6 of these was 30 days and the recorded total brain-weights were respectively 0.98,
1.0, 1.05, 1.13, 1.15, and 1.15 grams; 3 were 29 days old and the total brain-weights
were 1.29, 1.23, and 1.15 grams respectively. All of the material has been preserved
in formalin.

Michaelis (1907), in a number of cases of children autopsied by him in Leipzig,
reports 7 groups of embryos . Jackson ( 1 909) , in a study of prenatal growth, records
observations made upon 43 specimens ranging from 6 mm. to full term. In these
studies the measurements were by volume and not by weight, as in the other cases
recorded. The volume was determined in some cases by the amount of water dis-
placement of the brain of the embryo modeled in wax by the Born method; in others
by making enlarged drawings of the parts, measuring these drawings with a plani-
meter and calculating the volume by multiplying the areas by the thickness of the
sections. His conclusions are that the brain, although subject to considerable
variation, shows a fairly regular curve of growth. At the second month it forms
slightly more than 20 per cent of the total body-volume, the average dropping to
12.8 per cent in still-born fetuses. In the living-born, however, the average was
about 14.6 per cent.

Dockeray (1915), working toward the same end which is the purpose of my
own paper, and using similar methods, gives a complete volumetric study of the
brain parts in a fetus 156 mm. in length, and I have incorporated his results in this
paper.

METHODS AND MATERIAL.

Two points in the technic employed in my own studies were determined upon
because their worth had been proved, both in this laboratory and by other workers:

First, the use of the crown-rump or sitting-height as a standard for determining
the size of the material used: Both Le Bon (1879) and Donaldson (1909) agree
that body-length is a better criterion than body-weight from which to infer brain-
weight. The use of this method relieves us of the necessity of determining the leg-
length, an unsatisfactory procedure even in older fetuses.

Second, the method of preservation: All of the material used in this study had
been preserved in formalin which alone in varying strengths, or in combination with
other agents, is by far the most universally used preservative for animal tissues.
Hrdlicka (1906) experimented with the effects of various preservatives upon the
brains of both human and lower vertebrates and found formalin the most satis-
factory. Though confirming in part the work of others who claim that this agent
causes an increase in volume in the tissues preserved in it, in his summary he states
that the simple formalin solutions all show the same effects in all brains. These
consist of a sharp initial rise in the weight of the specimen, reaching a maximum
within less than a week, and a subsequent gradual, long-continued decrease. The
rise is in inverse ratio to the strength of the formalin solution, the percentage of
loss being apparently independent of the formalin percentage.

46 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

Many workers have tested out various fixatives with more or less indifferent
success, and while for special tissues and under favorable conditions many satis-
factory methods have been developed, still for general purposes, where availability
of the agent and the ease with which it can be prepared are the main desiderata,
none have yet been discovered which are as satisfactory and consequently as
widely used as is formalin in varying degrees of strength.

In order to secure the necessary data upon which to base conclusions, a great
number of embryos were studied and such stages selected as it was thought would
best show the developmental phases under consideration. Out of this number, 10
specimens were chosen, ranging in size from 4.3 mm. CR, estimated age 4^ weeks,
to a new-born infant of 367 mm. CR. All the material used is to be found in the
Carnegie Embryological Collection. In table 1 will be found listed the embryos
that were selected as being particularly suitable for the purposes of this study.
They are arranged serially in the apparent order of their development.- The
catalogue numbers given are those under which the embryos are listed in the
records of the laboratory. The measurements all signify the crown-rump length in
millimeters. The ages in weeks were determined from the last known menstrual
period and conform to the curve of age (based on length and weight) plotted by
Streeter (1920). All of the specimens had been preserved in formalin, and the first
8, in addition to fixation, had been embedded, cut, and stained. The material was
used as found, no attempt being made to account for any tissue changes dependent
upon the technic. The volume is therefore that of the embedded specimen. The
shrinkage resulting from the embedding process introduces a considerable error in
connection with actual volume, but as the shrinkage of the component parts of a
given brain is uniform, the error does not prevail in reference to the relative volume
of the separate parts.

In each of these younger embryos the brain was projected, drawn, and modeled,
according to the well-known method of Born. (See table 1 for magnifications used
in each instance.) The resulting models could then be easily subdivided into the
various parts decided upon and the weights and volumes ot these readily obtained.
The two older specimens were of a sufficiently advanced stage of development to
permit of dissection and weighing of the actual tissue.

TABLE 1. — List of specimens the brains of which were studied.

Catalogue
No.

Crown-rump
length.

Estimated
menstrual age.

Magnification
of model.

mm.

weeks.

diameters.

148

4.3

41

50

43

16

7

20

584a

25

8

20

96

50

11

10

1400-20

69.5

125

10

1400-22

SO

13*

10

1400-25

119

16

5

1400-28

156

19

5

230

26

Dissected

2558

367

40

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 47

The first question that arose was the selection of a method of subdivision which
could be adapted to all stages and which would yet permit the greatest number of
dissections and comparisons. Only simple determinations were possible in the
early stages, additions being made as the encephalon increased in complexity. In
seeking to establish definite landmarks by which to be guided in delimiting the
various parts, it was decided to base the initial procedure upon the early subdivision
of the encephalon into the primary brain vesicles and then to add such later sub-
divisions as could be accurately determined by a study of the gross models and a
microscopic study of the tissues. This plan permitted the development of a logical
method of procedure.

Since in all stages it was possible to delimit the three early vesicles — the
rhombencephalon,mesencephalon,and prosencephalon — the first point to determine
was that of landmarks by which these vesicles could be separated from each other
and the hindbrain from the spinal medulla. In the latter, of the three criteria
most commonly used — the highest root fibers of the first spinal nerve, the lowest
root fibers of the hypoglossal nerve, and the lowest crossing fibers of the pyramidal
decussation — the first proved to be the most satisfactory. Accordingly, the first
section was made in a transverse plane immediately anterior or cranial to the highest
rootlets of the suboccipital nerve. This section was the same in all cases, and
the resulting anterior mass, the encephalon, was the portion used for these studies.
The second section — the separation of the hindbrain from the midbrain — was made
in the youngest specimen (4.3 mm. CR length) in the constriction immediately
in front of the outward curve of the rhombic lip, the cerebellar rudiment. Later,
when it was possible to discern the inferior colliculi and the superficial transverse
pontine fibers, the incision was started dorsally, just posterior to the colliculi, to end
ventrally immediately anterior to these pontine fibers.

The mesencephalon is much more prominent proportionally in the younger
subjects than in the older, and forms the arch of the cephalic flexure. It is wedge-
shaped, the ventral surface being concave and shorter, the dorsal surface convex
and longer, and delimited in front and behind by transverse surface grooves. The
posterior incision was comparatively easy and was made, as described above, to
remove the rhombencephalon. The line of demarcation between the mesenceph-
alon and the prosencephalon was, however, more difficult to follow, since it was
found upon studying the tissues that the thickened lateral plates of the mesen-
cephalon project forward into the prosencephalon or, more properly speaking, are
invaginated as a result of the caudal growth of the prosencephalon, so that the line
of incision must extend anteromesially between these mesencephalic plates and the
more prominent overlapping masses of the thalamencephalon. This lateral thick-
ening of the mesencephalic plates, which is more pronounced in the basal parts
ventral to the iter (the site of the future tegmental nucleus), is sufficiently well
marked in these early stages to prevent error. Later, other landmarks make their
appearance; dorsally, the epiphjrsis, the habenular nuclei, the posterior com-
missure, and the increasing prominence of the posterior extremities of the thalami;
anteriorly, the pulvinars; posteriorly, the differentiation of the superior colliculi;

D 0 1

48 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

and ventrally, the developing mammillary bodies. In older subjects, therefore,
this anteromesial incision must, on each side, extend between the mesencephalon
and the thalami, to end just caudal to the epiphyseal attachment dorsally and the
mammillary bodies ventrally.

Having divided the entire encephalon into the three primary vesicles, the suc-
ceeding steps consist in separating these structures into their more prominent
constituent parts. In the 4 mm. embryo it was not possible to accurately remove
the cerebellar rudiment. In the 16 mm. embryo, in the case of the hindbrain, the
cerebellum could be removed from the stem. In the early stages this was accom-
plished by removing the prominent anterior rhombic lip on either side, which corre-
sponds to the cerebellar rudiment, as shown in figure 2. This mass, at first a promi-
nent roll or lip, becomes more and more sharply defined in the succeeding stages and
is delimited from the mesencephalon by the transverse groove mentioned above.
Owing to the difficulty of accurately separating the pons and the medulla, they were
considered as a unit mass. The mesencephalon was considered as a single mass
throughout the series. The fore-brain was then subdivided into the telencephalon
and the diencephalon. In the early stage this division was accomplished by cutting
along the thalamic margin, which extends as a slight groove along the dorsal border
of the optic evagination and its later developing stalk connected with the dien-
cephalon. In the older specimens the separation was effected by cutting through
the internal capsule between the thalamus and the corpus striatum dorsally, the
stria terminalis being used as a guide; ventrally the incision was made along the
lateral margin of the optic tract, leaving this structure connected with the thalam-
encephalon. Thus the diencephalon comprises the optic tracts, the thalamus,
the epithalamus, and the hypothalamus (except the hypophysis, which could not be
considered in all cases). The diencephalon was not subdivided. The telenceph-
alon in all stages was subdivided into the neopallium and the archipallium. In
the early stages the line of incision was made to include the shallow depression in
the lumen of the ventral wall of the first vesicle, lateral to the lamina terminalis
and anterior to the optic stalk. In the older specimens (16 mm. and upwards) this
rudiment is larger and consequently more easily separated from the remainder of
the vesicle. Here there is a considerable ventral prominence appearing lateral to
the lamina terminalis anterior to the torus opticus. This comprises the olfactory
bulb and tract. The latter extends along the lower or ventral margin of the
depressed cortical area over the base of the corpus striatum. In older subjects
(50 mm. and upwards) the anterior commissure, paraterminal body, fornix, and
hippocampus could easily be made out by microscopic study and as they became
differentiated they were removed and studied.

The method employed in isolating these various structures in the older modeled
embryos was essentially the same as that followed in dissecting the brain of the
term fetus, which was as follows : The cord was separated from the medulla at the
conventional level (in these specimens the bodies had been injected with a 10 per
cent solution of formalin and the brain, after removal, had been kept in the same
solution until studied) and the upper part of each cerebral hemisphere removed

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 49

by a transverse incision made just above the level of the fibers of the corpus callosum.
The occipital lobe on each side was then removed flush with the caudal margin of the
splenium and the basal surfaces of the temporal lobes were pared away until the
junction of the mesencephalon and diencephalon could be seen. The mesenceph-
alon was then separated from the rhombencephalon by an incision passing just
cranial to the transverse pontine fibers ventrally, and just caudal to the posterior
colliculi dorsally. The mesencephalon was next removed by an incision passing
anteromedially between the anterior colliculi and the pulvinars, just caudal to the
epiphyseal stalk dorsally and the corpora mamillaria ventrally. Next, the
olfactory bulbs and tracts were freed and removed by cutting each tract at its
junction with the base of the hemisphere. The two hemispheres were then
separated by cutting through the median sagittal plane of the corpus callosum, the
lamina terminalis, chiasma, and the basal structures, the incision passing between
the mammillary protuberances. The resulting blocks were treated alike; the
corpus callosum was lifted up and carefully dissected off from the fornix, exposing
that structure, as well as the thalamus, stria terminalis, and the corpus striatum.
The remainder of the cortex was removed anteriorly, with the head of the caudate
nucleus as a guide. Laterally, the cortex of the insula was pared down to the
lenticular nucleus and the temporal lobe removed, leaving the hippocampus, the
fimbrise, and the fornix intact. Following the fornix around anteriorly, its anterior
pillar and the mammillary and paraterminal bodies were removed together. The
thalamencephalon was then separated from the corpus striatum by cutting through
the internal capsule, following dorsally the line of the stria terminalis, and ventrally,
cutting just lateral to the optic tract, leaving that structure connected with the
diencephalon. The cerebellum was then separated from the remainder of the hind-
brain by cutting through the peduncles flush with the hemispheres.

This method of subdivision gave 8 separate units for study: (1) the medulla-
pons, (2) the cerebellum, (3) the mesencephalon, (4), the diencephalon, (5) the
fornix and hippocampus, including the paraterminal and mammillary bodies, (6) the
olfactory bulb and tract, (7) the telencephalon, and (8) the corpus striatum. In
all of the specimens a regrouping of these units was made, giving us the following
additional units: (1) the total hindbrain, which comprises both the cerebellum and
the medulla-pons ; (2) the total telencephalon which was subdivided into archipal-
lium (including the olfactory apparatus) and the neopallium — that is, the remainder
of the telencephalon, minus the corpus striatum, which in the more advanced
specimens was considered as a separate unit. In each instance the total weight and
volume of the encephalon were determined; then those of each of the subdivisions
were ascertained and the percentages calculated from the totals. The actual
weight and volume could both be determined in the case of the dissected specimens,
while for the modeled specimens the weight and volume of the entire model, or of a
part, were first determined. The weights of these wax models were simply carried
through, no attempt being made to ascertain the actual weight, since the relative
values would have been the same. The volume of the model was obtained by deter-
mining the volume of a gram of the wax used in making the model and multiplying

50 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

this by the weight of the part under consideration. The cubic centimeter value of
the wax being thus obtained in grams, it was a simple matter to calculate the
various amounts, percentages, etc. The actual volume was obtained by dividing
the model volume by the cube of the magnification used in the reconstruction.

In making this study the companion parts of the two sides have in every
instance been grouped together, thus giving one unit for comparison instead of
studying each of these paired structures separately. In all cases, too, the percent-
age weight has been chosen as the unit of comparison, since this method gives a
stable basis, whatever may have been the magnification used in the reconstruction,
and would apply equally well when the actual weights were considered, as in the two
dissected specimens. Weight, rather than volume, has been selected for comparison
for the same reason that model weights were considered instead of attempting to
calculate actual weights, since it was thought that by reducing the method employed
to the simplest possible working basis the percentage of error would likewise be
reduced to a minimum. It is to be remembered, also, that these percentage
weights are of only relative value, for the various parts of the encephalon present a
steady and consistent growth and development throughout the period of gestation,
and that the decrease in percentage weight, recorded for certain parts, as opposed to
increased percentages for others, means merely that those parts which present an
increase in weight grow so much more rapidly that they outstrip their more
sluggish neighbors in relative bulk, as will be seen in a study of both models and
tissues.

In every case, except the two oldest specimens, a careful check was kept upon
all steps by a microscopic study of the tissues. This was found to be especially
valuable in determining the landmarks to guide the incisions. However, while
occasional reference may be made to the degree of differentiation, it is the intention
at this time to deal entirely with the question of growth in bulk or mass, without
reference to histologic detail.

DESCRIPTIONS OF INDIVIDUAL SPECIMENS.

No. 148, 4.3 mm., estimated menstrual age 4^ weeks.
This embryo (see table 3 and fig. 1), which has
been the subject of a number of studies by various
investigators, presents some quite unusual
features and was selected for the reason that, for
a young stage, it presents the various develop-
mental features to such a degree as to render it
especially valuable as a basis upon which to build
a serial study of the various parts of the en-
cephalon. At this stage the central nervous
system is distinctly tubular in character and the
anterior neuropore is closed. Anterior and dorsal
to the eye-stalks, and separated from them by a
distinct fold, can be seen the olfactory region or
archipallium, and more dorsally the neopallium,
consisting at this stage of two faintly outlined
evaginations, one to either side of the mid-line.
The entire prosencephalon comprises 31.3 per
cent of the weight of the encephalon and can be

quite easily subdivided into the telencephalon,
weighing 7 per cent, and the diencephalon,
weighing 24.3 per cent. The further subdivision
of the telencephalon yielded an archipallium of
4.2 per cent and a neopallium of 2.8 per cent.
The cephalic flexure is well defined; its arch,
rather narrow, is formed by the mesencephalon,
which constitutes 14.3 per cent of the encephalon.
The rhombencephalon at this stage is propor-
tionately very large, comprising 54.4 per cent of
the total weight of the encephalon.

A study of this vesicle shows that the lateral
plates have a wide dorsal spread; the thin roof
plate is closely applied to the overlying ectoderm
and only a slight thickening is observed at the
point of junction of the roof and lateral plate —
the rhombic lip. The cerebellar rudiment is
merely a thickened fold which stretches across the
more anterior, cephalic portion of the dorsum or

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 51

roof, showing a tendency to overlap the con-
stricted caudal extremity of the mesencephalon.
The cerebellum is not sufficiently differentiated
at this stage to permit of its delimitation. The
site of the pontine flexure is not indicated at this
stage, though both the cervical and the cephalic
flexures are well marked.

No. 43, 16 mm., estimated menstrual age 7 weeks.

This specimen (see fig. 2) presents a much
more advanced stage of development. The
tubular character is less pronounced and the
pontine flexure is well defined, serving to ac-
centuate both the cervical and cephalic flexures.
That portion of the rhombic lip constituting the
cerebellar rudiment was sufficiently well dif-
ferentiated to permit of its removal and study,
„ oh

-r I '•
Q ; '-/• :
I' •

36.63 per cent of the total brain. Of this the
cerebellum comprises 9.74 per cent and the
medulla-pons 26.89 per cent, a slight increase for
the former, a considerable decrease for the latter.
The mesencephalon has increased to 14.76 per
cent, the total prosenccphalon to 48 per cent, to
which the diencephalon contributes 19 per cent—
a slight shrinkage, and the telencephalon 28.84
per cent, a gain of 7 per cent over the preceding
stage. A further study of the telencephalon
shows an archipallium of 1.43 per cent, a tre-
mendous drop as compared with 18.09 per cent
at 16 mm. The neopallium has, on the other
hand, increased to 27.41 per cent, a complete
reversal of relative values.

The basal ganglia, both the thalami and the
corpora striata, show a considerable increase in

:

' * :

•• •; •'••

FlQ. 2. — Left profile view of reconstruction of lirain
of embryo No. 43, crown-rump length
16 mm. X 10.

contributing 7.56 per cent of the total weight as
opposed to 38.11 per cent for the medulla-pons,
the combined weights giving a total of 45.67 per
cent for the rhombencephalon, a relative decrease
from the preceding specimen. The mesenceph-
alon has decreased to 12.52 per cent and the
diencephalon to 20.38 per cent, whereas the
telencephalon has increased to 21.43 per cent,
three times its weight in the younger embryo.
Of this the archipallium constitutes 18.09 per
cent and the neopallium 3.34 per cent, making a
total of 41.81 per cent for the prosencephalon, as
opposed to 31.3 per cent in the 4.3 mm. stage.

No. 584<z, 25 mm., estimated menstrual age 8 weeks.

This embryo presents a still further increase in
development, the rhombencephalon weighing

Rhombeneep/,

Archipal.

FlO. 1. — Left profile view of reconstruction of
the encephalon of embryo No. 148,
crown-rump length 4.3 mm., show-
ing manner in which the brain wag
subdivided. X 25.

bulk at this stage. The walls of the vesicles also
show a corresponding increase in thickness, so it
will be observed that, although there has been a
marked increase in size and weight of the ence-
phalon up to this stage of development, it con-
sists in cell proliferation rather than in dedif-
ferentiation and does not permit of much added
study of individual parts.

No. 96, 50 mm., estimated age 11 weeks.

This embryo (figs. 3, 4, and 5) has advanced
much farther in development and permits of a
greater amount of subdivision of the model.
The total brain-weight is 514.58 grams, of which
the rhombencephalon comprises 12.77 per cent,
the cerebellum contributing 3.79 per cent and
the medulla-pons 8.96 per cent, a relatively con-

52 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

siderable decrease in all three values. The
mesencephalon has likewise decreased, weighing
in this specimen but 7.09 per cent of the total, as
compared with twice that amount in the 8 weeks'
embryo. The total prosencephalon has nearly
doubled its relative weight, being now 80.14 per
cent, the diencephalon dropping to 12.77 per
cent, and the telencephalon increasing to 67.37
per cent, of which the archipallium constitutes
3.52 per cent and the neopallium 51.18 per cent,

well back over the diencephalon, encroaching
upon the mesencephalon. A curious feature,
noted only at this age, is that the corpora striata,
the diencephalon, and the rhombencephalon, are
all of equal weight value.

No. 1400-20, 69.5 mm., estimated age 12J^ weeks

This specimen still more closely approaches the
full adult differentiation. The cerebral hemi-
spheres have grown in all directions to a con-

f f':$; •;'• vyv^s

lU • iH

lOlfad.bulb /
Fornix G H ippocamp ''•<;.'..;.,-.

V

Corpus s/

'u~. .y^-.* •'*•''.'.'.??>•'• '. ". ''• f f'~',' •'"'

FIQ. 3. — Mesial view of right hemisphere of embryo No. 96, crown-rump length 50 mm., showing sub-
divisions of archipallium. X 5.

FIG. 4. — Left view of model of brain-stem (same embryo as shown in fig. 3), showing its subdivisions.
X 5.

FIG. 6. — Mesial view of hemisection of model shown in fig. 4, showing subdivisions of diencephalon.
X 5.

a tremendous increase over all preceding stages.
The corpora striata, which could be isolated in
this specimen, weighed 12.67 per cent of the
total brain. The various parts of the archi-
pallium could also be isolated, giving an olfac-
tory bulb weighing 0.3 per cent, and the para-
terminal body, fornix, and hippocampus, which
together weighed 3.22 per cent of the total. The
pontine flexure is much less acute and the cere-
bellum, as yet rudimentary, is a narrow, beak-
like structure (fig. 3). The mesencephalon
shows a distinct caudal projection overhanging
the median part of the roof of the cerebellar
rudiment. The cerebral hemispheres project

siderable degree, the total prosencephalon weigh-
ing 84.84 per cent of the whole. Of this the
diencephalon contributes 11.47 per cent, a
steady relative decrease, despite the increase in
bulk of the developing thalami. The archi-
pallium has relatively decreased, weighing but
2.87 per cent, of which the olfactory bulb con-
tributes but 0.15 per cent and the paraterminal
body, fornix, and hippocampus 2.72 per cent.
The neopallium, on the contrary, weighs 60.62
per cent and the corpus striatum 8.88 per cent,
making a total of 69 per cent of the entire
encephalon. The mesencephalon has dropped
to 5.97 per cent and the total rhombencephalon

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 53

to 10.19 per cent, of which the medulla-pons
represents 6.23 per cent and the cerebellum 3.96
per cent, thus showing the continued and rapid
increase of the prosencephalon, especially the
neopallium, over the other parts.

No. 1400-22, 80 mm., estimated age 13J^ weeks.

A still more decided advance in bulk and dif-
ferentiation is here observed (fig. 6). The
cerebral hemispheres are considerably more

ebel.

Flo. 6. — Left view of model of brain-stem, embryo No.
1400-22, crown-rump length SO mm. X 6.

extensive, the telencephalon totaling 80.56 per
cent of the entire weight of the encephalon. Of
this the neopallium constitutes 65.55 per cent
and the corpus striatum 11.23 per cent, a total
of 76.78 per cent. The archipallium weighs 3.78

creased to 88.6 per cent. The mesencephalon
presents a sligh decrease, weighing now 4.78
per cent. The total rhombencephalon has also
lost relatively, weighing 6.62 per cent, of which
the cerebellum comprises 2.89 per cent, the
medulla-pons 3.73 per cent.

No. 1400-25, 119 mm., estimated age 16 weeks.

The advance in general bulk is distributed in
this embryo (figs. 7 and 8) as follows: The total
prosencephalon constitutes 91.91 per cent of the
brain weight, of which the diencephalon com-
prises 6.29 per cent — a decrease, the telenceph-
alon having meanwhile increased to 85.62 per
cent. To this the neopallium contributes 69.55
per cent, the corpus striatum 11.64 per cent, and
the archipallium 4.43 per cent. Of the latter the
olfactory bulb shows a decreased weight of 0.28
per cent, and the paraterminal body, fornix, and
hippocampus a weight of 4.15 per cent — a slight
rise. The mesencephalon has fallen to 2.06 per
cent. The total rhombencephalon weighs 6.03
per cent, of which the cerebellum forms 2.74 per
cent and the medulla-pons 3.29 per cent, all of
which weights are practically identical with those
of the corresponding parts in the 80 mm. speci-
men, so that the telencephalic gain is balanced by
the losses sustained by the diencephalon and the
mesencephalon, the fractional increase in the
rhinencephalon being likewise accounted for.

Cerebel

JiMIil

\

Medulla

Hypothal. Paraterm. body

•*'-•>.-.

Paraterm.body

Olfact.bulb

8

HippocQmp.8t fornix

FIG. 7. — Mesial view of model of brain-stem of embryo No. 1400-25, crown-rump length 119 mm. X 2.5.
Flo. 8. — Mesial view of right half of telencephalon of embryo shown in figure 7. X 2.5.

per cent, the olfactory bulb contributing 0.35 per
cent and the paraterminal body, fornix, and
hippocampus 3.43 per cent; all of which agree
substantially with the figures given for the
corresponding parts of the preceding embryo.
On the other hand, the neopallium has increased
and the diencephalon has diminished to 8.04
per cent, the total prosencephalon having in-

No. 1400-28, 156 mm., estimated age 19 weeks.
This specimen (figs. 9, 10, 11, and 12) was
studied and published by Dockeray (1915), and
his findings have been rearranged and incor-
porated into this study. The total prosenceph-
alon has attained the enormous weight-value
of 93.69 per cent of the total brain weight, the
diencephalon dropping to 4.81 per cent and the

54 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

telencephalon weighing 88.78 per cent. Of the
latter the neopallium comprises 78.82 per cent
and the corpus striatum 7.62 per cent, the former
an increase, the latter a decrease as compared
with the preceding stage. The archipallium has
declined to 2.34 per cent, the olfactory bulb
weighing 0.14 per cent, the paraterminal body,
fornix, and hippocampus 2.2 per cent, the lowest
point yet reached by these parts. Themesenceph-

-VOIfacI B.

; \_Para'erm.B.

, basal qangli

' Dienceph.

Mesenceph .
•Cerebel.

•

• ;• •

Pons Medulla

FIG. 9. — Ventral view of a reconstruction of brain of embryo No.
1400-28, 150 mm. crown-rump length, showing manner in
which it was subdivided. X 2.

alon has also decreased, constituting but 1.44
per cent of the encephalon. The total rhomb-
encephalon weighs 4.97 per cent, slightly below
that of the 119 mm. specimen, the loss being
sustained by the meclulla-pons which now weighs
2.23 per cent, while the cerebellum weighs 2.74
per cent, the same weight-ratio as that in the
' preceding stage and virtually the same as that in
the 80 mm. stage. This structure thus main-
tains a relative weight-level extending over a
period of six weeks, which really means that there
has been an acceleration of growth during this
interval. The foregoing specimens were all
modeled and the models separated into these
various parts for study.

Fetus 230 mm., estimated age 26 weeks.

This stage is represented by a fetus of which
only the brain was available for study. Since
the specimen had readied a sufficiently advanced
state of development to permit the removal and
dissection of the brain, the actual weights and
volumes can be given. In form and fissures this
brain corresponds to a stage between the two
shown by Retzius (189G, plate 14, figures 3, 4, 5,

and figures 6, 7, 8), and has therefore been
placed at 340 mm. total length, or 230 mm. crown-
rump length, and its age estimated as 26 weeks.
These data were obtained from Dr. Streeter, who
dissected and studied the specimen.

The total prosencephalon weighed 93.32 per
cent of the encephalon, a fractional loss as com-
pared with the 156 mm. fetus. Of this the clien-
cephalon constituted 3.7 per cent, less than the
preceding one. The telencephalon, the total
weight of which wa 89.62 per cent, was sub-
divided into th • neopallium, weighing 79.21
per cent, and (he corpus striatum, weighing
8.33 per cent, both showing an increase. The
archipallium weighed but 2.08 per cent, the
lowest point yet reached. To this the olfac-
j tory bulb contributed 0.13 per cent, the para-
terminal body, fornix, and hippocampus, 1.95
per cent. The mesencephalon weigljed 1.42
per cent, practically the same as in the preced-
ing specimen. The total rhombencephalon
weighed 5.26 per cent, a somewhat greater
relative weight than at 19 weeks. This added
amount was contributed by, the cerebellum,
which weighed 3.15 per cent, as compared
with a medulla-pons weight of 2.11 per cent,
which is only a small fraction less than in
Dockeray's fetus, the last and largest of the
modeled specimens.

No. 2558, 367 mm., new-born white male.

This specimen had been injected and pre-
served in 10 per cent formalin. The sitting
height was 367 mm., body-weight 3,210 grams.
The head measurements were as follows:
length 130 mm., width 102 mm., circumfer-
ence 381 mm., biaural arc 248 mm. The brain
was removed by Dr. A. H. Schultz, of the Car-
negie Embryological Laboratory, to whom I am
indebted for the foregoing data. Judging from
macroscopic appearance, both body and brain
were normal and well developed. The enceph-
alon weighed 505.61 grams; the actual volume,
as determined by water displacement in a large
graduated cylinder where the indices could be
easily read, was 489.97 c. c. The cerebral
hemispheres were very large, richly convoluted,
and had thickened until the cavity of the lat-
eral ventricle was reduced to a mere slit. The
total prosencephalon weighed 467.12 grams, or
92.38 per cent of the whole brain, slightly
less than that of the preceding. Of this the
diencephalon contributed 2.63 per cent, a loss
of about 30 per cent as compared with the
230 mm. fetus. The telencephalon weighed
89.75 per cent of the whole, maintaining the
same weight-value as the preceding specimen.
Of this the neopallium weighed 424.3 grams,
or 83.92 per cent, and the corpus striatum
24.3 grams, or 4.8 per cent, a considerable
relative gain for the former. The archipallium

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 55

has declined to 5.2 grams or 1.03 per cent, which
is the lowest mark in the weight-curve for this
complex. Of this the olfactory bulb con-
tributed 0.2 per cent, and the paraterminal body,
fornix, and hippocampus, 0.83 per cent. The
mesencephalon likewise reaches its 'owest relative
level at this stage, weighing but 0.64 per cent of
the total weight of the encephalon. The rhom-
bencephalon weighed 35.28 grams, or 6.98 per
cent, the medulla-pons weighing 4.58 grams

H e o p ° :

(0.91 per cent) and the cerebellum 30.7
grams (6.07 per cent). This represents for
the former the lowest point in the series, and for
the latter a marked gain over preceding figures.
In this specimen the total brain constituted
15.75 per cent of the body-weight, a somewhat
greater proportion than is usually reported.
Jackson (1909) gives 14.6 per cent and Vierordt
(1893) 12.29 per cent.

N eo

' ° P „ ,

10

Hypothal.

Parat body
( pillar fornij!)

Olfact 8

FIG. 10. — Mesial view of right half of telencephalon of embryo shown in fig. 9. X 2.
Flo. 11. — Mesial view of the subdivided archipallium of embryo shown in figs. 9 and 10. X 2.
FIG. 12. — Drawing of a homisection of brain shown in figs. 9, 10, and 11, with subdivisions of basal
ganglia outlined. (Section b, slide 145.1 X 2.

BRAIN VOLUME AT DIFFERENT STAGES.

Since we have the volume of the models of the brain and its separate parts for
the different stages of development, it is possible to roughly calculate the actual
volume by dividing the model volume by the cube of the magnification used for the
reconstruction. It must be borne in mind, however, that since we are dealing with
brains that have been embedded in paraffine or cel-
loidin, the models represent the embedded brain and
one must take into account a considerable shrinkage
of the tissues. The shrinkage of a given brain is
more or less uniform throughout and therefore, in
itself, does not interfere with the determination
of the relative volume of its different parts. As
regards the actual volume, however, one must
take into consideration the primary swelling of
the brain when placed in a formalin solution and
its subsequent shrinkage in the course of its
preparation into serial sections. From our own
experience and that of others we can estimate

TABLE 2. — Volume of brain at different
stages of development, calculated by
dividing the model volume by the cube
of its magnification.

Age.

CR length.

Volume of
brain .

weeks.

mm.

c. c.

4

4

0.003

7

16

.041

8

25

.126

11

60

.603

12J

69.5

1.977

13J

80

1.503

16

119

8.750

19

156

13.120

26

230

54.290

40

367

489.970

56 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

that the volume of the brain as embedded is from one-tenth to one-third less
than the original, depending on the character of the tissues and their reaction
to formalin, and on the dehydrating and embedding media. The extent of this
shrinkage for the individual brains I have at present no way of accurately deter-
mining. In spite of the fact that my figures are only approximate (probably one-

TABLE 3. — Data on the relative weight and volume of the subdivisions of the brain at various stages of development.

With the exception of the two oldest stages, these data are based on wax-plate reconstructions. Where the structures
are bilaterally symmetrical the data for the right and left are combined.

1

Rhombencephalon

Medulla-pons

Cerebellum

Mesencephalon

Diencephalon

Telencephalon

Neopallium

Corpus striatum

Neopallium and
Corpus striatum.

Archipallium

Hippocampus, for-
nix and parater-
• minal body.

Olfactory bulb

_a

oa

.a

3
1

Embryo 4.3 mm. CR, No. 148. Brain
modeled X 60 by Streeter.

19.31
54 40
22 60

5.08
14.30
5.94

37.73
12.52
41.12

131.00
14.76
131.00

36.55
7.09

42.76

103.50
5.97
103.50

61.93
4.78
71.84

19.75
2.06

22.52

20.90
1.44
23.66

.78
1.42

.77

3.21
.64
3.11

8.63
24.30
10.10

61.41
20.38
66.94

175.50
19.77
175.50

65.70
12.77
76.87

199.00
11.47
199.00

104.13
8.04

120.79

60.33
6.29

68.78

69.76
4.81

78.97

2.05
3.70
2.01

13.32
2.63
12.91

2.48
7.00
2.90

64.56
21.43
70.37

256.00
28.84
256.00

346 . 68
67.37
405.61

1255.24
72.37
1255.24

1043 . 70
80.56
1210.69

820.80
85.62
935.70

1287.67
88.78
1456.65

49.52
89.62
48.65

453 . 80
89.75
439 . 77

.99
2.80
1.16

10.07
3.34
10.97

243 . 30
27.41
243 . 30

263.40
51.18
308.18

1051.40
60.62
1051.40

849.23
65.55
985.11

666.75
69.55
760.09

1143.20
78.82
1294.10

43.77
79.21
43.00

424.30
83.92
411.18

1 49

35.50
100.00
41.54

301.32
100.00
328.43

887.60
100.00
887.60

514.58
100.00
602.06

1734.50
100.00
1734.50

1295.53
100.00
1502.81

958.64
100.00
1092.85

1450.47
100.00
1639.85

55.26
100.00

54.29

505.61
100.00
489.97

4.20
1.74

Embryo 16 mm. CR, No. 43. Brain
modeled X 20 by Streeter.

137.62

45.67
150.00

325.10
36.63
325.10

65.65
12.77
76.81

176.67
10.19

176.67

85.77
6.62
99.49

57.76
6.03
65.85

72.14
4.97
80.57

2.91
5.26
2.86

35.28
6. 98
34.18

114.83
38.11
125.16

238.70
26.89
238.70

46.10
8.96
53.94

108.00
6.23

108.00

48.31
3.73
56.04

31.48
3.29
35.89

32.33
2.23
36.50

1.17
2.11
1.15

4.58
.91
4.43

22.79
7.56

24.84

86.40
9.74
86.40

19.55
3.79

22.87

68.67
3.96

68.67

37.46
2.89
43.45

26.28
2.74

29.96

39.81
2.74
44.07

1.74
3.15

1.71

30.70
6.07
29.75

54.49
18.09
59 40

Embryo 25 mm. CR, No. 584a. Brain
modeled X 20 by Jenkins.

12 70

1 43

12 70

Embryo 50 mm. CR, No. 96. Brain
modeled X 10 by Snyder & Sherrick.

65.18

12.67
76.26

154.10
8.88
154.10

145.46
11.23
168.73

111.60
11.64
127.23

110.61
7.62
124.21

4.60
8.33
4.52

24.30
4.80
23.55

328.58
63.85
384.44

1205.50
69.50
1205.50

994.69
76.78
1153.84

778.35
81.19
887.32

1253.81
86.44
1418.31

48.37

87.54
47.52

448.60
88.72
434 . 73

18.10
3.52
21.18

49.74
2.87
49.74

49.01
3.78
56.85

42.44
4.43

48.38

33.86
2.34
38.34

1.14
2.08
1.13

5.20
1.03
5.04

16.56
3.22
19.37

47.10
2.72
47.10

44.40
3.43
51.50

39.73
4.15
45.30

31.89
2.20
36.10

1.07
1.95
1.06

4.20
.83
4.07

1.54
.30
1.81

2.64
.15
2.64

4.61
.35
5.35

2.71
.28
3.08

1.97
.14
2.24

.07
.13

.07

l.(est)
.20
.97

Fetus 69.5 mm. CR, No. 1400-20.
Brain modeled X 10 by Jenkins.

Fetus 80 mm. CR, No. 1400-22.
Brain modeled X 10 by Sherrick.

Fetus 119 mm. CR, No. 1400-25.
Brain modeled X 5 by Snyder.

Percentage of total weight

Fetus 156 mm. CR, No. 1400-28.
Brain modeled X 5 by Dockeray.

Fetus 230 mm. CR, Brain dissected
by Streeter.

Actual volume, c. c

Fetus 367 mm. CR, No. 2558. Brain
dissected by Jenkins.

Actual volume, c. c

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 57

tenth to one-third less than the original volume) , the interest attached to the ques-
tion of the volume of the fetal brain perhaps warrants my including the data given
in table 2, which were obtained on the basis of the volume of the models. From
these same figures one can readily calculate the volume of any individual part of
the brain by multiplying the total volume by the percentage of the total brain
formed by that particular part (table 3) .

SUMMARY.

A study of the preceding data shows the rapid growth of the brain as a whole
and the relatively enormous rate of growth of some of its component parts as com-
pared with others; for, while all parts of the brain consistently increase in size, they
do not all grow at the same rate. The telencephalon, for example, shows a relatively
rapid increase throughout the entire series, greatly surpassing all other parts; the
cerebellum shows a similar, though less marked increase during the latter half of
gestation; while all of the other parts show a relative decrease in weight- values.

Starting with an embryo of 4 3/2 weeks, 4.3 mm. long, with an actual total brain
volume of approximately 0.003 c. c., and in which it is possible to definitely outline
little more than the three primary brain vesicles common to all vertebrates, one can
follow the rapid growth and development until birth, when this complex and highly
specialized organ attains a total actual volume of 490 c. c., a gain in less than 30
weeks of over 160,000 times the initial volume.

An analysis of the values for the different parts shows a steady upward curve
for the prosencephalon, from 31.3 per cent in the early embryo to 92.38 per cent at
term, a gain of nearly three times its initial bulk, despite the fact that all of its
component parts, except the telencephalon, have shrunken considerably in weight-
values. The percentage of the total brain-weight formed at different stages by the
five chief subdivisions of the brain is shown graphically in chart 1.

The most striking feature in this growth curve, as well as the most significant
one, is to be found in the enormous proportional and actual increase in the size of the
telencephalon, which gives the amount of cortical expanse necessary to provide for
the control of all subsidiary parts of the nervous system, as well as being the seat
of the higher psychic functions. Beginning with an initial weight of 7 per cent at
4>£ weeks, the telencephalon increases rapidly up to 13.K weeks, when it constitutes
80.56 per cent of the entire encephalon. Then follows a period of more gradual
growth up to the twenty-sixth week, when the telencephalon attains its maximum
relative weight (89.62 per cent) which it maintains until term. Assuming that the
specific gravity of the brain tissue is very little over 1.0, the actual increase in weight
is from approximately 0.0002 gram (as calculated from volume) at 4>^ weeks to
493.8 grams at birth.

Another noteworthy point in the growth rate of the cerebral hemisphere is that
it attains its maximum relative weight before any folding occurs to increase the
extent of its periperhal surface, showing that in the later weeks its development is
one of complexity rather than of bulk, a fact which would appear to argue that these
later changes consist in perfecting the development of already existing structural
units, rather than in the acquisition of new ones.

58 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

These figures for the telencephalon are rendered even more striking when we
study the growth conditions of its various subdivisions. It was possible in all cases
to separate this structure into the neopallium and the archipallium. At 4.3 mm.
the undivided neopallium weighed 2.8 per cent; at 25 mm. 27.4 per cent; at 50 mm.
51.18 per cent, the curve continuing to mount rapidly to a maximum of 83.92 per
cent at term. At the stage of 50 mm. it was possible to isolate the corpus striatum
which weighed 12.67 per cent. From this point its weight-curve declines steadily
to a minimum of 4.8 per cent at term. The archipallium at 4.3 mm. weighed 4.2
per cent, twice the initial weight-value of the neopallium, reaching a maximum of 18
per cent at 7 weeks, from which point the weight-curve declined steadily to 1.03 per

[J Telen

cephalon

I Mesencephalon
Cerebellum

i;

CHART 1. — Diagrammatic representation of the percentage weights of the brain parts at different stages

of development.

RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN. 59

cent in the new-born. At 11 weeks the archipallium could be subdivided into the
olfactory bulb, which weighed 0.3 per cent, and the paraterminal body, fornix, and
hippocampus, which together weighed 3.22 per cent. Both of these units show a
consistent shrinkage in relative weight until birth, when the former weighed 0.2
per cent and the latter 0.83 per cent.

The diencephalon could also be isolated in all cases, and in the 4.3 mm. embryo
had a weight-value of 24.3 per cent, or more than three times that of the telenceph-
alon. From this high initial percentage the curve declined gradually to 2.63 per
cent at term, notwithstanding the considerable increase in the size of the thalami.

The mesencephalon, weighing in the youngest specimen 14.3 per cent, drops
steadily to 0.64 per cent at term, declining from a structure of considerable size
to a mere slender, connecting stem, which attains its chief differentiation through
the development of secondary centers connected with the fiber systems passing
through it.

The total rhombencephalon starts with a maximum weight of 54.4 per cent of
the whole in the 4.3 mm. specimen, from which point its curve sinks rapidly until
at 156 mm. it reaches a minimum of 4.97 per cent, then rises gradually to 5.26 per
cent at 230 mm., and 6.98 per cent at term. This unusual curve is explained when
we study the growth-rate of the two component units of the rhombencephalon,
which behave in a widely different manner. The curve for the cerebellum begins in
the 16 mm. embryo at 7.56 per cent, ascends to 9.74 per cent at 8 weeks, its maximum
point during prenatal growth. From this level it falls to 2.74 per cent at 13>£ weeks,
remaining there until the end of the nineteenth week when it rises rapidly, reaching
6 per cent at term, its actual weight being 30.7 grams. This reversal of the usually
observed conditions shows the increasing importance of the cerebellum at these
stages of development, its individual growth-rate becoming sufficiently marked to
maintain an unvarying level during six weeks of changing values in other parts. It
then gains a new impetus which is sustained throughout of the remainder gestation.
The medulla-pons in the 16 mm. embryo comprises 38.11 per cent of the weight of
the entire encephalon. The curve falls rapidly to 3.73 per cent at 13>^ weeks, then
more gradually to its minimum of 0.91 per cent at term, the general weight-curve
for this part being comparable only to that of the diencephalon, though the relative
loss is much greater in the medulla-pons. It would seem probable that the early
large size of the medulla-pons is due to its association with the development of the
cranial nerves which are relatively very large at this time.

60 RELATIVE WEIGHT AND VOLUME OF COMPONENT PARTS OF FETAL BRAIN.

REFERENCES.

BISCHOFF, 1880. Das Hirngewicht des Menschen. Bonn.
BOYD, R., 1861. Tables of the weights of the human body

and internal organs in the sane and insane of

both sexes at various ages. Phil. Trans. Royal

Soc. of London, vol. 151.
DOCKERAY, F. C., 1915. Volumetric determinations of the

parts of the brain in a human fetus 150 mm.

long (crown-rump). Anat. Rec., vol. 9.
DONALDSON, H. H., 1895. The growth of the brain. London.
, 1908. A comparison of the albino rat with man in

respect to the growth of the brain and spinal

cord. Jour. Comp. Neur., vol. 18.
, 1909. Relation of body-length to body-weight and

to the weight of the brain and of the spinal cord

in the albino rat. Ibid., vol. 19.
HRDLICKA, ALES, 1906. Brain and brain preservatives.

Proc. U. S. Nat. Mus., vol. 30.
• , 1919. Anthropometry. Amer. Jour. Phys. An-

throp., vol. 2.
JACKSON, C. M., 1909. On the prenatal growth of the various

organs and parts. Amer. Jour. Anat., vol. 9.
LE BON, G., 1879. Recherches anatomiques et math£ma-

tiques sur les lois des variations du volume de

cerveau et sur leur relations avec 1'intelligence.

Rev. d'anthrop., Paris, vol. 2.

LEWANDOWSKY, 1910. Handbuch der Neurologic, vol. 1.
MALL, F. P., 1910. Determination of the age of human

embryos and fetuses. Manual of Human

Embryology, Keibel & Mall, vol. 1.
MANOUVRIER, 1885. La quantitfe dans 1'encaphalie.

Paris.
MARCHAND, F., 1902. Ueber das Hirngewicht des Menschen.

Abh. der math. phys. Cl. d. Konigl. Sachs.

Gesellsch. d. Wiss., vol. 27.
MARSHALL, JOHN, 1892. The brain of the late George Grote,

with comments and observations on the human

brain and its parts generally. Jour. Anat. &

Phys., vol. 27.

MARTIN, 1914. Lehrbuch fur Anthropologie. Jena.

MATIEGKA, J., 1903. The significance of the weight of the
brain in man. Sitzungber. der Kon. kornischen
Gesellsch. d. Wiss. Math. Natrwiss. Classe
Jahr., vol. 20.

MICHAELIS, P., 1907. Das Hirngewicht des Kindes.
Monatschr. f. Kinderh., vol. 6.

PEARL, RAYMOND, 1905. Some results of a study of varia-
tions and correlations in brain-weight. Jour.
Comp. Neur., vol. 15.

, 1905. Biometrical studies on man. Variation and

correlation in brain weight. Biometrika, vol. 4.

RETZIUS, G., 1896. Das Menschenhirn. Tafel 2. Stock-
holm.

, 1900. Ueber das Gehirngewicht der Schweden.

Biol. Untersuch., N. F., vol. 9.

SPITZKA, E. A., 1907. A study of the brains of six eminent
scientists and scholars belonging to the
American Anthropometric Society. Amer.
Phil. Soc., n. s., vol. 21, part 3.

STREETER, G. L., 1920. Weight, sitting-height, head-
size, foot-length, and menstrual age, in the
human embryo. Contritubions to Embry-
ology, vol. 10, Carnegie Inst. Wash. Pub. 274.

VIERORDT, 1890. Das Massenwachsthum der Korperorgane
des Menschen. Arch. f. Auat. u. Phys., Anat.
Abth., Supp. Bd.
, 1893. Daten und Tabellen, 3 Aufl., Jena.

WEST, G. M., 1894. Anthropometrische Untersuchungen
tiber die Schulkinder in Worcester, Mass.
Archiv f. Anthrop., vol. 22.

ZIEHEN, 1899. Macroskopische Anatomie des Gehirns.
Bardeleben'a Handbuch d. Anat., vol. 4.

, 1903. Gehirngewichte. Monatschr. f. Psych, u.

Neurol., vol. 13.

CONTRIBUTIONS TO EMBRYOLOGY, No. 60.

ABNORMALITIES OF THE MAMMALIAN EMBRYO OCCURRING

BEFORE IMPLANTATION,

By GEORGE W. CORNER,
From the Anatomical Laboratory, Johns Hopkins Medical School.

With two plates and one figure.

61

ABNORMALITIES OF THE MAMMALIAN EMBRYO OCCURRING

BEFORE IMPLANTATION,

The occurrence of monstrosities, degenerative changes, and other abnormal
conditions of the mammalian embryo brings up so many questions of scientific
and practical interest, attended by such difficulties of approach, that it seems in
order to report some recent observations which, though few in number, afford
evidence of a positive character with regard to one of the points now in dispute.

Mall, in his several contributions on the nature and significance of abnor-
malities of the human embryo, never hesitated to assert the opinion that all these
changes could be attributed to disturbances of nutrition due to faulty implantation
of the ovum in the maternal uterine mucosa. This conviction was based on two
grounds; first, that defective implantation could actually be observed in many of
the abnormal human embryos which had come into his hands; second, that experi-
mental teratologists had produced in amphibia, by the application of chemical
poisons which presumably hindered the embryonic nutrition, all sorts of abnor-
malities resembling many of the known mammalian terata.

This hypothesis has never been subjected to critical test in the human species,
because of the lack of specimens from stages prior to implantation; but the possibility
of attacking it on logical grounds was seized by Kellicott, who pointed out (partly
on the basis of his own remarkable experiments on the production of abnormalities
in fish embryos by the action of low temperatures) that the onset of such changes
might date from the earliest stages of cleavage of the ovum. He suggests that the
cause of abnormal and monstrous development is to be found in disturbance of the
normal organization of the ovum and he definitely considers the possibility that in
mammals the disturbing cause might be found operative before as well as after
implantation.

A year before Kellicott's contribution Huber had published an account of a few
observations on abnormalities in very young embryos of the albino rat. His series
included cases of separation of the first two blastomeres into half -embryos, incom-
plete or retarded segmentation, degeneration of ova at the end of the segmentation
period, and abnormal formation of the segmentation cavity, all occurring in embryos
which were lying free in healthy uteri. Since faulty implantation could be excluded,
Huber was inclined to consider these abnormalities due to disturbances inherent
in the ova.

The specimens described herein were obtained during the collection of a series
of early embryo pigs. Figure 1 shows two of these embryos that were found
among others in a healthy, normal uterus. One (A) is a normal blastodermic vesicle
which would appear, by comparison with Assheton's well-known stages, to be about
7 days old; figure 7 A, plate 2, illustrating the same ovum, displays its delicate, filmy
texture, with clearly outlined nuclei, and the developing embryonic area below the

63

64 ABNORMALITIES OF MAMMALIAN EMBRYO BEFORE IMPLANTATION.

ectoblast. It is also shown in section in figure 2, plate 1. The other vesicle (B) is
collapsed and wrinkled, its texture granular and almost opaque (the opacity is
slightly exaggerated in fig. 7v), the nuclei hardly visible. Microscopic section
confirms the external appearance of degeneration, showing the cells to be granular
and their relations distorted (fig. 3). It is interesting to note that the preservation
of the cells is better at one pole of the vesicle than at the other. In the ovaries of this
sow there were 7 corpora lutea; in the uterus 6 vesicles, of which 2 were entirely
normal, 2 normal in texture but collapsed and cup-shaped, and 2 abnormal, as
illustrated. In addition there was 1 unsegmented ovum, thus accounting for all the
ruptured follicles.

Figure 4 shows another case of the same sort with cupping of the vesicle and a
partial break-down of the cells. This was obtained from another healthy uterus
containing 1 normal vesicle and 4 of the collapsed type.

Inner cell mass.

Fio. 1. — A normal (A) and a path-
ological (B) ovum obtained from
the same uterus. The more de-
tailed structure of these speci-
mens is shown in figure 7, plate
2. The pathological ovum is
wrinkled and compressed. Here
it is shown from a side view; in
figure 7 its broader surface is

Nuclei. Bhown'

A B

A third sow contained a still more interesting series of 4 vesicles. These were
5 mm. in diameter, a size attained by normal pig embryos at about the twelfth day,
according to the studies of Assheton. In 3 the abnormality consisted of total
absence of an embryonic area; in order to make sure of this point the vesicles were
fixed in the fully dilated condition by injection of Bouin's fluid. I am indebted to
my colleague, Dr. C. H. Heuser, for friendly aid in the preparation of these speci-
mens. When dilated the 3 vesicles were of even thickness at all points of their
walls. Microscopic section of one of them showed it to be composed of the usual
two layers of cells.

The fourth specimen from this uterus showed, on the contrary, a distinct mass,
slightly larger than the embryonic area found in a normal vesicle of the same
dimensions, but opaque, white, and projecting from the surface of the vesicle by a
pedicle, so that altogether it was not unlike a little mushroom in appearance.
Figure 6 shows a section of this curious specimen, in comparison with a normal
embryo from a vesicle of the same size (fig. 5). The little tumor is composed of
cells of diverse staining reaction, disposed in an irregular way, but including three
or four minute vesicular cavities, one of which appears in the section. The whole
arrangement rather suggests a malignant papilloma; whether the tumor is at the
site of the embryonic area itself or at some other part of the vesicle (the embryonic
area being absent, as in the other blastodermic vesicles in the same uterus) can not

ABNORMALITIES OF MAMMALIAN EMBRYO BEFORE IMPLANTATION. 65

be decided; but in any event the specimen may be regarded as a true monster, the
earliest ever seen in a mammal. This and the other related blastodermic vesicles,
like those described in the previous paragraphs, were found lying unattached in the
cavity of a healthy uterus which was quite normal and which proved on microscopic
examination to be lined by a mucosa in every way similar to that of uteri containing
normal embryos. The Fallopian tubes presented no gross abnormalities; they were
not examined microscopically.

In view of the positive evidence afforded by the few specimens gathered by
Huber and the writer, we are forced to the conclusion that the organism is liable to
pathological changes before its attachment to the uterus. Whether these are due
to faults inherent in the germ-cells, or to traumata which assail the ovum during
its passage from the ovary to the uterine mucosa (such, for example, as chemically
abnormal secretions of a seemingly normal uterus) our specimens of course do not
explain; they merely increase the difficulty of the problem by demonstrating the
inadequacy of the theory of faulty implantation to account for all developmental
aberrations. Most likely it will be proved in the end that the germ-cells and their
product are liable to the onset of abnormalities at all stages of their history; certainly
hints of such a possibility are given by two lines of investigation, as yet but little
applied to mammalian forms. The first of these comprises those experiments
already mentioned, which show that the eggs of fish and frogs are susceptible to
damage by all sorts of chemical and physical factors in their environment, at stages
of development so early as to be comparable with the segmenting mammalian
ovum during its passage through the tube and while in the uterus before implanta-
tion; the other line of approach is through the studies of Morgan and others, who
have discovered the occurrence in plants, insects, and even mammals, of inheritable
factors which, when brought into play by some unfavorable mating, cause such
deviation from the anatomical or physiological norm as to kill the individual at one
or another stage of its development. For instance, according to Cuenot, Castle,
and Little, it is impossible to breed mice which are homozygous for yellow coat-
color; it appears that the inheritance of this character through both parents is
invariably fatal to the embryo; and, indeed, the recent report of Kirkham hints
strongly that in this case the lethal factor may attack the embryo shortly before the
time of implantation, while it is still in the morula or vesicle stage. It seems not
impossible, therefore, that in future the application to mammals of both the
teratological and the genetic modes of experimentation may aid in solving the
complex but pressing questions arising from abnormalities of development of the
human organism.

66

ABNORMALITIES OF MAMMALIAN EMBRYO BEFORE IMPLANTATION.

AUTHORS CITED.

ASSHETON, R., 1898. The development of the pig during

the first ten days. Quart. Jour. Microsc. Sci.,

vol. 41, p. 329.
CASTLE, W. E.t and C. C. LITTLE, 1910. On a modified

Mendelian ratio among yellow mice. Science,

n. B., vol. 32, p. 868-870.
CUENOT, L., 1908. Sur quelques anomalies apparentes des

proportions Mendeliennes. Arch. Zool. Exp.,

vol. 9; Notes et Revue II, p. vii.
HUBEB, G. C., 1915. The development of the albino rat.

II. Abnormal ova; end of the first to the end of

the ninth day. Memoirs Wistar Inst. Anat.

and Biol., No. 5.
KELLICOTT, W. E., 1916. The effects of low temperature on

the development of Fundulus. Amer. Jour.

Anat., vol. 20, p. 449-482.

KIRKHAM, W. B., 1917. Embryology of the yellow mouse.
Anat. Rec., vol. 11, p. 480-481. (Proc. Amer.
Soc. Zool.)

MALL, F. P., 1908. A study of the causes underlying the
origin of human monsters. Jour. Morph.,
vol. 19, p. 1-361.

, 1910. Pathology of the human ovum. In: Keibel

and Mall. Manual of Human Embryology,
vol. 1, p. 202-242.

, 1917. On the frequency of localized anomalies in

human embryos and infants at birth. Amer.
Jour. Anat., vol. 22, p. 49-72.

LITTLE, C. C., 1917. The relation of yellow coat color and
black-eyed white spotting of mice in inherit-
ance. Genetics, vol. 2, p. 433^444.

MORGAN, T. H., 1917. The mechanism of Mendelian in-
heritance. New York.

, 1919. The physical basis of heredity. Philadelphia.

CORNER

PLATE 1

r

Embryonic shield

.e^e*'
a.*

J.F.Didusch fecit

A.Hoen& Co imp

(2) Section illustrating structure of a normal 7-day pig ovum. This is the same specimen shown in text-figure 1 a. X 600 ; i. c. m.,
inner cell mass. (3) Section of abnormal ovum shown in figure 1 b, illustrating flattening of the vesicle with partial degeneration of its
cells. X 600. (4) Section of a degenerating flattened embryonic vesicle of about the 7th or 8th day. X 600. (5) Section through
normal blastodermic vesicle of the 1 1th day. X 150. (6) Section through tumor mass in wall of early blastodermic vesicle, showing
irregular growth of embryonic cells. X 150. Drawings made directly on stone.

CORNER

PLATE 2

J. F. Didusch fecit.

FIG. 7. — Two blastodermic vr.-irlc •sustained from the normal uterus of a pig. One of these
(A) is normal and has a development of about 7 days; the other (B) is patho-
logical. X GOO.

CONTRIBUTIONS TO EMBRYOLOGY, No. 61.

THE DEVELOPMENT OF THE EXTEKNAL GENITALIA IN THE

HUMAN EMBBYO.

BY MILO HERRICK SPAULDING,

Of the University of Montana, Slate College of Agriculture, Bozeman.

With four plates and two text-figures.

67

THE DEVELOPMENT OF THE EXTERNAL GENITALIA IN THE

HUMAN EMBRYO,

The present investigation is concerned with the development of the external
genitalia in the human embryo, the determination of the stage at which sex differ -
ences make their appearance, and how early sex can be recognized from these
external structures. In addition to the morphological interest in the structural
development of the phallus region, its embryology presents several other phases
of almost equal importance. Thus the ability to recognize sex at an earlier period
than has hitherto been possible is of great practical value to the clinician as well as
to the embryologist. At the same time, this point is largely dependent upon the
verification of the existence of a definite "indifferent" ("undifferentiated," Pohlman
1904) period through which all embryos have been generally believed to pass before
assuming their definite male or female characteristics.

As a result of this study I have found that there is apparently no real indifferent
period. On the contrary, the younger embryos show constant differences in the
phallus and almost from the earliest differentiation of the genital tubercle (i. e.,
embryos 14 mm. GL) they can be divided into two groups. This division is based
upon the marked difference in the length of the urethral groove upon the caudal
slope of the phallus. This seems to be quite constant and without intergradations
and can be traced backward from the older embryos, in which sex can be definitely
recognized, to the younger stages which heretofore have been included in the so-
called "indifferent" period. In one group the urethral groove extends from the
base of the phallus practically to its apex, i. e., onto the region of the future glans;
these I consider to be males. In the second group the groove is much shorter and
terminates some distance below the apex of the phallus — i. e., it does not extend
onto the region of the future glans; these I consider to be females.

I wish to extend my sincerest thanks to Dr. G. L. Streeter for his kindness in
according me the privilege of examining the extensive collection of human embryos
in the Carnegie Laboratory of Embryology, as well as for the great interest he has
taken and the many helpful suggestions he has made throughout the course of the

investigation.

MATERIAL AND METHODS.

The material studied in the preparation of the present paper has been limited
to the collection of human embryos in the Carnegie Laboratory of Embryology.
This includes a large number of specimens which are graded 2 and 3; that is, they are
unsuitable for sectioning because of faulty preservation, mutilation, or some other
defect; also a considerable number of sectioned embryos up to 50 mm. in length.
While perhaps more attention has been paid to the external features of these unsec-
tioned embryos, as far as possible this evidence has been verified by a study of
sectioned specimens of corresponding sizes, a number of these having been recon-

69

70

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

TABLE 1. — Condensed tabulation of specimens exam-
ined, arranged in length groups of 10 mm. (length
decades).

Sex.

Length.

Male.

Female.

Unde-
termined.

8 to 13mm.

33

17

31

20 to 29 mm.

41

40

30 to 39 mm.

15

48

40 to 49mm.

11

26

50 to 59mm.

16

14

60 to 69mm.

25

23

70 to 79mm.

6

5

80 to 89mm.
90 to 100 mm.

8
14

3

7

structed in order to correlate the development of the internal urogenital organs
with that of the external genitalia.

Only a condensed tabulation of the embryos examined, with the sex classified, is
included in the present report, the complete list, including the serial number of each
embryo with remarks upon its condition, having been placed on file among the
records of the Carnegie Laboratory of Embryology. The sex grouping adopted in
this table is based upon the diagnostic features brought out by the investigation.

Table 1 shows that there is considera-
ble variation in the relative frequency of
the sexes in each of the length decades
and also that there is a greater number of
females than males (145 to 84) in the period
between 14 and 50 mm. length. This vari-
ation in the different groups would indicate
that as yet a sufficient number of embryos
has not been examined to permit definite
deductions concerning the sex-ratio at this
period of embryonic life. Conclusions,
therefore, will have to be deferred until
much more extensive data have been
accumulated.

Critical examination has been made of practically all of the embryos in the
collection between the lengths of 14 and 50 mm. About 80 of these, however,
have been discarded from present consideration because, owing to mutilation or
poor state of preservation, a clear picture of the external genitalia could not be
obtained; therefore, no interpretation of sex was possible. Of the older stages
(51 to 100 mm. CR) only selected specimens have been studied. The accom-
panying list, therefore, includes all the well-preserved embryos between the length
of 14 and 50 mm., both sectioned and unsectioned, and a small proportion of the
older specimens.

No attempt was made to identify the sex of embryos smaller than 14 mm.,
although it is believed that a careful study of a sufficient number of such younger
stages will show that the sex differences here pointed out can be traced back to the
very beginnings of the genital tubercle.

HISTORICAL.

When the early scientists began the study of human embryology with the few
specimens that came into their possession, the questions of sex recognition and the
relations between the sexes attracted their attention to the study of the external
genitalia. When we of the present day consider the paucity of their material, as
well as the comparative crudity of their instruments and methods, we can not but
be surprised at the accuracy of their observations and conclusions. That there
were misinterpretations was due largely to insufficient material and defective
apparatus, rather than to faulty observation.

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 71

As the result of this early work we find, in reviewing the literature on the sub-
ject, that various views were held by these early scientists concerning the sex of
the younger embryos during the so-called indifferent period. Many of them
considered that the younger embryos represented a stage intermediate between
males and females. Others, for example Tiedemann (1813), who is conspicuous
for the accuracy of his observations, believed that, because so many more females
than males were observed, all human embryos were at first female and that the
males resulted from an advance in development over this more primitive condition.
The subsequent investigations of Joh. Muller (1830), Rathke (1832), and Bischoff
(1842) resulted in but minor additions to the observations of Tiedmann. In fact,
Muller considered that the latter were so remarkably complete that additional
investigations could not be expected to produce anything new. Ecker (1851-59)
published some excellent figures of the external genitalia of human embryos, which
have served as a basis for the somewhat inaccurate Ecker-Ziegler wax models of
this region. No other notable papers were published until about 1889, when the
development of modern instruments, the introduction of modern methods of tech-
nique, and the consequent improvement in the quality of the preparations, resulted
in a renewed interest in all branches of biological study. In the embryology of the
human urogenital system this era was heralded by the work of Tourneux (1889)
upon a series of 35 specimens ranging in length from 24 mm. to 35 cm. Only 10 of
these were less than 40 mm. in length, while the majority were fetuses over 50 mm.
in length. Although Tourneux described the external genitalia of the majority
of these specimens, several of which are shown by fairly accurate figures, the
number of young embryos was relatively so few that his attention was not directed
toward possible sex differences in these early stages; hence, more consideration
was given to the histological study of the urethral canal and the formation of the
prostate. In the same year Nagel published the results of his study of a small
series of human embryos. This was also devoted mainly to the internal urogenital
system, although Nagel made some observations upon the external genitalia which,
in part, predicted some of the results of the present study. He believed, for
example, that the urethral groove should furnish a clue to the sex of an embryo at a
much earlier period than it had previously been possible to recognize sex, although
his observations indicated that the males possess the shorter groove, while my
findings tend to show quite the opposite. Nagel furthermore found that the
outer genital (labio-scrotal) swellings do not arise as a more or less complete ridge
surrounding the phallus, as had previously been figured, but as separate, lateral
swellings which only later become connected in front of the phallus in the female,
and in the male are shifted to a caudal position to form the scrotum. The classical
investigation of Keibel (1896), giving an exhaustive account of the development of
the internal urogenital organs, based upon a series of 13 embryos (3 to 25 mm. in
length), was probably the first study of this system by the modern method of plate
reconstruction whereby it has been possible to correlate the development of the
external with the internal organs. While Keibel figures the external genitalia in
several of his reconstructions, especially those of embryos 11.5 and 25 mm. long,

72

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

which show the shorter urethral groove that I have found to be characteristic of
female embryos, no special description is given ot these structures. In fact, he
realized that for the investigation of the development of the external genitalia a
large series of embryos was necessary, and this portion of his paper is for the most
part devoted to a plea to his colleagues for more specimens.

While I realize that I have been more fortunate than any of my predecessors
in having at my disposal a larger and more complete series of embryos, I still would
reiterate Keibel's plea for more material; because it seems that the more specimens
one has at his disposal, the more one realizes the need of additional material in order
to study some of the points concerning which there is still a conflict of opinion.

In 1904 Herzog pointed out that the direction of the phallus made it possible
to recognize the sex of an embryro at about the beginning of the third month, in-
stead of towards its close. This difference in direction he showed to consist in a
greater caudal decurvation of the female phallus, while that of the male remained
more nearly at right angles to the axis of the body. His series consisted of 16
embryos, varying in length from 20 to 190 mm. Of these, he figures the external
genitalia of only 9, none of which bring out very clearly his point of directional dif-
ference of the phallus. Some excellent figures of the external genitalia are given by
Otis (1906), although the discussion of the genitalia does not form a logical part of
his paper.

CLANS AREA
EPITHELIAL TAG
LATERAL BUTTRESS

ANAL PIT
•ANAL TUBERCLES

CLUTEAL FOLD

TEXT-FIGURE 1. — Drawing of external genitalia of an embryo 16.8 mm. long (Specimen No. 492),
illustrating the genital-tubercle period. X 25.

From 1890 to 1907 a number of valuable contributions upon the development
of the urogenital system in mammalia, especially the domesticated forms, made
their appearance. While a number of these (notably the series of investigations
carried out by Fleischmann and his pupils) pertained more or less to the develop-
ment of the external genitalia, they have proved of but little use in the present
investigation, since these papers failed to bring out any early sex differences in the
forms studied.

Felix (1912) gave an exhaustive summary of the accumulated observations
upon the development of the human urogenital system, although his discussion of

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

73

the external genitalia is rather meager as compared to that of the internal organs.
Certain of my observations have led me to conclusions different from those reached
by Felix; the discussion of these moot points will be taken up in the section describ-
ing the development of the external genitalia.

STAGES OF DEVELOPMENT OF THE EXTERNAL GENITALIA.

While the greater part of the investigation has been devoted to a study of the
changes occurring after the embryos have attained a length of 14 mm., a few
observations have been made upon younger specimens which are here included as
forming a starting-point for the present description.

For the sake of convenience, the sequence of development of the external
genitalia has been divided into three periods, each of which is represented by one or
more stages: (1) genital-tubercle period, characterized by the more or less conical
form of the genital eminence prior to the formation of the labio-scrotal swellings; (2)
phallus period, beginning with the definite appearance of the labio-scrotal swellings
which separate the conico-cylindrical outgrowth of the genital tubercle from the
surrounding body areas; and (3) definitive period, characterized (particularly in
male embryos) by the transition of the primary external genitalia into what is
essentially their final form.

CLANS

EPITHELIAL TAG

•m

CORONARY SULCUS

SHAFT
-UROCENITAL OPENING

^H/t URETHRAL FOLDS r ^

LATERAL PHALLIC GROOVED
LAB1O.SCROTAL SWELLINGS

ANAL TUBERCLES

RAPHE
POSTERIOR COMMISSURE

B

TEXT-FIGURE 2. — Drawings illustrating the parts of the external genitalia at the beginning of the definitive period.
A. male embryo 45 mm. long (Specimen No. 948); B, female embryo 49 mm. (Specimen No. 2256). X 20.

GENITAL-TUBERCLE PERIOD.

Stage 1, 9 mm. (fig. 1, plate 1). — The earliest stage that I have studied is shown
in an embryo 9 mm. long. In this specimen the genital tubercle is a very low
conical eminence between the umbilical cord and the base of the tail. Its apex is

74 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

separated from the rest of the tubercle by a pair of converging depressions which
unite in the midventral line to form a longitudinal groove extending to the caudal
border, giving to the entire groove a Y-shape. Caudal to the arms of the Y and
lateral to its stem is a pair of broad, swollen areas. This condition is slightly
different from the stage described by Keibel (1896) of the beginning development of
the genitalia in a 3-mm. embryo. In his specimen he described the cloacal tubercle
as consisting of a pair of eminences separated by the cloacal membrane. The
difference between the stage described by Keibel and my first stage may be accounted
for entirely by the difference in the size of the two specimens.

The term genital tubercle, which I use to describe the entire genital eminence,
has a slightly different meaning from the one given to it by Felix (1912, p. 948).
That author speaks of the genital eminence at first as the cloacal tubercle, which is
later divided into a basal genital tubercle and a terminal phallus. According to
his usage, the former develops into the genital (labio-scrotal) swellings, while' the
latter forms the phallus proper. My observations indicate that the entire primor-
dial genital eminence should be considered as the precursor of the phallus and as
such should be termed the genital tubercle. As will be shown later, my observations
also indicate that the labio-scrotal swellings apparently originate from the outlying
tissue and not from the basal portion of the primary tubercle. But even without
this modified interpretation of their origin, the term genital tubercle seems most inap-
plicable to these swellings (which are decidedly secondary, both in genital structure
and in function) and is therefore much more a propos of the genital eminence as the
precursor of the phallus.

Stage 2, 8 mm. (fig. 2, plate 1). — The second stage, represented by the recon-
struction model of an 8-mm. embryo, shows considerable advance in the develop-
ment of the genital tubercle. This now forms a somewhat rounded mass which
occupies almost the entire area between the umbilical cord and the base of the tail.
Its cranial margin is but barely indicated by a slight groove between its apex and the
umbilical cord. The apex is broadly rounded and from it the slightly convex caudal
surface slopes toward the base of the tail. The caudal slope is almost bisected by
the urethral groove, a shallow depression extending from the base to the tip of the
tubercle. The basal end of the groove is separated from the anal pit by a narrow
transverse bar. It is significant that, while there is this external separation of the
primitive cloacal groove into the urethral groove and anal pit, the internal division
of the cloacal cavity into urogenital sinus and rectum is not yet complete, at least in
this embryo. The margins of the urethral groove are but slightly elevated into the
urethral folds, although the margins of the anal pit are more pronounced. Laterally
the caudal surface of the tubercle is rounded out into pronounced swellings.
Almost in front of the tubercle the ventral body-wall bears a pair of swellings lying
in the umbilico-phallic angles, the significance of which has not yet been definitely
ascertained. These masses show a gradual increase in size until the embryo
reaches a length of 16 mm. They then apparently disappear in some embryos,
while in others they undergo a caudal shifting; this would seem to indicate that
they are the primordia of the labio-scrotal swellings, which definitely make their

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 75

appearance in embryos 17 to 19 mm. long. As yet, however, a sufficient number of
these younger embryos has not been examined to permit definite conclusions on this
point.

Stage 3, 8 to 12 mm. (figs. 3 and 4). As growth proceeds, the genital tubercle
is transformed into a compressed, conical protuberance. This is brought about
by the deepening of the umbilico-phallic groove, so that the cranial slope of the
tubercle is nearly straight (i. e., approximately at right angles with the body axis),
while the caudal slope remains decidedly convex. At the same time the caudal
outline has become markedly triangular by the broadening of the base until it
occupies practically the entire area between the bases of the legs. The conspicuous
lateral slopes form the "lateral buttresses" which, arising from the cranial border
of the tubercle just proximal to its apical area, have a decidedly caudal trend, so
that they finally disappear basally opposite the caudal border of the tubercle. The
apex of the tubercle is now clearly marked off from the more proximal portion by a
shallow circula.1- depression, indicating it as the future glans and separating it from
the basal shaft. The urethral groove is a long, lancet-shaped depression, broadest
and deepest basally, narrowing distally into a shallow slit limited by a very small
"epithelial tag" just proximal to the primitive glans area. The urethral folds
(margins of the groove) are elevated as slight rolls of tissue which distally merge
into the glans region, while basally they become more tumid and broaden out to
surround the anal pit as anal tubercles. Lateral to these urethral folds, the caudal
surface of the tubercle is somewhat swollen in the younger specimen illustrating this
stage (fig. 3), in marked contrast to the decided concavity of these regions in the
older embryo (fig. 4) . In the former (fig. 3) the urethral and anal membranes have
not ruptured. The older embryo (fig. 4) agrees with the finding that (in the major-
ity of cases) these membranes rupture at about the stage of 12 to 13 mm., although
several specimens were found, among both the sectioned and unsectioned material
in which the membranes were still imperforate at 17 mm. The perforation of this
membrane transforms the shallow urethral groove into a gutter-like, primitive
urogenital opening as a direct communication between the phallic portion of the
urogenital sinus and the exterior. This is accompanied by an increase in the
definiteness of its outlines, as the result of which its sex difference in length is cor-
respondingly emphasized. Because of the variation in the time of rupturing of this
membrane, as well as on account of normal differences in the breadth of the open-
ing thus formed, it seems advisable to continue to use the term urethral groove,
except when referring to embryos that clearly show this feature as an opening.

While there is an apparent sex difference in the lateral outline of the tubercle
in embryos 10 to 16 mm. in length, in that in some its tip is placed slightly more
cranially than in others, thus producing a wider separation between the tubercle
and the tail in the first group than in the second, a sufficient number of embryos has
not been examined to permit of definite conclusions on this point, because of the
possibility thafc some distortion of the embryos may have been brought about by
manipulations during fixation.

76 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

Stage 4, 14 to 15 mm. (figs. 5, 7, male; figs. 6, 11, female). — In this stage the
sex difference in the length of the urethral groove becomes clearly evident. But
slight changes have taken place in the general shape of the genital tubercle. In the
male (figs. 5, 7) the glans area is more clearly indicated than in the female (figs. 6, 11).
In both embryos the urethral groove is sharply outlined, making it possible to con-
trast its difference in length in the two specimens. In the female the epithelial tag
is more pronounced than in the male and the entire tubercle is slightly more swollen.

PHALLUS PERIOD.

Stage 5, 16 to 17 mm. (fig. 8, male; fig. 12, female). — By the time the embryo
has attained a length of 16 to 17 mm. the genital tubercle has elongated into a
narrow, conical organ which, because of its modified shape and its separation from
the surrounding body areas by the newly formed labio-scrotal swellings, will now
be called the phallus. The male embryo representing this stage (fig. 8) shows the
phallus as more nearly cylindrical than in any of the younger embryos, the lateral
buttresses having, to a marked extent, merged into its body. The groove limiting
the glans is present, although not clearly shown in the photograph. The urogenital
opening is a narrow orifice extending almost the full length of the phallus, limited
distally by a pronounced epithelial tag. The urethral folds are considerably
broader than in any of the younger stages, their outer margins being sharply
separated from the remains of the lateral buttresses by a concave depression. The
most pronounced change, however, is the presence of the labio-scrotal swellings as a
pair of distinct, rounded ridges, one on each side of the base of the phallus and
separated from it by a broad lateral phallic groove. As regards this feature, it is
quite likely that this embryo represents a slightly older stage, as the male specimen
of the next stage shows more clearly the probable beginnings of these swellings, and
the female selected as the counterpart for this stage does not show them to any
marked degree, if they are present at all. It is, of course, very possible that there is
normally more or less variation in the time of development of these structures,
and that this embryo (fig. 8) is merely slightly precocious in this respect.

The female of this stage is represented by the dissected-out phallus region of an
embryo of the same length. The phallus has the same general configuration as that
of the male just described, with the exception of a shorter urogenital opening ter-
minated by a larger epithelial tag, a greater fullness in the genital folds, the presence
of a straight postanal bar which is grooved by a slight median longitudinal depres-
sion flanked on each side by anal tubercles, and the apparent absence of the labio-
scrotal swellings. The presence or absence of these swellings in this specimen must,
however, remain an open question, although the indications point strongly to their
absence. In this respect the embryo seems to be more nearly normal (so far as
sequence in the appearance of the modified parts of the external genitalia is con-
cerned) than its coordinate male.

Stage 6, 19 mm. CR (fig. 9, male; fig. 13, female). — In the male of this stage the
phallus is still of narrow, conical form, with the glans area rather more sharply
indicated by a broad, band-like depression, not as clearly shown in the photograph

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 77

as in the embryo itself. The remains of the lateral buttresses arise just proximal
to this depression, spreading laterally for a short distance, then continuing basally
almost parallel to the axis of the phallus until they finally merge into the tissue at
its base. Unlike most of the younger embryos examined, there is almost no caudal
trend to these buttresses. The urogenital opening has deepened and become more
pronounced. It extends from the base of the phallus practically to its tip. In
this specimen it is not limited distally by an epithelial tag. The lack of this
appendage, useless as it at present seems, is probably due to some loss prior to or
during fixation. The distal portion of the groove (the glans portion) forms a
diamond-shaped dilatation extending to the proximal limits of the glans, the
remainder of the groove being constricted into a deep, narrow slit. The urethral
folds are rather pronounced elevations which diverge distally to merge into the
glans, while proximally they are considerably thickened into a pair of caudally
projecting, diverging masses (bulbo-urethral swellings?) which merge laterally
into the outlying labio-scrotal swellings. They are separated from the cavernous
portion of the phallus by broad concavities. Perhaps the most interesting feature
of this embryo is that it shows the apparent beginnings of the labio-scrotal swellings.
These are a pair of flat, elevated areas on either side of the base of the phallus with
which they seem to be continuous. The cranial margins of these areas are nearly
straight, extending at diverging angles into the inguinal regions, where their
cranio-lateral angles continue for a short distance as ridges of tissue. The caudal
margins appear as continuations of the diverging wings of the basal portions of the
urethral folds, forming sinuous curves to the caudo-lateral angles, from which
points they sweep cranially in rounding curves to the cranio-lateral angles, where
they unite with the cranial margins. This is the only specimen which shows these
swellings definitely connected with the base of the phallus. In all others in which
they are developed there is a groove (lateral phallic groove) on each side which
separates them from the base of the phallus.

As regards the formation of the labio-scrotal swellings, my observations diverge
from the description given by Felix (1912, pp. 948-949) :

"The cloacal tubercle is thus divided into the almost circular base of the phallus and
the scmilunar genital tubercle (fig. 639). This latter represents an originally impaired
structure lying cranial to the phallus and surrounding it on two sidfs; from it there are
formed later the two genital swellings. The genital tubercle is separated from the phallus
in female embryos by a deep groove which is the anlage of the later sulcus nympho-labialis,
the groove between the labia majora and minora. In the male embryos this groove is
absent from the beginning."

My own findings indicate that the labio-scrotal swellings arise as paired out-
growths from the outlying tissue and not from a basal segment of the primary
tubercle. Furthermore, all but one of the specimens I have examined, both male
and female, show these swellings as separated from the phallus by the lateral phallic
grooves, a condition which is in striking contrast to the concluding statement made
by Felix.

78 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

The coordinate female (fig. 13) shows, more clearly than any of the younger
specimens, the distinction between the glans and shaft, although the division is
indicated more by changes in the surface modeling than by the formation of a defi-
nite coronary sulcus. There is also a very slight indication of the caudal decur-
vation which was first pointed out by Herzog (1904) as an early sex difference.
The urogenital opening is clearly limited to the shaft, being broader basally and
narrowing distally into a mere slit. The urethral folds are still quite broad and are
separated from the cavernous portion by shallow depressions. The labio-scrotal
swellings resemble those already described for the male of the preceding stage.

Stage 7, 21 to 23 mm. CR (figs. 10, 15, males; figs. 14, 19, females).— While there
is a slight increase in the length of the phallus at this stage, the most conspicuous
changes are the increase in the relative sex differences in the length of the urogenital
opening and the greater development of the genital folds and the labio-scrotal
swellings. In the males figured, the greater length of the urogenital opening is clearly
seen to be in sharp contrast to the much shorter opening of the females . The urethral
folds at this stage are more pronounced than in any of the embryos previously
described. As a result of the deepening of the lateral depression between these
folds and the cavernous portion of the shaft, their median portions now project from
the shaft as compressed ridges. Distally they merge into the terminal glans area,
while proximally they broaden out into the overhanging pre-anal enlargements
(bulbo-urethral swellings?). As a result of this modification, the shaft of the
phallus now shows clearly two distinct regions: the heavily thickened, cavernous
portion, formed by the medial migration of the lateral buttresses, and the caudal
urethral portion, represented by the projecting urethral folds outlining the uro-
genital opening, a condition which persists throughout the remainder of the phallus
period. The labio-scrotal swellings have now grown into /-shaped enlargements,
separated from the phallus by grooves more pronounced in the males than in the
females. Cranially, these swellings are separated from each other by a triangular
prolongation of the base of the phallus, so that there is no indication of the horse-
shoe shape which is so characteristic of most of the older illustrations. The caudal
extremities of these swellings are joined together by a low, curving postanal ridge
(slightly larger in the female), which is almost bisected by a median sagittal groove.

Stage 8, 24- to 28 mm. CR (figs. 16, 17, males; figs. 20, 21, females). — This stage
is represented by an exceedingly interesting pair of twins. In the male (fig. 16)
the nearly cylindrical phallus projects at approximately right angles to the body
axis, in marked contrast to a pronounced decurvature of that of its mate (fig. 20).
The glans is not so sharply demarked as in the female, although there are slight indi-
cations of its extent. The urogenital opening extends almost to the tip of the
phallus as a shallow, open groove deepened basally by the caudally extended, plate-
like urethral folds. As a result of this increase in the folds, the base of the phallus is
considerably longer than its apex. The labio-scrotal swellings are separated from
the base of the phallus by broad, shallow grooves; their tips curve cranio-laterally
away from the phallus, leaving a rather broad, unswollen area between the latter
and the umbilical cord.

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 79

In the female (fig. 20) the glans has become more clearly defined, in a negative
way, by changes which have taken place in the character of the urethral folds.
As a result of this modification the entire shaft is terminally decurved upon the
sharply demarked urethral folds, which now apparently form plate-like supports for
the thicker cavernous shaft, the terminal portion of which (the future glans) is
slightly more knob-like than that of the male. The inclosed urogenital opening is
narrow and lanceolate, clearly restricted to the region of the genital folds and not
extending onto the decurved glans. In general, the labio-scrotal swellings are
similar to those of the male; their caudal prolongation into a curving, connecting
postanal ridge is, however, more pronounced. Here it is without tubercular
enlargements, but is almost bisected by a median groove.

Some of the older embryos in this stage show further changes of a minor nature
in the phallic region. Thus the glans may be more strongly indicated by a further
increase in the density of its tissue (corpus cavernosum glandis) , so that it appears
as an opaque white tip in sharp contrast to the more translucent shaft. The post-
anal ridge, joining the tips of the labio-scrotal swellings, becomes more V-shaped,
while the inclosed anal opening shifts from a transverse to a longitudinal slit.

Reconstruction models of three embryos of this stage show quite clearly the
correlation between the sex differences of the external and internal organs. In one
specimen (No. 584a, 25 mm. CR) showing externally the long male type of uro-
genital opening, the tip of which extends onto the glans, the internal organs are
male also. The gonad is rather short, barely reaching to the rim of the pelvis,
while the tubar portion of the urogenital folds, containing the Wolfnan and Miil-
lerian ducts, unite with the dorsal wall of the bladder without the formation of a
genital cord (primordium of the uterus).

In the females modeled (No. 840, 24.8 mm. CR, and No. 782, 28 mm.) the
opposite condition is found. Externally the phallus of each shows a short uro-
genital opening. Internally the organs are decidedly more of the female type.
The gonads are slightly longer, while the tubar portions of the urogenital folds
(containing the Mlillerian ducts) unite to form a genital cord before fusing with
the dorsal wall of the bladder, thus producing a sinuous ridge which projects from
the dorsal wall of the bladder and partially subdivides the cavity of the pelvis.
In the older specimen the gonads are relatively larger and more compact.

Stage 9, 30 to 38, mm CR (figs. 18 to 23, males; figs. 22 and 29, females) .—The
changes in form between this and the previous stage are slight and consist mainly
of a further decurvation of the female phallus and a great condensation of the
glans area, although as yet the definite coronary sulcus has not been formed. In
the male, the urogenital opening still extends the entire length of the phallus and
there is no indication of the raphe".

DEFINITIVE PERIOD.

Stage 10, 38 to 45 mm. CR (figs. 24 to 28, 31, 32, males; figs. 30, 35, 36, fe-
males).— The series of important changes occurring between the lengths of 38 and
45 mm. culminates in the final differentiation of the external genitalia. The forma-

80 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

tion of an open coronary sulcus demarks the terminal, knob-like glans from the
remainder of the phallus. The most significant changes, however, are those in-
volved in the separation oi the urogenital opening and anus. These modifications,
more extensive in the male than in the female, are brought about by the trans-
formation of the open, gutter-like urogenital sinus into the tubular urethra, due to
the almost complete fusion of the urethral folds to form the raphe". At the same
time the labio-scrotal swellings migrate from their primitive lateral position to
a new one caudal to the penis, forming the scrotum.

An early stage in this series of modifications is shown in figure 24. The
ridge-like raphe now separates the anal opening from the proximal end of the
lenticular urogenital opening, the distal end of which continues to the tip of the
glans as a narrowing slit. Coincident with this fusion of the urethral folds to form
the raphe, the labio-scrotal swellings shift slightly, so that their greatest tumidity
now lies caudal to the penis instead of lateral to it, as heretofore.

Regarding the formation of the scrotum, Felix (1912) makes such conflicting
statements that it is not at all clear just what his conclusions are. In his prelimi-
nary account (p. 953) he states:

"The basal growth of the pars pelvina must produce a new area, interposed between
the base of the penis and the anal opening, and the best name for this is the unpaired
scrotal area (fig. 642)."

The figure referred to by him, according to my interpretation, is that of a
female specimen, and the area he labels as the unpaired scrotal area I should de-
signate as the posterior commissure. He continues:

"In embryos of 60 mm. head-foot length this (unpaired scrotal area) becomes raised
up in toto and forms the unpaired scrotal swelling, into which the two genital swellings,
which we may now term the scrotal swellings, extend from above .... As soon as
the descensus is complete this unpaired scrotal area alone forms the scrotal sack, the paired
scrotal swellings to the right and left of the penis fading out into the surrounding areas."

Later, in referring to the development of the labia majora in the female, he

says (p. 955) :

"The two genital swellings on this account are not prolonged anally to form an un-
paired scrotal tubercle, and consequently they do not fade out as in the male but persist
as the labia majora."

Finally, in his resume* of the development of the external genitalia he con-
cludes (p. 958) :

"In the male the tuberculum genitale fades out in all its parts and in its place there is
formed at the anal periphery of the phallus, from the unpaired scrotal area, a new swelling,
the unpaired scrotal sack."

An analysis of the above shows that he has made two statements implying
that the labio-scrotal swellings are transformed into the scrotum (in toto or in part),
and two other equally emphatic statements implying that they play no part in its
formation. Furthermore, if we accept his first statement, it is still doubtful

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 81

whether he considers that there is a caudal migration of the labio-scrotal swellings
into the scrotum, or whether the final formation of the latter is accomplished by an
anal prolongation of these swellings with an accompanying cranial atrophy.

Subsequent growth results in the gradual reduction of the broad urogenital
opening into a small aperture of varying size and shape, located just proximal to the
coronary sulcus. The reduction in the size of this opening is usually accompanied
by the development of a more or less pronounced epithelial rim and an increase in the
size of the epithelial tag into a comb-like ridge of varying size and shape. This
sequence of events is so clearly shown in figures 24 to 28, 31, and 32 that no detailed
description of the individual steps is necessary. Attention should be called, how-
ever, to the ridge-like development of the penial raphe in the embryo illustrated in
figure 27, where it forms a penial-scrotal frenulum, binding the penis to the scrotum
with a consequent decurvation. The production of this frenulum is quite character-
istic for most of the older embryos, although as a rule it does not become noticeably
developed until after the embryos have attained a length of 55 mm.

The changes taking place in the female at this time are not so extensive. As
in the male, a definite coronary sulcus is formed. The other modifications may be
considered as a faint paralleling of the more profound metamorphosis which the
male genitalia have been undergoing and consist chiefly of changes in the labio-
scrotal swellings. The caudal ends of these draw toward each other (fig. 35) and
finally fuse to form the posterior commissure (figs. 37, 38), definitely separating the
urogenital opening from the anus, although this is slight in comparison to the
corresponding separation in the male. In this manner, the labio-scrotal swellings
are transformed from separate, lateral swellings into a cranially open, horseshoe-
shaped rim partially encircling the phallus. Unlike the corresponding change in
the male, this fusion is not accompanied by either a caudal migration of these
swellings or by the formation of a definite raphe".

Regarding the formation of the posterior commissure, my observations are
decidedly at variance with the conclusions drawn by Wood-Jones (1913) that in the
adult (human) female this is not definitely present as a commissural bar separating
the vaginal orifice and anus, but that it merely shows as such when the genitalia are
placed in an unnatural position.

The posterior commissure may be considered as marking the advent of the
definitive period in the female, and from now on we may refer to the component
parts of the external genitalia by their adult terms. The labio-scrotal swellings
form the labia majora, connected by the posterior commissure. The cavernous por-
tion of the phallus becomes the clitoris, divided into glans and shaft. The urethral
folds constitute the labia minora, margining the persisting primary urogenital
opening (the urethro-vaginal orifice) . It must be pointed out, however, that the
use of these latter terms at this time is an arbitrary one, because the division of the
female phallus into clitoris and labia minora is never entirely complete and is con-
summated only with the formation of the frenulum clitoridis at some time after the
close of the period covered by the present investigation (100 mm. CR length).

82 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

For this reason, the strict usage of these terms would require the continued use of
the inclusive term phallus until the separation is culminated.

The phallus period is thus terminated by the formation of the raph.6 and the
scrotum in the male, and the posterior commissure of the labia majora in the
female. Thus the definitive characteristics of the external genitalia are organized
and the recognition of the sex of the embryos is correspondingly simplified. These
observations, in the strict sense, diverge from the statement made by Felix (1912,
p. 949) :

" The beginning of the sexual differentiation can hardly be assigned to a definite period,
since it takes place quite gradually and even in advanced stages presents difficulties to
diagnosis. We base our diagnosis on the position of the ostium urogenitale relative to
the coronary sulcus on the one hand and the anal opening on the other. Male embryos
are distinguished by the distal (distal here has reference to the base of the phallus)
circumference of the ostium urogenitale always lying in the coronary sulcus, while the
proximal circumference becomes more and more distant from the anal opening. Female
embryos may be recognized by the urogenital opening retreating away from the coronary
sulcus by the gradual closure of its distal part, while its proximal part always lies close
to the anal opening; the differentiating moment is, accordingly, both positive and nega-
tive in each sex."

While this statement, if it be interpreted broadly, may to a certain extent be
considered as containing the germ of the diagnostic point which I have found to be
of value in the early recognition of sex in the human embryo, it plainly does not
agree with my own observations, in that Felix limits it to the final period when I
find definitive changes taking place in the external genitalia, and at which time the
modifications are so characteristic that examination of the external genitalia almost
at once enables one to definitely say that one embryo is male and another is female.
It must be emphasized, however, that it is still necessary to make careful examina-
tion of the genitalia to obtain a correct diagnosis, because from this time throughout
the rest of the fetal period individual variations are more pronounced than in the
preceding embryonic period. In many of the females, for example, the clitoris
attains considerable size, and if this alone is considered, or if only a cursory examina-
tion is made, such an embryo might easily be mistaken for a male. (A comparison
of figs. 33 and 34, males, with figs. 37 and 46, females, will illustrate this point.)
In fact, in the majority of males between the lengths of 50 and 100 mm. CR,
the exposed penis seems to be comparatively shorter than the clitoris in females of
the same size, an effect which is largely produced by the varying development
of the penial-scrotal frenulum.

So few pronounced morphological changes take place in the developing geni-
talia during the rest of the period covered in this study that it will not be necessary
to divide them into successive stages.

Although I have not yet been able to make a complete study of the develop-
ment of the prepuce, my observations upon its formation from the external ex-
amination of older embryos may be of value and are therefore included in the
present account as suggesting the apparent formation of this structure, although

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 83

the final corroboration of this will have to be left for future investigation. For
this reason, it seems inadvisable at this time to enter into any comparison of the
various explanations that have been made to account for its origin.

After the separation of the phallus into shaft and glans by the formation of a
coronary sulcus, the glans remains as a naked, knob-like termination of the shaft
(in both sexes) until the embryo reaches a length of about 65 mm. CR. In males
of this size the first traces of the prepuce may be noted. The skin of the shaft
becomes elevated into a pair of conspicuous rolls on either side of the urethral
opening (fig. 41). Gradually these rolls join on the dorsal side of the shaft (fig. 42)
to form a flat ridge whose distal margin has grown out to cover the proximal edge
(corona) of the glans (embryos 75 mm. CR). The continued outgrowth of this
fold results in the gradual inclosure of the glans (figs. 47 to 50) until (embryos 98 to
100 mm. CR) the glans is completely covered by the prepuce.

Accompanying the development of the prepuce, there is a gradual shifting of
the urethral opening until (in embryos 85 to 90 mm. CR) it occupies a subterminal
position in the frenular notch of the glans. With the completion of the prepuce
this primary opening is entirely closed. Some time later a new (the secondary or
definitive) terminal urethral opening is formed near the tip of the glans.

But few important changes take place in the female genitalia in this later period
(50 to 100 mm. CR length), these being mainly concerned with the continued
growth of the labia majora and the beginnings of prepuce formation. The labia
majora increase somewhat in height, so that the inclosed structures become more
submerged in the rim which they form, although this submergence is by no
means completed at this period.

The formation of the prepuce is considerably more involved than in the male
and all of the folds included in its complete development do not make their appear-
ance until some time after the close of the period under consideration. The only
observations made are concerned with the glandular portion. This fold apparently
develops in much the same way as it does in the male, although it has not been
studied as closely and is not as well shown in the figures. Its growth is by no means
as rapid as in the male, with the result that the glans of the clitoris is not completely
covered by it at the close of this period (100 mm. CR). As has already been noted,
the complete separation between the clitoris proper and the labia minora does not
take place until considerably later.

84 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

SUMMARY.

A more comprehensive idea of the sequence of events in the development of
the external genitalia may be gained if we briefly review their development in each
sex.

DEVELOPMENT OF THE MALE GENITALIA.

The genital tubercle, from which is derived the phallus and which includes the
urethral and anal openings, arises as a broad conical eminence barely separated
cranially from the umbilical cord, but with a well-defined caudal slope. Its
rounded free end, as the primordium of the glans, is faintly separated from the basal
portion. The lateral slopes, below the glans area, extend into the outlying basal
tissue as the "lateral buttresses, " which I interpret as enlargements for the develop-
ing corpora cavernosa. The caudal slope is bisected by the shallow urethral groove
whose margins are slightly elevated into the urethral folds. Distally, these margins
merge into the glans area, while proximally they continue into the basal enlarge-
ments surrounding the anal pit. In the majority of embryos which have attained
a length of 12 mm. the urethral membrane is ruptured to form the primitive uro-
genital opening, although a few cases have been found in which this perforation
had not taken place at the stage of 16 to 17 mm. Practically from its first appear-
ance the urethral groove shows a sex difference in its length, this difference being
further accentuated upon the formation of the urogenital opening.

As development proceeds, the elongation of the tubercle and the medial migra-
tion of its lateral buttresses transform it into the somewhat cylindrical phallus,
whose base is separated from the surrounding body areas by the newly formed
labio-scrotal swellings (embryos 17 to 19 mm.). The latter, appearing first as
broad, elevated areas extending laterally from the base of the phallus, are soon
transformed into swollen ridges separated, in both sexes, from the base of the
phallus by the lateral phallic grooves. It may be emphasized that in the majority
of specimens the labio-scrotal swellings seem to be from the first separated from the
phallus. In fact, only one embryo was found which showed these areas as definitely
merging into the phallus, while in all of the others of about the same age the swell-
ings were already separated from the phallus by pronounced grooves, the lateral
phallic grooves.

Accompanying the elongation of the tubercle into the phallus, the uro-
genital opening becomes gradually more sharply outlined and the sex differ-
ence in its length correspondingly emphasized. The urethral folds which form
the margins of the opening also increase in definiteness, partly through their
own elevation and partly through their lateral demarcation from the cavernous
portion of the shaft; the result of these combined factors being that the folds
extend as plate-like caudal projections from the more cylindrical shaft. At the
same time the terminal glans has increased in density so that it now forms an opaque
white area in contrast to the translucent shaft, although the limiting coronary
sulcus is not formed until the close of the phallus period.

As a result of these combined changes the male phallus at this time is markedly
different from that of the female. The length of the urogenital opening is still,

DEVELOPMENT OF EXTERNAL GENITALIA IN 1HE HUMAN EMBRYO. 85

however, the chief diagnostic feature in the two sexes. In the male it consists of a
long slit extending from the base to the apex of the phallus, whereas in the female
it extends only to the base of the glans. In the male it is more open distally, the
apposition of the urethral folds reducing it proximally to a mere slit. This is in
marked contrast to the condition in the female, where the opening (restricted to the
shaft) has exactly the opposite shape, being broadest basally, while the terminal
portion is narrowed into a slit.

No further pronounced morphological changes take place in the male genitalia
until after the embryo has reached a CR length of 38 mm. At this time a series of
changes begins which result in the transformation of the male genitalia into approxi-
mately their final form (about 45 mm.) and in the complete separation of the uro-
genital opening from the anus. The gutter-like urogenital sinus is transformed
into the tubular urethra by the merging of the basal portions of the urethral folds
into the raphe", reducing the primitive urogenital orifice to an irregularly shaped
opening on the distal portion of the shaft. At the same time the glans becomes
sharply defined by the formation of a wide coronary sulcus and the labio-scrotal
swellings assume their final position caudal to the penis to form the scrotum, the
halves of which become more or less closely approximated to the raphe in the mid-
ventral line, although they never entirely lose their bilateral character.

This period (45 mm. CR length) marks the completion of the definite sex
differentiation, characterized by the development of definite structural features in
the male in sharp contrast to the lack of these characters in the female.

As growth continues, the gradual constriction of the urethral opening syn-
chronous with the formation of the prepuce, to be described later, is accompanied
by an outgrowth of its rim which, in embryos 60 to 85 mm. CR length, results in
the formation of a decided cup in the bottom of which the opening is located. With
the later stages of prepuce formation the urethral opening is shifted to a more
terminal position in the frenular notch of the glans. It then becomes entirely closed
and eventually there is formed the new, permanent terminal urethral opening on the
apex of the glans.

The first evidence of the prepuce is found in embryos of about 65 mm. CR
length, at which time it can be recognized as a pair of swellings on each side of the
urethral opening. Gradually these swellings fuse together over the dorsum of the
shaft to form a flattened ridge of skin whose distal margin has enveloped the
proximal margin (corona glandis) of the glans (embryos 75 mm. CR length).
Subsequent growth results in the progressive inclosure of the glans by this distally
migrating fold of skin until at about 100 mm. CR length the originally naked glans
is entirely covered. It must be emphasized that these observations upon the forma-
tion of the prepuce are based only upon the external examination of the genitalia
of these older embryos and have not been confirmed by histological study. For
this reason they must be considered merely as suggestions of what apparently takes
place and may later be corroborated or contradicted when it becomes possible to
make a study of sectioned embryos showing this development.

80 DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

DEVELOPMENT OF FEMALE GENITALIA.

The entire process of development of the external genitalia in the female is
accompanied by fewer pronounced morphological changes than occur in the
male. It is noteworthy, however, that (in spite of this greater simplicity in struc-
ture) completion of development is more protracted, so that the final differentiation
of the female genitalia, although brought about by comparatively minor changes,
does not synchronize with the more complete transformation of the male (45 mm.
CR length), but instead, beginning with a slight change at a length of about 50 mm.,
extends as a gradual process throughout a considerable period of early fetal life.

From its beginning until the stage of 21 to 25 mm. the genital tubercle of the
female closely resembles that of the male except for the shorter urethral groove.
About this period the female shows the beginnings of the caudal decurvation, which
is apparently brought about by an excess in the growth of the cavernous over the
urethral regions of the phallus. At the same time the urethral folds have become
compressed into plate-like caudal projections supporting the slightly overhanging
glans, which in this way is more clearly defined than in the male. As has already
been pointed out, the coronary sulcus is not formed in either sex until the embryos
reach a length of about 45 mm. From 25 mm. to 45 mm. this caudal decurvation
becomes a more and more pronounced characteristic of the female phallus, and for
this reason is a diagnostic feature of increasing importance as development proceeds.

In the female important changes occur at about the stage of 45 to 50 mm. CR
length, and these likewise mark the termination of the phallus period. While much
less extensive than the correlated changes in the male, they are nevertheless char-
acteristic and indicate the approximate assumption of the final form. The most
pronounced modification occurring at this time is that the caudal ends of the
labio-scrotal swellings grow towards each other and finally join in the midventral
line to form the posterior commissure (50 mm. CR length). In this manner these
originally paired swellings are transformed into a cranially open, horseshoe- shaped
rim, inclosing the rest of the external genitalia and separating them from the anus.

The formation of the posterior commissure in the female thus synchronizes
with the formation of the raphe in the male and may be considered as representing
the advent of the final differentiation, and from now on we may refer to the genitalia
by their adult terms. The labio-scrotal swellings form the labia majora. The
glans and cavernous portion of the phallus may be considered as the clitoris, and
the urethral folds as the labia minora. The inclosed primary urogenital opening
may now be called the urethro-vaginal orifice. It must be pointed out, however,
that the application of these terms at this time is an arbitrary one, because the actual
separation of the female phallus into these more definitely adult structures does not
take place until some time after the close of the period included in the present study
(100 mm. CR). Strict accuracy would demand that, until such division had been
completed by the formation of the frenula clitoridis, the inclusive term phallus be
retained.

Because of the persistence of the urethral folds (labia minora) in the female and
their failure to fuse together as they do in the male, the female phallus retains a more
conical outline than does the penis of the male.

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO. 87

In female embryos of 60 to 100 mm. CR length there is shown a partial develop-
ment of the prepuce, although the complicated set of folds involved in its complete
formation is not produced at this time and only partial growth of the glandular
portion is completed at the close of this period. While the glandular prepuce is
apparently formed in much the same way as it is in the male, its growth is markedly
slower, and in embryos 100 mm. long the glans is not completely surrounded by it.

In these older embryos (60 to 100 mm. CR length) there is also some increase
in the height of the labia majora, so that the inclosed portions are somewhat
submerged in the rim thus formed, although this submergence is by no means as
complete as it becomes in later fetal life. It should also be noted that in fetuses up
to 100 mm. CR the labia majora are still cranially separated and there is no indi-
cation that they play any part in the formation of the mons veneris.

CONCLUSIONS.

1 . There is a morphological sex difference in the external genitalia of the human
embryo practically from their first appearance, characterized by marked difference
in the length of the urethral groove upon the caudal slope of the genital tubercle.
In some embryos the distal end of the groove extends onto the glans region (males),
while in others it does not reach to this region (females) .

2. This should prove to be a fairly reliable character for the recognition of sex
in human embryos at an earlier period than has heretofore been possible.

3. From the first appearance of the genital tubercle, there apparently is no
indifferent or undifferentiated period in the development of the external genitalia
before they acquire their definite male or female characteristics.

4. The caudal decurvation of the female phallus, first pointed out by Herzog,
is a valid diagnostic characteristic and is applicable at an earlier period than was
first indicated.

5. The phallus period is definitely terminated in both sexes by the assumption
of approximately the adult form of the genitalia. In the male this is brought
about by the fusion of the urethral folds into the raphe" and the accompanying caudal
migration of the labio-scrotal swellings to form the scrotum. In the female it is
characterized by the fusion of the caudal ends of the labio-scrotal swellings to form
the posterior commissure.

88

DEVELOPMENT OF EXTERNAL GENITALIA IN THE HUMAN EMBRYO.

BIBLIOGRAPHY.

VAN ACKEREN, F., 1888. Beitrage zur Entwicklungs-

geschichte der weiblichen Sexualorgane de3

Menschen. Zeitschr. f. wiss. Zool., vol. 48.
ALLEN, B. M., 1904. The embryonic development of the

ovary and testis of the mammals. Amer.

Jour. Anat., vol. 3. 89-154.
BAYER, 1903. Entwicklungsgeschichte des weiblichen Geni-

talapparatea. Vorlesungen u. allg. Geburts-

hilfe, vol. I1, 1-104.
B£>HM, J., 1905. Die ausseren Genitalien des Schafea.

Morph. Jahr., vol. 34, 248-320.
BUCHANAN, G., and E. A. FRASER, 1918. The development

of the urogenital system in Marsupialia, with

special reference to Trichosurus vulpeeula.

Part I. Jour. Anat., vol. 53, 35-95.
DIMPFL, H., 1906. Die Teilung der Kloake bei Cavia

cobaya. Morph. Jahrb., vol. 35, 17-65.
DISSE, J., 1905. Untersuchungen uber die Umbildung der

Kloake und die Entstehung dea Kloaken-

hockera bei Talpa europea. Anat. Hefte,

Abth. 1, vol. 27, 479-533.
DCRBECK, W.. 1907'. Die ausseren Genitalien deb

Schweines. Morph. Jahrb., vol. 36, 517-543.
, 19072. Die ausseren Genitalien der Hauskatze.

Morph. Jahrb., vol. 36, 544-565.
—, 19073. Tabellarische Ubersicht der Genitalent-

wicklung bei Saugetieren. Morph. Jahrb.,

vol. 36, 566-569.
ECKER, A., 1851-59. Icones Physiologicae. Erlauterungs-

tafeln zur Physiologie und Eutwicklungsge-

Bchite. Leipzig.
FELIX, W., 1912. The development of the urogenital organs.

Manual of Human Embryology (Keibel and

Mall). Philadelphia.
FLEISCHMANN, A., and pupils, 1902-1907. Morphologische

Studien uber Kloake und Phallus der Amuioten.

Morph. Jahrb.:

IV. Die Saugetiere (Fleischmann), vol. 30, 653-666.
V. Die Stilistik dea Urodaum und Phallus bei den

Amnioten (Fleischmann), vol. 30, 666-675.
VI. Kloake und Phnllua des Schafes und Schweines

(Schwartztrauber), vol. 32, 23-57.
VII. Historisch-kritische Betrachtungen (Fleibchmann),

vol. 32, 58-96.
VIII. Die Stilistik des Urodaums (Fleischmann), vol. 32,

97-103.
IX. Die ausseren Genitalien des Schafes (Bohm), voi. 34,

248-320.
X. Die Teilung der Kloake bei Cavia cobaya (Dimpfl),

vol. 35, 17-64.
XI. Das Analrohr dea Schafes (Schwartztrauber), vol.

35, 65-74.
XII. Bau und Entwicklung der ausseren Genitalien bei

Cavia cobaya (Gruber), vol. 36, 1-26.

XIII. Die ausseren Genitalien des Schweines (Durbeck),

vol. 36, 517-543.

XIV. Der ausseren Genitalien der Hauskatze (Durbeck),

vol. 36, 544-565.
XV. Tabellarische Ubersicht der Genitalentwicklung bei

Saugetieren (Durbeck), vol. 36, 566-569.
XVI. Die Stileharactere am Urodaum und Phallus

(Fleischmann), vol. 36, 570-601.
FRASER, E. A., 1919. Development of the urogenital

system in the Marsupialia, with special

reference to Trichosurus vulpeeula. Part II.

Jour. Anat., vol. 63, 97-129.
FHIEDENTHAL, H., 1911. Das epithelhornchen am Penis dea

Menschenfotus als Rest eines auruckgebildeten

Stachelhbrnchens. Arbeiten aus dem Gebiet der

Experimentellen Physiologie, vol. 2, p. 87. Jena.
GRDBER, K., 1907. Bau und EntwickJung der ausseren

Genitalien bei Cavia cobaya. Morph. Jahrb.,

vol. 46, 3-26.
HENNEBERQ, B., 1914. Beitrag zur Entwicklung der

ausseren Genitalorgane beim Sauger. Erster

Teil. Anat. Hefte, vol. 50, Abth. 1, 423-497.
, 1917. Same; zweiter Teil. Ibid., vol. 55, Abth. 1,

227-414.

HERZOQ, F., 1904. Beitrage zur Entwicklungsgeschichts

und Histologie der mannliehen Harnrohre.

Arch. f. mikr. Anat., vol. 63, 710-747.
KEIBEL, FRANZ, 1896. Zur Entwicldungsgeschichte des

menschlichen Urogenitalapparates. Arch. f.

Anat. u. Physiol., Anat. Abth., 55-156.
KOLLMANN, J., 1907. Handatlas der Entwicldungsgeschichte

des Menschen, 2 vols., zweiter Teil. Jena.
KRASA, F. C., 1917. Die Entwicklungsgeschichte des Uro-

genitalsystems beim Maulwurf (Talpa europea).

Anat. Hefte, vol. 55, Abth. 1, 443-508.
MiiLLER, JOB., 1830. Bildungsgeschichte der Genitalien.

Dilsseldorf.
NAGEL, W., 1889. Ueber die Entwicklung des Urogenital-

syatems beim Menschen. Arch. f. mik. Anat.,

vol. 34, 269-384.
, 1892., Ueber die Entwicklung dor Urethra und

des Dammea beim Menschen. Ibid., vol. 40,

264-287.
, 1894. Ueber die Entwicklung der inneren and

ausseren Genitalien. Arch. f. Gyn., vol. 45,

453-477.
, 1897. Entwicklung und Entwicklungsfehler der

weiblichen Genitalien. Veil's Handb. d. Gyn.,

1st ed., vol. 1,519-628. Wiesbaden.
OTIS, W. J., 1906. Die Morphogenese uud Histogenese des

Analhockers nebst Beobachtungen uber die

Entwicklung des Sphincter ani externus beim

Menschen. Anat. Hefte, vol. 30, Abth. 1, 199-

258.
POHLMAN, A. G., 1904. Has a persistence of the Mullerian

ducts any relation to the conditions of crypt-

orchidism? Amer. Med., vol. 8, 1003-1000.
, 1911. The development of the cloaca in human em-
bryos. Amer. Jour. Anat., vol. 12, 1-23.
RATHKE, H., 1832. Abhandlungen zur Bildunga- und

Entwicklungsgeachichte dea Mensthen und dor

Tiere.
RETTERER, E., 18901. Sur 1'origine et 1'evolution do la

region anogenitale des Mammiferes. Jour, de

1'Anat. etde Physiol., vol. 26, 126-151, 153-216.
, 18902. Du developpement du prepuce, de la

couronne du gland et du col du penis chez

1'embryon humain. C. R. Soc. de Biol., No.

29, p. 528.

, 19051. Du role de 1'epithelium dana le developpe-
ment des organes genito-urinaires externea.

C. R. Soc. Biol., Paris, vol. 57, 1. 1040-1043.
, 19052. Du developpement et de la structure des

raphea des organes genito-uriuaires. C. R.

Soc. Biol., vol. 57, 2, 22-25.
, 1914. Developpement et histogenese comparee dea

organes genitaux externea. Jour, d'urol. rned.

et chir., vol. 6, 157-173.
RoacHULETZ, V., 1890. Beitrage zur Entwicklungageschichte

des Genital-hockers beim Menachen und beim

Schwein. Diss., Berlin.
SCHWARTZTRAUBER, J., 1904. Kloake und Phallus des

Schafea und Schweines. Morph. Jahrb., vol.

voi. 32, 23-57.
,1906. Das Analrohr dea Schafes. Ibid., vol. 35, p.

65-74.
SPULER, A., 1910. Ueber die normale Entwicklung des

weiblichen Genitalapparat.es. Veit's Handb.

f. Gyn., 2d ed., vol. 5, 573-650.
TIEDEMANN, F., 1813. Anatomie der kopflosen Missge-

burten. Landshut.
TOURNEUX, F., 1889. Sur le developpement et 1'evolution

du tubercule genital chez le foetus humain

dana les deux sexea, avec quelquea remarques

concernant le developpement des glandes pro-

statiques. Jour, de 1'Anat. et de la Physiol.,

vol. 25, 229-263.
WOOD-JONES, F., 1913. Some points in the nomenclature of

the external genitalia of the female. Jour.

Anat. and Physiol., vol., 148, 73-80.
, 1914. The morphology of the external genitalia of

the mammals. Lancet, 1017-1023; 1099-1103.

SPA'JLDING

PLATE I

J. F. Didusch fecit. J

FIG. 1. — No. 407, 9 nun. X 21.

FIG. 2. — No. 792, 8 mm.

FIG. 3. — No. 2702, 8 mm. X Iv5.

FIG. 4. — No. 1784«, 12 mm. X 21.

FIG. 5.— No. 1936, 14 mm., male. X 1 I ">-

FIG. 6.— No. 2023, 15 mm., female. X 14.5.

NOTE: Figures 1, 3, 4, 5, and G are drawings made directly from the specimen. Figure 2 was drawn fn nn thr model.

remaining figures are photographs: -<>IIH> of these, owing to difficulty in obtaining clear negatives, liavr I n retouched to

bring out certain details of structure that would otherwise be lost.

SPAULDING

PLATE 2

FIG. 7.— No. 1930, 14 mm., male. X 12.

FIG. 8. — No. 955, 17 mm., male. >< 12.

FIG. 9. — No. 2S, 19 mm., male. X 0. (Insert, lateral view.)

FIG. 10. — No. 3S9a, 21 mm., male. X 6. (Insert, lateral vicw.i

FIG. 11. — No. 2023, 15 mm., female. X 12.

FIG. 12.— No. 1750, 17.2 mm., female. X 12.

FIG. 13.— No. 684, 20 mm., female. X 6.

FIG. 14.— No. 1!I4, 21 mm., female. X 6.

22

FIG. 15.— No. 590. 23 mm., male. X 6.

FIG. 10. — No. 1900-60/*, 25 mm., male. X 6.

FIG. 17.— No. 879r, 29.0 mm., male. X 6.

FIG. 18. — No. 1022rf, 32 mm., male. X 6. (Insert, lateral view.)

FIG. 19. — No. 2393, 23 mm., female. X 6.

FIG. 20. — No. 1900-60a, 25 mm., female. X 6.

FIG. 21.— No. 950, 29 mm., female. X 6.

FIG. 22.— No. 13586, 33 mm., female. X G.

SPAULDING

PLATE 3

FIG. 23. — No. 1206, 37 mm., male. < 4.
FIG. 24. — No. 948, 45 mm., male. X 4.
FIG. 25.— No. 652k, 44 mm., male. X 4.
FIG. 26. — No. 217, 45 mm., male. < 4.
FIG. 27.— No. 693, 45 mm., male. < 4.
FIG. 28. — No. 607, 38 mm., male. < 4.
FIG. 29. — No. 22506, 36 mm., female. X 4.
FIG. 30.— No. 2250a, 40 mm., female. X 4.

FIG. 31. — No. KiM'i, -It; nun., male (permeal view). X 4.

FIG. 32. — Same (lateral view). X 4.

FIG. 33. — No. 1474rf, 5lS mm., male. X 4.

FIG. 34. — No. S47. 5s. s mm., male. X 4.

FIG. 35. — No. 15976, 44 mm., female (perineal view). X 4.

FIG. 36. — Same (lateral view). < 4.

FIG. 37.— No. 13SX, 51 mm., female. X 4.

FIG. 38. — No. 16256, 54 mm., female. X 4.

SPAULDING

PLATE 4

t

51

FIG. 39. — No. 1183, 60 mm., male (perincul view1!. X 4.

FIG. 40. — Same (lateral view). X 4.

FIG. 41. — No. 1163, 68 mm., male. X 4. (Insert, lateral view).

FIG. 42. — No. nn34t, 75 mm., male. X 4.

FIG. 43. — No. 907, 60 mm., female (perineal view). X 4.

FIG. 44. — Same (lateral view). X 4.

FIG. 45.— No. 12826, 65 mm., female. X 4.

FIG. 46. — No. 1542, 69 mm., female. X 4.

Fie;. 47. — No. 2026, 80-90 mm. (est), mak.
FIG. 4*.— No. 1705o. S3. 2 mm., male. X 4.
Flu. 4!!.— No. ls.-i2. 95 mm., male. X 3.
FIG. 50. — No. 834, 98 mm., male. X 3.
FIG. 51. — No. 1455, 7* mm., female. X 4.
FIG. 52. — No. 1474!), 84 mm., female. X 4.
FIG. 53.— No. 1831, 93.5 mm., female. X 4.
FIG. 54.— No. 1476, 100 mm., female. X 4.

X 4.

CONTRIBUTIONS TO EMBRYOLOGY, No. 62.

FURTHER EXPERIMENTAL STUDIES ON FETAL ABSORPTION,

III. THE BEHAVIOR OF THE FETAL MEMBRANES AND PLACENTA OF THE GUINEA-
PIG TOWARD TRYPAN BLUE INJECTED INTO THE MATERNAL BLOOD-
STREAM.

IV. THE BEHAVIOR OF THE PLACENTA AND FETAL MEMBRANES OF THE RABBIT
TOWARD TRYPAN BLUE INJECTED INTO THE MATERNAL BLOOD-STREAM .

BY GEORGE B. WISLOCKI,

Of the Anatomical Laboratory, Johns Hopkins Medical School.

With one plate.

89

FURTHER EXPERIMENTAL STUDIES ON FETAL ABSORPTION,

III. THE BEHAVIOR OF THE FETAL MEMBRANES AND PLACENTA OF THE GUINEA-PIG
TOWARD TRYPAN BLUE INJECTED INTO THE MATERNAL BLOOD-STREAM.

Our knowledge of the behavior of vital dyes in the pregnant animal rests upon
the careful observations of Goldman (1909) upon mice and rats. He observed that
when vital dyes, such as trypan blue and pyrrhol blue, were injected into the
maternal circulation, they were not transmitted through the placenta to the fetus.
The barriers which prevent the dye from entering the fetus are the chorionic ecto-
derm of the placenta and the epithelium of the vitelline membrane. In the rat and
mouse the latter constitutes the outermost fetal covering. He also noted the
interesting fact that the placenta acts as an attraction center for vital dyes, since,
with the initiation of pregnancy in an animal which has been vitally stained, the
dye is to a great extent given up by the liver, spleen, and other tissues in which it is
stored and is conveyed by the blood-stream to the developing placenta, where it is
immediately absorbed by the chorionic epithelium and the endodermal cells of the
vitelline membrane.

The present studies, which are a continuation of a series of investigations on
fetal absorption (Wislocki, 1920), concern the behavior of the placenta and fetal
membranes of the guinea-pig and rabbit toward trypan blue. Although similar in
many respects to the observations of Goldmann on the rat and mouse, there are
differences sufficiently great to give interest to a description of the process of vital
staining in these rodents.

The guinea-pigs used in these experiments were stained by the technique
usually employed in administering vital dyes. There was no opportunity to study
the placenta and fetal membranes in the early stages of their development, and the
following description applies only to the period from the time the fetus reaches a
length of about 15 mm. up to term.

In an animal killed at the end of a week after six injections of dye-stuff the
maternal tissues were found to be deeply stained. Each fetus was completely
surrounded by its vitelline membrane, which was stained dark blue, the color being
deepest in a villous zone adjacent to the placenta and fading somewhat towards
the antimesometral pole. The amniotic fluid was usually colorless but sometimes
showed a slight tinge in deeply stained animals. The amnion, when stripped away
from the vitelline membrane and examined in salt solution, as a rule was seen to be
uncolored; in rare instances, however, it showed a blue tint, attributable to large
and repeated injections of the dye. The fetus and umbilical cord likewise were
usually entirely unstained, although a barely perceptible blue coloration of the fetus
was noted in several instances, associated with a bluish tinge in the amniotic fluid
and membrane.

Goldmann noticed in the mouse and rat the staining of the amniotic membrane
and fluid, but never any of the fetus. He believed that the dye in the amnion must

91

92 EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

be derived from the neighboring vitelline membrane and this, he thought, was an
argument in favor of the origin of the amniotic fluid, in part at least, from the
maternal blood. It is more probable that the trypan blue occasionally seen in
the amniotic fluid of rodents first enters the fetal blood-stream through both the
vitelline membrane and the placenta. A trace of the dye in the fetal blood might
readily escape detection, whereas, as soon as it had diffused from there into the
colorless amniotic fluid it would attract attention. That it immediately enters
the amnion from the vitelline membrane, without first entering the fetal circulation,
seems improbable, for the vitelline membrane is well vascularized by the omphalo-
mesenteric vessels and, should a dye diffuse from its epithelium through the base-
ment membrane, it would necessarily be swept into these large vascular channels
before it could enter the amnion.

The outside of the placenta appeared blue and on gross section the entire
organ was found similarly stained with the exception of an area about 5 by 3 mm.,
which occupied a central position between the placental labyrinth and decidua (fig. 1) .
As will be shown later, this white tissue consists of chorionic villi possessing a peculiar
arrangement and designated by Duval as "the roof of the central excavation."

In many rodents, such as the rat, mouse, and guinea-pig, the vitelline mem-
brane, as described by Duval and Sobotta, becomes the outermost fetal covering
and persists throughout intrauterine life. Its epithelial surface comes in direct
contact with the uterine wall and serves to nourish the fetus by the absorption of
embryotrophe and the assimilation of extra vasated maternal blood. This is more
readily accomplished by the development of numerous villi which project into the
uterine space and are covered by a single layer of cylindrical, phagocytic epithelial
cells. This layer stains very rapidly and deeply with trypan blue (figs. 1 and 2).
A single injection suffices, after a few hours, to cause the cytoplasm of the cells to
become heavily laden with fine blue granules. One is justified in speaking of the
cells covering the villi as an attraction center for vital dyes, for nowhere else in the
body are the dye-stuffs so rapidly taken up.

The villi show their greatest development near the line of attachment to the
placenta, while towards the equator of the vitelline membrane they become
progressively lower. Finally they disappear, so that the membrane at its anti-
mesometral pole is perfectly smooth. During this transition the epithelium changes
from a cylindrical to a cuboidal type. The latter is probably less phagocytic than
the former, for the cuboidal cells contain relatively few trypan-blue granules.

The villi cease suddenly near the attachment of the vitelline membrane to the
placental surface (figs. 1, 2), and the cylindrical, vitally staining epithelium changes
abruptly to a delicate, flattened type which fails to absorb the dye. This un-
stained, flattened tissue completely lines the angle (sinus entodermaticus) formed
by the membrane at its attachment to the placenta. As the vitelline membrane
(ectoplacental endoderm) spreads out laterally over the surface of the placenta, its
epithelium again undergoes a change (fig. 2). The cells proliferate, forming in
many places small tufts or clusters, the older cells of which gradually become
constricted off by new cells arising from the basement membrane.

EXPERIMENTAL STUDIES ON FETAL ABSORPTION. 93

Duval noted the remarkable papillary hyperplasia of the epithelium in this
region but did not ascribe to it any function. These cells do not stain vitally, as do
those of the neighboring villi, and therefore their function possibly is not one of
phagocytosis or absorption. It is interesting to note that the greatest proliferation
of the epithelium occurs during the middle of pregnancy and that towards term it
ceases. At birth the vitelline membrane in this region has been reduced to a single
layer of columnar or cuboidal epithelial cells resting upon a simple basement
membrane.

We shall now consider the vitally stained placenta. From a knowledge of the
behavior of the placenta of the mouse and rat towards trypan blue, we should
expect to find the dye abundantly present in the chorionic epithelium. We should
look for it particularly in the giant cells and syncytium of fetal origin which invade
and destroy the maternal tissue and attach the placenta to the uterine wall, and also
in the delicate epithelial syncytium of the labyrinth which covers the villi and forms
a barrier between the maternal blood-stream and the fetal capillaries.

To the unaided eye the placenta appears quite blue, with the exception of a
tiny flattened area near its center, which always remains unstained (figs. 1, 4). On
microscopic examination one is surprised to find very little of the dye stored within
the chorionic epithelium. If the animal has not been deeply stained one may fail to
find any at all. In a well-stained guinea-pig, it can be distinguished as a fine blue
tracery in the chorionic epithelium of the placental labyrinth (fig. 3). The stain is
faint, however, as compared with its exhibition in the same cells of the mouse and rat.

In the giant cells and syncytium of fetal origin which invade the decidua and
form an irregular boundary between the fetal and maternal tissues, commonly
referred to as the central zone or "Umlagerungszone," no trypan-blue granules are
found. This is surprising, since in the injected mouse and rat these cells are filled
with blue pigment throughout gestation. Possibly the fetal syncytium stains
vitally in the guinea-pig in younger stages, when the ectoplacental cone is sending
out syncytial roots and is burrowing its way into the decidua. However, in the
youngest fetus (15 mm.) observed in these experiments no dye was found, nor was
any found in the syncytium which, in the guinea-pig, divides the placental laby-
rinth into numerous lobules, the so-called interlobular syncytium.

We have already mentioned that the roof of the central excavation, a structure
peculiar to the placenta of the guinea-pig, stands out in the gross as an unstained
area. When the mesoderm containing allantoic vessels invades the ectoplacenta,
the latter is virtually hollowed out and a central core of mesodermal tissue is formed.
This has been termed by Duval the central excavation. Part of this mesoderm
finally invades the central zone, hollows out the masses of trophoblast situated there,
spreads out over a broad area, and sends into the decidua a number of villous
processes covered by several layers of ectodermal cells. This conspicuous structure,
the function of which is unknown, has been designated by Duval as the roof of the
central excavation. It persists throughout gestation but the decidual tissue which it
embraces between its slender, finger-like processes gradually degenerates and, as
term approaches, becomes converted into a homogeneous substance. The villi

94 EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

themselves, together with the vessels supplying them, appear to degenerate to a
great extent. It is interesting to note that this same area is conspicuously blue
when the vital dye is injected into the fetal instead of into the maternal blood-
stream (figs. 4, 5).

The behavior of the surface of the placenta, just beneath the vitelline mem-
brane, is also noteworthy. The proliferation of the epithelium of the ectoplacental
endoderm has already been described and need not be referred to further. Very
early in the development of the placenta a layer of fetal syncytium comes to lie
directly beneath the membrane. The periphery of this syncytium, however, very
soon undergoes a change. It shapes itself into two or three layers of ectodermal
cells with large round nuclei surrounded by a wide zone of vacuolated cytoplasm.
Duval called attention to these cells on the surface of the placenta of the guinea-pig
and referred to them as giant cells. In a section through the border of the placenta,
at about the middle of pregnancy, one sees the vitelline membrane beneath which
are layers of these giant cells, beneath which in turn is a layer of border syncytium
containing large lacunse, and finally the placental labyrinth (fig. 6) . In an animal
which has been stained with trypan blue the ectoplasm of the giant cells is filled
with fine granules of the dye (fig. 6) . As the end of gestation approaches, these cells
become shrunken and in part disintegrate and disappear.

In the decidua serotina groups of free macrophages, containing varying
amounts of trypan blue in their cytoplasm, were encountered. Many of these wan-
dering cells appeared to be undergoing degeneration. They were similar to those
described by Goldmann in the mouse and rat, which were thought by him to be
important in the transfer of nutrient material, such as glycogen, from the tissues
and vessels of the decidua to the fetal ectoderm. In the myometrium and serosa,
in the neighborhood of the placental attachment, hundreds of similarly stained
cells were observed. Goldmann believed that in the mouse and rat many of these
are derivatives of the serous cells covering the peritoneal surface of the uterus.
This may also be the case in the guinea-pig, for the peritoneal cells were con-
spicuously loaded with the dye.

One further aspect of the behavior of trypan blue is worthy of description.
In several of the guinea-pigs used in these experiments large injections of trypan
blue were administered, resulting in mild inflammatory changes in the placenta, to
judge from a poor staining reaction on the part of the syncytium and the presence
of numerous polymorphonuclear leucocytes in the sinusoids. It was observed that
the polymorphonuclear leucocytes in the maternal blood-spaces contained numer-
ous tiny granules of trypan blue within their cytoplasm (fig. 7) . It has been the
general belief that none of the elements of the circulating blood normally stain with
vital dyes. Downey (1917), however, has attempted to show that the polymorpho-
nuclear leucocyte is capable of ingesting vital dye under certain conditions. In
the circulating blood conditions are unfavorable, he claims, for phagocytosis of any
kind to occur, and he bases his argument on the failure of the polymorphonuclear
leucocytes, while moving rapidly, to take up bacteria, vital dye, or any other par-
ticulate matter. Downey has shown experimentally that in the areas where the

EXPERIMENTAL STUDIES ON FETAL ABSORPTION. 95

velocity of the blood-stream is diminished, or outside of the blood-vessels, the
leucocytes stain with vital dyes, and he believes that this staining is quite compar-
able to that seen in macrophages. The question is still debatable whether or not
his assumption is correct and whether the staining should be considered as physio-
logical rather than as an indication of cell injury.

The placentse in which the staining of the polymorphonuclear leucocytes was
observed showed evidence of slight inflammatory changes, but there was nothing
in the appearance of the leucocytes themselves to condemn them as moribund.
Further evidence is necessary, however, before it can be decided definitely whether
or not true vital staining of polymorphonuclear leucocytes ever occurs.

In conclusion, a few observations upon the fate of a fine suspension of carbon
particles, such as one finds in india ink, injected into the blood-stream of the preg-
nant guinea-pig, will be described. Several animals were injected and killed on the
second, fourth, and sixth day thereafter, respectively. In these animals the liver
and spleen appeared very black while the color of the other abdominal organs was
almost normal. On opening the thorax the lungs were seen to contain considerable
carbon. The bone-marrow was also conspicuously black, and traces of carbon were
visible to the unaided eye in many other tissues, such as the ovaries, pancreas,
omentum, diaphragm, and mesentery. In the uterus, placenta, or vitelline
membrane, however, no carbon was recognizable. The fetuses appeared perfectly
normal, the amniotic fluid clear and colorless. Under the microscope particles
of carbon were found in great abundance in the liver, spleen, and lungs, where
they had been phagocytosed by endothelial and connective-tissue cells. In the
omentum, mesentery, bone-marrow, and diaphragm, similar but less numerous
particles were encountered ; in the placenta and fetal membranes, on the other hand,
there was not even a trace of carbon.

In order to determine whether the cells of the vitelline membrane refuse under
all circumstances to take up carbon particles, a small quantity of ink was injected
directly into the uterine cavity in the region of the villi. Microscopic examination
of the villi after this experiment showed that the cells covering them did not take
up a single particle of carbon into their cytoplasm, although they were seen actively
engaged in the absorption of hemoglobin. It appears, therefore, that the endoder-
mal cells of the vitelline membrane and the chorionic epithelium of the placental
labyrinth are unable to phagocytize particles as coarse as those of india ink, although
they are capable of absorbing a substance as finely dispersed as trypan blue.

96 EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

SUMMARY.

1. Trypan blue, injected into the pregnant guinea-pig, stains the placenta and
the vitelline membrane but does not enter the amniotic fluid or the fetus, except
in traces.

2. A zone of the vitelline membrane covered by villi absorbs the dye very
rapidly and in large amount, and may well be termed an attraction center for the
dye. The portion of the vitelline membrane covering the placenta, theectopla-
cental endoderm, fails to stain vitally.

3. Trypan blue is absorbed and stored to a slight extent by the ectoderm cover-
ing the chorionic villi of the placental labyrinth. It is not taken up at all by the
interlobular syncytium or by that part of the ectoderm, composed of giant cells and
syncytium, which occupies the central zone (Umlagerungszone) .

4. The dye appears in the form of fine granules in the layer of giant cells
described by Duval, situated beneath the ectoplacental endoderm.

5. It is also taken up by a host of free macrophages which are found throughout
the decidua serotina and the uterine wall.

6. Carbon particles injected into the maternal blood-stream are not phago-
cytosed by any cells of the placenta or fetal membranes.

EXPERIMENTAL STUDIES ON FETAL ABSORPTION. 97

IV. BEHAVIOR OF THE PLACENTA AND FETAL MEMBRANES OF THE RABBIT TOWARD
TRYPAN BLUE INJECTED INTO THE MATERNAL BLOOD-STREAM.

Successive doses of trypan blue were administered intravenously to six pregnant
rabbits, after which the animals were killed. Gestation was more than half
completed at the time the last injection was done, and therefore no observations
were made during the early stages of development.

On opening the abdomen the usual distribution of color was observed. The
conspicuous uterus, which ordinarily contained from 6 to 12 fetuses, was deeply
stained, both on surface view and on section. The placentae were all dark blue.
The vitelline, or outermost fetal membrane, was dark blue, the color being deepest
in an equatorial zone between the mesometral and antimesometral poles of the
fetal sac. The stain faded gradually toward the antimesometral pole, while in the
opposite direction it ceased abruptly in the region of the terminal sinus. A band
of unstained membrane — the chorion Iseve — consequently remained between the
terminal sinus and the border of the placenta.

The vitelline membrane, loosely attached to the amniotic membrane by strands
of mesoderm, was easily stripped, revealing the delicate amnion inclosing the amni-
otic fluid. The amnion itself was faintly stained and the amniotic fluid contained
traces of the dye.

When large and repeated doses of trypan blue are injected into the maternal
blood-stream, the amniotic fluid may become quite deeply stained. In such cases
the fetus also shows a light staining. In the less deeply stained animals, in which
the amniotic fluid contains only a trace of dye, a hardly appreciable staining of the
fetus may be observed. Such a staining as this might easily escape detection. It
is the writer's belief that the fetus probably contains at all times as much trypan blue
as the ajnniotic fluid, but that the dye, when present only in traces, is less readily
visible in the fetal tissues than in the amniotic fluid. The escape of traces of dye
from the maternal into the embryonic circulation and the amniotic fluid appears to
impair in no way the vitality of the fetus.

The observation that soluble dye-stuffs of various kinds, injected into the
maternal blood-stream, can be detected frequently in traces in the amniotic fluid
but less often in the fetal tissues, has led a number of investigators to conclude that
fluid diffuses directly from the maternal blood-stream into the amniotic fluid, and
that it is not necessary for the substances to enter first the fetal circulation; nor are
they supposed to reach the amniotic fluid through the placenta, but are thought to
enter it directly through the fetal membranes. One theory uses these observations
to prove that the amniotic fluid is in whole or in part a transudate from the maternal
vessels.

Leaving the human being out of consideration, the question of the possibility
in animals (from observations on which the claim is based) of the formation of the
amniotic fluid in this way meets with several grave objections. None of the
observers working on rodents have taken into consideration the fact that the
amniotic sac is completely surrounded by the vitelline membrane. This membrane
possesses an unbroken epithelial surface whose cells are concerned chiefly in the

98 EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

transfer of products from the embryotrophe to the fetal circulation. Furthermore,
it is richly vascularized and consequently the amniotic fluid is completely surrounded
by fetal blood-vessels which would make it difficult to conceive of dye-stuffs diffus-
ing through the membrane without first entering the fetal circulation. Moreover,
from the present observations and those on the guinea-pig recorded in the study just
preceding, it appears probable that dyes such as trypan blue, when administered
in large doses, stain the fetus as well as the amniotic fluid. The portal of entry of
these traces of dye could be either the placenta or the vitelline membrane, or both,
but it appears likely that, whichever route the dye pursues, it first enters the fetal
blood-stream.

In the human being it is theoretically conceivable that fluid may diffuse
directly from the maternal blood-stream into the amniotic fluid. The wall of the
amniotic cavity contains no fetal vessels and the chorion is intimately fused with
the uterine mucosa in which prominent maternal vessels are visible. If fluid
escapes from these vessels it would seem quite possible that some of it might diffuse
through the amniotic membrane directly into the amniotic sac.

It is of interest to consider the differences which exist in various animals in
regard to the staining of the amniotic fluid after injection of a vital dye into the
maternal blood-stream. Goldmann (1909) first noted the staining of the amniotic
fluid in the rat and mouse after the administration of vital dyes. The author has
observed the passage of traces of trypan blue into the amniotic fluid of the guinea-
pig and rabbit. In similar experiments on cats (Wislocki, 1920), however, the
amniotic fluid, as well as the allantoic fluid, remained unstained. The conclusion
to be drawn from these observations is that the fetal membranes and placentae of
rodents are slightly permeable to ultra-microscopic particles, such as trypan blue,
while those of carnivorous animals are impermeable.

Homer (1905) made parallel observations upon the permeability of the mem
branes and placenta in different classes of animals. He injected tetanus antitoxin
into pregnant sheep, cows, rabbits, guinea-pigs, and human beings. He observed
that the antitoxin passed readily from mother to fetus in the human, that it was
transmitted only occasionally in the guinea-pig and rabbit, and that transmission
never occurred in sheep or cows. What is the explanation of this striking difference
in behavior? Romer concluded from his experiments that the more heterogeneous
the substance the more readily is it transmitted by the placenta. Thus tetanus
antitoxin, which was derived from the horse, was extremely heterogeneous for man
but only slightly so for the cow and sheep. Another explanation suggests itself,
both for Romer's results and for the passage of traces of trypan blue through some
placentse and not through others. We must recall that Grosser (1909) has classified
placenta3 according to the degree of union which exists between the fetal and maternal
tissues. Thus, in the pig, cow, sheep, etc., the simplest types, there occurs merely
an apposition of the chorionic ectoderm to the unbroken surface of the uterus. In
carnivora, such as the cat and dog, which represent the next highest type, the
chorionic ectoderm has invaded the uterine mucosa so that the placenta is made up
of maternal blood-vessels intimately surrounded by fetal tissue. In rodents and

EXPERIMENTAL STUDIES ON FETAL ABSORPTION. 99

man, the highest types, a still more intimate union of fetal and maternal tissues
occurs. Here the maternal blood eventually flows in channels completely lined by
fetal ectoderm, and hence only a layer of syncytium and delicate connective tissue
separates the maternal blood from the fetal capillaries. It is apparent that in these
animals (rabbit, guinea-pig, mouse, rat, and man) in which tetanus antitoxin or
traces of trypan blue have been found in the fetus or the amniotic fluid, an extremely
thin barrier of cells separates the maternal from the fetal blood-stream. In the
cat, sheep, and cow, on the other hand, where the union is less intimate and numerous
cells intervene between the two circulations, the transmission of antitoxin and trypan
blue has not been observed.

The vitelline membrane of the rabbit is composed of a single layer of columnar
cells, resting on a basement membrane beneath which are numerous blood-vessels.
The surface of the membrane at its antimesometral pole is smooth, while in the
neighborhood of the terminal sinus, where it ends in a ragged edge, it is covered by
villi which project into the semi-fluid embryotrophe imprisoned between them and
the uterine wall. In the injected animal the vitelline membrane is deep blue on
gross appearance, and microscopically granules of trypan blue are found in large
numbers in the columnar cells composing the membrane (fig. 8). These large
granules, usually 20 or more to a cell, are scattered throughout the cytoplasm. The
single oval or round nucleus contains none of the dye. No granules are visible in the
basement membrane or beneath it in the walls of the fetal vessels. The amniotio
membrane contains no trace of dye.

In the rabbit the labyrinth comprises the bulk of the placenta during the second
half of pregnancy. Septa of fetal connective tissue, containing branches of the
umbilical vessels, divide it roughly into lobules. The umbilical vessels break up
within the lobules into capillaries which are completely lined by endothelial cells
and are accompanied by delicate supporting strands of mesoderm. The space
between the capillaries is occupied by a syncytium composed of fetal ectoderm, the
meshes of which surround innumerable tiny spaces in which the maternal blood
circulates. In consequence of this arrangement the maternal and fetal blood-
streams are separated by only a thin layer of syncytium and the delicate endo-
thelium of the fetal capillaries. In fact, in many places even the syncytium
appears to be absent, so that the maternal blood-cells come in direct contact with
the fetal endothelium, an arrangement no doubt greatly facilitating the interchange
of substances.

Trypan blue was found abundantly present in the placental labyrinth as
aggregations of tiny granules throughout the syncytium (fig. 9). None of it was
seen in the cells forming the fetal capillaries. Beneath the labyrinth, and uniting
it to the uterine wall, lies a layer of giant cells. These cells, which early in gestation
play a prominent part in the growth of the placenta and the nutrition of the fetus,
are reduced during the latter half of gestation to a relatively inconspicuous layer.
In these vitally stained rabbits the cytoplasm of these cells became filled with
numerous tiny blue granules, showing that the cells, even with approaching
maturity, possess the power to assimilate materials brought to them.

100 EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

Beneath the layer of giant cells lies the outermost or decidual layer of the
placenta, composed of decidual cells, many of which were found to be undergoing
degeneration. This layer is traversed by conspicuous maternal vessels which pass
to and from the labyrinth. Try pan blue was present only in the cells which were
degenerating and whose cytoplasm and nuclei, as a result, stained diffusely. The
uterine musculature was noteworthy because of the number of vitally stained
macrophages visible in its connective-tissue septa.

SUMMARY.

1. Trypan blue injected into the maternal blood-stream of the pregnant rabbit
stains the placenta and vitelline membrane.

2. The dye passes in traces from the maternal into the fetal circulation, staining
the fetus and amniotic fluid very slightly.

3. The dye is stored in the form of granules in the cells of the vitelline membrane
and in the syncytium and giant cells of fetal origin in the placenta.

CONCLUSIONS.

Our knowledge concerning the behavior of the placenta and fetal membranes
toward colloidal dyes injected into the blood-stream may be summed up as follows:

Finely dispersed colloids, such as trypan blue and pyrrhol blue, when injected
intravenously into the pregnant animal, reach the placenta and are there absorbed
and stored in the form of granules in the chorionic ectoderm. In the mouse and
rat the dye passes in traces into the amniotic fluid but fails to stain the fetus
(Goldmann). In the guinea-pig and rabbit it likewise passes in traces to the amni-
otic fluid, but in addition it faintly stains the fetus. In the cat vital dyes are not
transmitted, even in traces, to the amniotic fluid or to the fetus (Wislocki, 1920).
This variation in behavior may be explained on comparative anatomical grounds,
differences of architecture making the placenta of carnivora less permeable than
that of rodentia. Vital dyes are also absorbed and concentrated into granules in
the cytoplasm of the cells of the vitelline membrane, which in rodentia forms the
outermost fetal covering. Granules of dye are deposited in a structure peculiar to
carnivora, termed in the cat the "brown border," which is a modified portion of the
chorionic membrane bordering the placenta.

EXPERIMENTAL STUDIES ON FETAL ABSORPTION.

101

REFERENCES CITED.

DOWNEY, H., 1917. Reactions of blood and tissue cells to
acid colloidal dyes under experimental con-
ditions. Anat. Record, vol. 12, p. 429.

, 1918. Further studies on the reactions of blood and

tissue cells to acid colloidal dyes. Anat.
Record, vol. 15, p. 103.

GOLDMANN, E. E., 1909. Die aussere und innere Sekretion
des gesunden und kranken Organismus im
Lichte der "vitalen Fiirbung." Beitr. z. klin.
Chir., vol. 64, p. 192.

GROSSER, O., 1909. Vergleichende Anatomie und Entwick-
lungsgeschichte der Eihaute und der Placenta.
Vienna.

ROMER, A., 1904. Weitere Studien zur Frage der Intrau-
terinen und extrauterinen Antitoxintibertrag-
ung von der Mutter auf ihre Nachkommen.
Beitr. z. exper. Therap., Heft 9.

WISLOCKI, G. B., 1920. Experimental studies on fetal
absorption. I. The vitally stained fetus.
II. The behavior of the fetal membranes and
placenta of the cat toward colloidal dyes in-
jected into the maternal blood-stream. Con-
tributions to Embryology, vol. 11, Carnegie
Inst. Wash. Pub. 274.

DESCRIPTION OF PLATE.

ch. ep., Chorionic epithelium.

d. «., Decidua serotina.

ec. endo., Ectoplacental endoderm.

endo. v., Endodermal villi.

g. c., Giant cells.

p. I., Placental labyrinth.

r. c. e.. Roof of the central excavation.

s. endo.. Sinus endodermaticus.

». m., Vitelline membrane.

Fio. 1. Gross specimen showing appearance of the vitelline membrane and placenta of a fetal guinea-pig, measuring
18 mm. after injection of trypan blue into the maternal blood-stream. The deeply stained villous
portion of the vitelline membrane is readily distinguishable, as is also the unstained roof of the central
excavation.

FIG. 2. Section through vitelline membrane of guinea-pig at its attachment to the surface of placenta. The char-
acter of the epithelium covering the various portions of the vitelline membrane is clearly shown. The
endodermal cells clothing the villi are loaded with trypan-blue granules. The curious papillary hyper-
plasia of the ectoplacental endoderm is well illustrated.

Fia. 3. Section through placental labyrinth of guinea-pig, showing tracery of trypan blue in the chorionic epithelium
after repeated administration of the dye.

Fio. 4. Specimen showing distribution of trypan blue in the placenta of guinea-pig after injection of the dye into
maternal blood-stream. Note that the only unstained area is the roof of the central excavation.

FIG. 5. Specimen showing the distribution of trypan blue in the placenta of guinea-pig, after injection of the dye
into the fetal circulation. Note that when the dye reaches the placenta through the fetal vessels, the
roof of the central excavation becomes stained, but none of the dye diffuses into the decidua.

Fia. 6. Section through the surface of the placenta of guinea-pig (about the middle of pregnancy) showing the vitally
stained giant cells beneath the ectoplacental endoderm.

Fia. 7. Section of the placental labyrinth of guinea-pig showing polymorphonuclear leucocytes containing granules of
trypan blue.

Fio. 8. Section through the vitelline membrane of the rabbit, showipg the columnar cells covering the villi filled with
granules of trypan blue.

Fia. 9. Section from labyrinth of rabbit's placenta, showing syncytium in which numerous tiny granules of trypan blue
are visible.

W1SLOCKI

PLATE 1

L-A

-/'

y*—

m
&

,„.. _

,s, V It!

.,

'.

•-. •

v^..: . -; • •, - \

H^

1 V

,:

: .,

.'
"

-

*'-*-'"'' w*

* .

..

.-\ -

~.

-

«»

"> * ^~

.v*^ *a

P 0

a

©

• » «
i

-, \^..^

%

fA

I

E. Piotti feel!.

CONTRIBUTIONS TO EMBRYOLOGY, No. 63.

THE DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA,

BY G. B. WISLOCKI AND J. A. KEY.

Of the Anatomical Laboratory, Johns Hopkins University, and the Laboratory of
Surgical Research, Harvard Medical School.

With one plate.

103

•*.*•*

lu-i L I S R A

THE DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA, T*"

Since mitochondria were discovered by Altmann in 1890, a voluminous litera-
ture has accumulated on this subject. Mitochondria have been described in almost
every type of animal and plant cell and have been credited with functions varying
from the transmission of hereditary characters to differentiation into neurofibrils,
muscle fibrils, connective-tissue fibrils, and zymogen granules. Other observers
believe that they are concerned in the metabolic processes of the cell. Recently
the literature has been ably reviewed by Duesberg (1912) and Cowdry (1918),
and therefore no attempt will be made to do so here.

The only reference to mitochondria in the placenta that we find in this exten-
sive literature is a note by De Kervily (1916). This writer states that mitochondria
are abundantly present in the Langhans cells of the human placenta throughout
gestation. He makes no mention of their occurrence in the syncytium or other
placental elements. It was consequently deemed of interest to investigate the
mitochondria of the placenta more in detail and in different types of animals. We
were guided in our choice by the classification of Grosser (1910), who arranged the
placentae of animals in an ascending scale dependent upon the union of the chorion
with the uterine wall. Following this classification, we have chosen the pig, cat,
guinea-pig, and human, as representing several rather distinct types.

According to Grosser, the pig represents the simplest type, in which the chorion
is merely apposed to the folds of the uterine mucosa, in consequence of which the
maternal and fetal blood-vessels remain widely separated. In the cat (Carnivora)
the union is more intimate as the result of an invasion of the uterine mucosa and its
stroma by the chorion, whereby the fetal ectoderm eventually surrounds the mater-
nal blood-vessels, and the distance between the maternal and fetal circulation
is greatly reduced. In the guinea-pig (Rodentia), a still more intimate fusion of
the chorion with the uterine wall occurs. Here the chorion finally erodes the
maternal vessels so that their endothelium disappears and the maternal blood flows
through channels completely lined by fetal ectoderm. The barrier between the
maternal and fetal blood-streams has been reduced to a single layer of chorionic
epithelium and a delicate stroma. In the human, conditions are similar to those in
the guinea-pig, although there are marked differences in the finer architecture.

MATERIAL AND METHODS.

A series of mature placentae from the pig, cat, guinea-pig, and human were
obtained and fixed fresh. Placentae of earlier stages of development in the cat and
the guinea-pig were also obtained in order to check up the various types of cells
found in these animals.

Small pieces of the material were fixed either in the acetic-osmic-bichromate
mixture, formalin-Zenker mixture of Bensley, or the various formalin-bichromate
mixtures of Regaud. The sections were stained by the fuchsin-methyl green

105

106 DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

method of Bensley, the copper-chrome-hematoxylin method of Bensley, the
fuchsin-picric acid method of Altmann, the alizarin-anilin-gentian violet method of
Benda, and the iron-hematoxylin method of Regaud. The best results were obtained
by fixing according to method IV B of Regaud and staining by the fuchsin-methyl
green method of Bensley. The great variation in the number and size of mito-
chondria renders differentiation by the iron hematoxylin method very difficult, as
the minute mitochondria in some cells are completely decolorized while adjacent
cells containing an abundance of large mitochondria remain deeply stained.

In Regaud's method IV B the fixative is made up of neutral formalin 1 part, 3
per cent aqueous solution of potassium bichromate 4 parts. The formalin is neutral-
ized with magnesium carbonate. The tissues are fixed 4 days, the solution being
changed daily; they are then mordanted for 8 days in a 3 per cent aqueous solution
of potassium bichromate, washed 24 hours in running water, dehydrated in 50
per cent, 70 per cent, 95 per cent, and absolute alcohol, cleared in xylol and
imbedded in paraffine.

Sections were cut 2 to 4 n in thickness and fixed to the slide by the water-
albumin method. The sections were freed from paraffine by xylol and passed
through graded alcohols to water, placed for half a minute in a 1 per cent aqueous
solution of potassium permanganate, then for half a minute in a 5 per cent aqueous
solution of oxalic acid. They were then washed thoroughly in water and stained by
Bensley 's acid fuchsin-methyl green method, which consists of the use of two stain-
ing solutions: i. e., (1) acid fuchsin, 20 gms.; anilin water, 100 c. c.; (2) 1 per cent
aqueous solution of methyl green. In the use of this method the following procedure
is to be carried out : Pour the acid fuchsin solution (a comparatively fresh solution
must be used) on the slide and heat the slide over a flame until the solution begins to
steam. Allow the stain to act about 6 to 8 minutes, heating 3 or 4 times to the
steaming-point. Pour off the stain, blot, and then wash the slide thoroughly in dis-
tilled water. Blot and pour 1 per cent methyl green on the slide. This must be
given time to stain the nuclei and protoplasm of the cells and is usually allowed to
act about half a minute. Then blot and dehydrate in absolute alcohol, clear in
xylol, and mount in balsam. If the sections are stained too deeply with methyl
green the minute mitochondria are obscured. If such sections are dipped for an
instant in 95 per cent alcohol part of the green is taken out and good differentiation
can be obtained.

With this method the mitochondria are stained a bright red by the fuchsin and
the protoplasm is light green or unstained. The chromatin of the nucleus stains
green and the nucleoli either red or green. Unless otherwise stated, the following
descriptions are based upon sections prepared by the above technique.

PLACENTA OF THE PIG.

In the pig the chorion is covered by a single layer of cuboidal or columnar cells
which is applied directly to the uterine mucosa. A thin layer of structureless
substance, the so-called embryotrophe, which stains light green, lies between the two
tissues. The chorionic epithelium and uterine mucosa are thrown into inter-

DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA. 107

locking folds. The cells of the chorion vary in height in different parts of each fold;
at the apices of the chorionic folds they are of low columnar form, on the sides of
the folds they are cuboidal, while in the depths of the folds a transition again occurs
to columnar cells 20 to 30 M in height. Situated in the basal half of each cell is a
single large, oval nucleus which takes the green faintly, has a well-defined chromatin
network, and one or two large fuchsinophilic nucleoli. The protoplasm of the basal
half of the cell also stains a faint green. In the apical zone the protoplasm is clear
and unstained. In many cells it is extensively vacuolated and degenerated cells
with pycnotic nuclei and diffusely fuchsinophilic protoplasm are common. In a
number of the chorionic epithelial cells a fuchsinophilic, colloid-like substance is
found in the basal zone (usually between the nucleus and the base of the cell),
varying in amount from one or two small granules to large masses 5 to 6 ^ in
diameter. The larger masses are surrounded by a clear zone as though they had
contracted during fixation. The cdls containing this colloid substance present
no evidence of degeneration.

The mitochondria are concentrated in the apical half of the cells facing the
uterine mucosa (fig. 1). They are quite abundant in the apical zone and appear
as minute granules and very delicate, slightly curved rods. The rods and rows of
granules are parallel to the long axis of the cell and appear as though streaming
from its apex towards the nucleus. They steadily decrease in number as the
distance from the apex increases, and it is rare to find any in the region of the
nucleus and none is seen in the zone between the nucleus and the base of the cell.
The cuboidal and low columnar cells of the sides and apices of the folds are similar
to the high columnar cells in the depths of the folds, except that the various regions
of the cell are less well marked and fewer cells contain colloid droplets.

The chorionic epithelium rests on a very thin basement membrane which is
supported by a delicate connective-tissue layer containing the smaller fetal blood-
vessels. This layer is composed of small spindle or stellate cells which do not
differ from fibroblasts found in other tissues. They contain a few minute granular
and rod-shaped mitochondria. The endothelium of the fetal vessels also contains a
few granular and rod-like mitochondria and herein does not differ from ordinary
vascular endothelium. This delicate connective-tissue layer merges into a thin
layer of coarser fibrous tissue which contains the larger fetal blood-vessels, the cells
of which are larger than those of the preceding one and contain larger and more
abundant mitochondria. This fibrous layer is covered by the allantoic mem-
brane, which is composed of a single layer of flattened epithelial cells resting on a
delicate basement membrane. The allantoic epithelium is quite rich in minute
granular mitochondria which are scattered uniformly through the protoplasm.

The uterine mucosa facing the chorionic epithelium is composed of a single
layer of low cuboidal cells resting on a basement membrane. Each cell contains a
single relatively large, round or oval nucleus which stains deep green and contains
one or two large fuchsinophilic nucleoli. The protoplasm stains lightly with green
and contains an abundance of rather coarse mitochondria (fig. 1). These are in
the form of coarse granules and curved rods and are distributed quite unfiormly

108 DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

through the protoplasm. Just beneath the basement membrane is a rich capillary
network which rests on the connective-tissue stroma containing the maternal blood-
vessels and uterine glands. The mitochondria of the connective-tissue cells and of
the endothelium are similar to those of the corresponding cells in the chorion.

The uterine glands are lined by a single layer of medium-sized columnar epithe-
lial cells 15 to 20 IJL in height. Each cell contains a single round nucleus, located
near the base of the cell, which stains light green and has a well-defined chromatin
network and one to four fuchsinophilic nucleoli. The cell protoplasm stains a
faint green and contains a rather large number of mitochondria, most of which are
in the form of heavy, slightly curved rods (fig. 2). A few medium-sized granular
mitochondria are found in either extremity of the cell. The rods are all arranged in
the long axis of the cell and often a single rod extends the entire distance from the
base to the apex of the cell. Many of these cells contain coarse globules of a substance
similar in staining to the mitochondria. These are larger in diameter than the
mitochondria and fairly uniform in size. In senile cells with pycnotic nuclei they
may entirely replace the mitochondria. They are not uncommon, however, in the
apical zones of apparently normal cells. In cells in a later stage of degeneration
the globules are replaced by vacuoles.

The muscle fibers of the pregnant uterus contain numerous minute mitochon-
dria in the form of short rods and granules. They are very numerous in a narrow
zone around the nucleus and fairly numerous in the peripheral zone of the muscle
fiber just beneath the fiber sheath. In lesser numbers they are scattered through
the substance of the fibers.

PLACENTA OF THE CAT.

In the cat the placental labyrinth consists of alternating columns of fetal
ectoderm and loose connective tissue. The maternal blood-vessels occupy the
center of the ectodermal columns, while the fetal vessels are located in the connective
tissue separating them. The layer of ectoderm surrounding the maternal blood-
vessels is composed of giant cells and numerous smaller cells. The cytoplasm of the
latter coalesces gradually, as gestation advances, to form a syncytium which
incloses the maternal blood-vessels and the giant cells.

The cells lining the maternal vessels differ from ordinary endothelium in that
their protoplasm is slightly more abundant and much richer in mitochondria
(fig. 3). The mitochondria are so numerous that, unless the preparation is very
thoroughly differentiated, the protoplasm of these cells stains a uniform deep red.
In a well-differentiated section, forms varying from fine to coarse granules and curved
rods are seen closely packed in the faint green protoplasm. The nuclei stain light
green and have a green-staining nucleolus as in ordinary endothelium. In the
earlier stages of the placenta these cells possess more abundant protoplasm rich in
small mitochondria.

The giant cells, isolated or in small groups, alternate with the maternal vessels
to form the center of the column. Occasionally, giant cells lie alongside of a maternal
vessel between it and the border layer of tetal ectoderm. These are round to oval

DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA. 109

in shape, vary from 15 to 40 // in diameter, and contain from one to three nuclei.
The average giant cell is about 20 M in diameter and contains only one nucleus,
which is large, round, and stains light green. Near its center is a single large
fuchsinophilic nucleolus from which delicate strands of chromatin radiate to the
nuclear membrane. The protoplasm stains faint green and may contain numerous
minute vacuoles, especially in the older cells. The mitochondria are fairly abund-
ant in the form of minute granules and rods arranged in an eccentric zone around
the nucleus. The peripheral zone of the proptoplasm is entirely free from mito-
chondria or other granules. In some cells an eccentric ring of clear protoplasm is
present between the nucleus and the zone of mitochondria. In cells containing
two nuclei the zone of mitochondria may surround both and not invade the proto-
plasm lying between them; or, there may be a separate mitochondrial zone around
each nucleus. In the giant cells of younger placentae the mitochondria are slightly
larger and more numerous in proportion to the amount of protoplasm and are not
so definitely concentrated in the zone around the nucleus.

The small ectodermal cells form a layer which surrounds the giant cells and
maternal vessels and separates them from the connective-tissue stroma containing
the fetal vessels. At first these cells have quite distinct boundaries, but during
the latter part of gestation then1 cytoplasm coalesces to form a syncytium which
closely invests the giant cells and the maternal vessels. The smaller ectodermal
cells are low columnar, about 10 ^ in height, and are quite uniform in size. In the
central portion of each cell is a single round nucleus which stains deep green and
contains a single deep green nucleolus and several heavy masses of chromatin. The
protoplasm contains a large number of vacuoles of various sizes, more abundant at
the pole of the cells nearest the maternal vessels. In osmic preparations these
vacuoles are seen to be filled with fat. The mitochondria, in the form of small
granules and short rods, are fairly numerous and lie in the protoplasm between
the fat vacuoles (fig. 3). They are not concentrated in any particular part of
the cell. In the younger placentae transition can be traced between the small ecto-
dermal cells and the giant cells.

Lying between and over the columns of fetal ectoderm is a delicate connective-
tissue stroma which contains the fetal blood-vessels. In the young placenta the
cells of the stroma are quite rich in mitochondria, but in the mature placenta they
possess no more mitochondria than do ordinary connective-tissue cells. The
endothelial cells of the fetal vessels possess a few, but they are in no way remarkable.

The chorionic membrane of the cat consists of a single layer of cuboidal to
columnar cells resting on a basement membrane supported by a very vascular
connective-tissue stroma. On the opposite side this stroma is covered by the
allantoic membrane. Numerous folds project from the chorionic surface toward
the uterine mucosa.

The chorionic epithelium near the placental margin is composed of high
columnar cells, but as the distance from the placenta is increased the cells become
lower and more cuboidal. They are crowded together, their outlines being distorted
by pressure from the surrounding cells. Many cells are apparently in process of

110 DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

being pinched off by pressure from new cells arising next to the basement membrane.
Usually, each cell contains a single nucleus which may be located in any part of the
apex, depending upon the pressure from surrounding cells. However, cells contain-
ing two or three nuclei are not rare. The nucleus has a very indefinite limiting
membrane and its outline is irregularly stellate, apparently from the pressure of
vacuoles in the protoplasm which indent its surface and may almost completely
divide it. It stains rather deep green and contains a single large fuchsinophilic
nucleolus. The cytoplasm stains a faint green and is crowded with vacuoles of
fairly uniform size, about 1 M in diameter. In the apices of a few of the cells are large
vacuoles containing fragments of phagocytosed red blood-cells. The mitochondria
are not numerous and are seen as small granules and short rods lying between the
vacuoles. They are not concentrated in any particular part of the cell.

The allantoic membrane consists of a single layer of flattened cells resting on a
delicate basement membrane. The free surface of the cells is covered with debris.
The nucleus is oval or flattened, stains a rather deep green and contains a single
fuchsinophilic nucleolus and several irregular masses of chromatin. The pale-green
protoplasm is finely vacuolated and contains a few small granular mitochondria.

PLACENTA OF THE GUINEA-PIG.

The vitelline membrane of the guinea-pig consists of a single layer of columnar
cells of medium height resting on a basement membrane which is supported by a
layer of connective tissue rich in small blood-vessels. The membrane is thrown into
folds and numerous villi project from its surface. The columnar cells have dome-
like apices which bulge out from the points of contact with the surrounding cells
and this, in addition to the variation in height of the cells, gives an irregularly
serrated edge to a cross-section of the membrane. In the central portion of each
cell there is a light-green nucleus which contains from one to four fuchsinophilic
nucleoli. The protoplasm of the apical zone stains a light green, while that of the
basal zone is clear and unstained. Two to six small, clear vacuoles are usually
seen in the apical zone. In the more senile cells the protoplasm contains vacuoles of
varying size. They may be quite large and may occupy any part of the cell.

A variable number of fuchsinophilic granules are constantly present in these
cells. They vary from a few small granules to a large number of coarse globules
2 n in diameter. When only a few are present they are located in the basal zone
and around the nucleus, leaving the basophilic apex clear. They are apparently
not products of degeneration, as they are present in all cells.

The mitochondria are not very numerous. They are mostly in the form of
small rods, though a few minute granules are constantly present. In the average cell
they are limited to the base of the cell and the zone around the nucleus and are
distributed throughout the protoplasm. Apparently there is no quantitative rela-
tion between the mitochondria and the fuchsinophilic granules, as cells containing
a large number of granules may contain either a very few or a relative abundance of
mitochondria. Senile cells contain very few or no mitochondria. Senility, how-
ever, apparently does not affect the number of fuchsinophilic granules, as wide
variations are found in cells with pycnotic nuclei and vacuolated protoplasm.

DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA. Ill

The vitelline membrane covering the placenta likewise presents a single layer
of columnar cells. The cells here are considerably larger than those just described
and in gross appearance resemble those of the chorionic membrane of the cat, in
that they are crowded together with their nuclei at different levels and their cell-
outlines are distorted by pressure from neighboring cells. Also, the older cells are
constantly being constricted at their bases and pinched off by new cells arising
next to the basement membrane.

Each cell possesses a single large oval nucleus which is usually found in the
central zone of the cell unless influenced by pressure, when it may be found any-
where from the base to the apex. It stains deep green and contains one to three
basophilic nucleoli.

The protoplasm stains faintly and in nearly all cells is extensively vacuolated,
the vacuoles being largest and most numerous in the cells which are constricted at
the base. Occasionally large vacuoles in the apical zone contain fragments of
phagocytosed red blood-cells. The mitochondria, present as small granules and
short rods, are fairly numerous, and especially so in the young cells next to the
basement membrane. They are not concentrated in any particular part of the cells
but are distributed uniformly through the protoplasm, the rods usually lying with
their long axes parallel to the long axis of the cell. In the senile cells with irregular
pycnotic nuclei the mitochondria are less numerous and many bulb-like swellings
and mitochondrial vesicles are present. However, the mitochondria persist as long
as the cell remains attached to the basement membrane.

Just beneath the thin basement membrane supporting these cells is a layer, one
to four cells in thickness, of large round or oval giant cells. In the mature placenta
most of these cells are senile and have vacuolated protoplasm and pycnotic nuclei,
although a few can be found which appear to be quite healthy and normal. As a
rule they are mononuclear, though cells containing two to four nuclei are occasionally
seen. The nuclei are large, round to oval in shape, stain slightly with green, and
contain a single large, deep-green nucleolus from which strands of chromatin
radiate to the nuclear membrane. The protoplasm stains a faint green and is
quite rich in minute granular and rod-like mitochondria. These may be con-
centrated in a zone around the nucleus or distributed uniformly through the proto-
plasm. In the degenerating cells few or no mitochondria can be seen. No other
granules are visible in these cells.

Beneath the layer of giant cells is a well-defined layer of connective tissue
which separates it from the cortical syncytium covering the labyrinth. From
the inner surface of the cortical syncytium heavy bands of interlobular syncytium
project into the labyrinth and divide it up into lobules. The structure of the cortical
and interlobular syncytium is the same and they will be described together.

The syncytium is a coarse reticulum of heavy protoplasmic bands in which no
cell boundaries can be distinguished. The interspaces of the reticulum are the
sinusoids in which the maternal blood circulates and, as they are not lined by endo-
thelium, the blood comes in direct contact with the syncytium. The small round
to oval nuclei are very numerous and are distributed quite uniformly through the

112 DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

protoplasm. They stain a deep green and contain a single large basophilic
nucleolus. The protoplasm stains a bright green and is extremely rich in minute
granular and short, rod-like mitochondria (fig. 4). These are quite uniform in size,
the rods rarely being over 2 pin length, and are distributed uniformly through the
protoplasm. No other granules are demonstrable. In some areas the protoplasm
is finely vacuolated and here the mitochondria are somewhat less numerous.

The labyrinth consists of a dense network of maternal blood-spaces, inclosed
in a thinned-out syncytium, and fetal blood-vessels supported by a small amount
of connective tissue. The syncytium resembles endothelium, but retains the rich-
ness in mitochondria (fig. 5) characteristic of the border and interlobular syncytium,
and also exhibits a similar staining reaction.

The fetal blood is inclosed in delicate capillaries whose endothelium is similar
to that found in other localities. Its nuclei and protoplasm stain a lighter green
than do the nuclei and protoplasm of the syncytium and the protoplasm is .about
one-fourth as rich in mitochondria as is that of the syncytium. The mitochondria
are much more variable in size and many curved rods, 6 to 8 /x long, are seen.

A section through the roof of the central excavation shows irregular columns
of degenerating decidual tissue separated by papillary processes of fetal ectoderm,
one or two cells in thickness. The fetal cells are small, cuboidal in shape, and
contain relatively large round or oval nuclei which stain deep green and contain
basophilic nucleoli. The small amount of green-staining protoplasm is quite rich
in small granular and rod-like mitochondria. The cells of the decidua in this
locality have highly vacuolated protoplasm and pycnotic nuclei. The mitochondria
are limited to a few minute granules lying between the vacuoles in the protoplasm.

HUMAN PLACENTA.

The labyrinth of the mature human placenta is made up of chorionic villi sur-
rounded by maternal blood-spaces. The fetal blood-vessels occupy the center of
each villus and are supported by a rather coarse connective-tissue network. The
villi are covered by a syncytium which everywhere separates the fetal vessels from
the maternal blood. Just beneath this syncytium are varying numbers of clear
Langhans cells. The syncytium is a layer of protoplasm of variable thickness in
which no cell boundaries can be made out. The numerous nuclei are usually
arranged in a single layer and are quite close together. However, in places where
the syncytium is thick, as in the angle or at the junction of two large villi, there may
be 4 or 5 layers of syncytial nuclei. They are small, round or oval in shape, stain
deep green, and contain a medium-sized basophilic nucleolus.

The protoplasm of the syncytium takes the green faintly and is usually homo-
geneous, though in some areas it is finely vacuolated. It is very rich in minute
granular and short rod-like mitochondria (fig. 6). These are distributed uniformly
through the protoplasm and are so numerous and crowded that it is difficult to
distinguish the individual granules except in very thin, well-differentiated sections.
They are somewhat less numerous in the vacuolated areas.

DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA. 113

The cells representing the remains of the Langhans layer lie just beneath the
syncytium. They are usually seen in groups of 2 to 4 cells, though many isolated
cells are present and occasional groups of 10 to 15 cells can be found. Many of
these cells are senile and have vacuolated protoplasm and pycnotic nuclei. Others
appear to be quite healthy and normal. They range from 10 to 20 /u in greatest
diameter and vary in shape from isolated oval or spindle cells to rows of low
cuboidal cells. The nucleus is rather large, stains light green, and contains 1 to 4
medium-sized basophilic nucleoli and a delicate chromatin network.

The protoplasm is clear and unstained and contains a moderate number of
mitochondria. These are usually small, irregularly curved rods of moderate length,
and many of them have small nodules at their extremities. A few granular forms
are present. The mitochondria are sharply differentiated in the clear unstained
protoplasm and this enables one to recognize these cells, regardless of their form or
position. As the signs of degeneration increase the mitochondria decrease in number
and size. No vesicles or other granules are seen in these cells. The Langhans cells
are easily found in all mature human placentae.

The decidual cells are found in an irregular layer next to the uterine wall as
short columns projecting into the labyrinth. They are large round to oval cells 10
to 30 M m diameter, and resemble somewhat the giant cells found in the placentae
of the pig and the cat, although they are entirely different in origin. They contain
one or two large round or1 oval nuclei which stain a faint green and contain in turn
one or two large basophilic nucleoli and a delicate chromatin network.

The protoplasm stains more intensely than the nucleus and is quite rich in
minute mitochondria. They are in the form of short rods and granules and are
distributed uniformly through the protoplasm. The protoplasm of many of these
cells contains a colloid-like substance which stains deeply with fuchsin. It varies
in amount from a few small droplets to large irregular masses 4 to 5 M in diameter.
An apparently normal cell may contain a large amount of this substance and the
surrounding cells be entirely free from it. It is not present in the degenerating
cells. Decidual cells are present in all stages of degeneration.

As senility advances, the nucleus becomes homogeneous and stains deep green,
the protoplasm loses its affinity for the green, and the mitochondria decrease in
number and size and finally disappear entirely.

CONCLUSION.

From the foregoing description, it will be seen that mitochondria are abundantly
present in the various types of placentae investigated. They are particularly numer-
ous in those cells which form the barriers between the maternal and fetal circulation
and through which the respiratory and nutrient interchange between the two
organisms is constantly proceeding.

To review briefly, in the pig, mitochondria are very plentiful in the layers of
epithelial cells, namely the chorionic ectoderm and the uterine mucosa which form
the bridge between the maternal and fetal blood-streams. They are also found in
great abundance in the epithelium of the uterine glands and this is significant,

114 DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

because an important role is ascribed to these structures. The uterine glands of
the pig's placenta are credited, according to several investigators (Ercolani, 1869;
Bonnet, 1882) with liberating a profuse secretion, termed the uterine milk, which
finds its way onto the surface of the uterine mucosa, where it is absorbed by the cho-
rionic epithelium and thus constitutes one of the main sources of fetal nutriment.

Mitochondria are equally numerous in the placenta of the cat. They occur
chiefly in the broad layer of chorionic epithelium which intervenes between the
circulation of the mother and that of the offspring. They are also abundant in
the delicate layer of endothelium which incloses the maternal blood-channels.
Little has been reported concerning the histochemistry of the cat's placenta
beyond the fact that the chorionic epithelium contains numerous fat droplets
(Melissenos, 1906). Recently, one of the authors (Wislocki, 1920) has studied the
behavior of the cat's placenta toward trypan blue injected into the maternal blood-
stream. It was found that trypan blue does not pass from mother to fetus but is
stored in the form of minute granules in the cytoplasm of the chorionic epithelium
and the endothelial cells which line the maternal blood-vessels. These are the cells
whose cytoplasm, as described, is filled with mitochondria.

In the guinea-pig mitochondria are found in great profusion in the syncytium
which comprises the bulk of the placental labyrinth and separates the maternal from
the fetal blood-stream. Fat and glycogen are also readily demonstrable in the
syncytium and trypan blue is deposited there when injected into the maternal cir-
culation. The placenta of the guinea-pig possesses numerous degenerating cells of
both fetal and maternal origin and it is interesting to note that these cells lose their
mitochondria while undergoing degeneration.

In the human placenta, also, it is the syncytium and Langhans cells which con-
tain mitochondria in greatest abundance. It need only be emphasized that these are
the cells which separate the two circulations and in which products of metabolism,
such as glycogen (Driessen, 1907) and fat (Hofbauer, 1910) are amply demonstrable.

In conclusion it is interesting to consider what significance can be ascribed to
the mitochondria in the placenta. It is striking that they are most numerous in
those cells of the placenta which are presumed to play an important role in the
metabolic transfer between the two organisms. From observations such as we have
made, however, it is impossible to assign to them any specific function. The only
conclusion one can safely draw concerning mitochondria is that they represent
material which is probably used in the multiphasic activities of the placental cells.

DISTRIBUTION OF MITOCHONDRIA IN THE PLACENTA.

115

BIBLIOGRAPHY.

BONNET, R., 1882. Die Uterinmilch und ihre Bedeutung fur
die Frucht. Beitrage zur Biologic zu v. Bis-
choffs 50 jahrigem Doktorjubilaum. Stuttgart.

COWDBY, E. V., 1918. The mitochondrial constituents of
protoplasm. Contributions to Embryology,
vol. 8, Carnegie Inst. Wash. Pub. No. 271.

DE KEBVILT, M., 1916. Le chondriome des cellules de
Langhans du placenta humain. Compt. rend.
Soc. Biol., vol. 79, pp. 589-590.

DRIESSEN, L. F., 1907. Ueber Glykogen in der Placenta.
Arch. f. Gyn., vol. 82, p. 278.

DUESBERO, J., 1912. Plastosomen, "Apparato reticolare
interne " und Chromidial-apparat. Ergeb.
der Anat. u. Ent., vol. 20.

EBCOLANI, W., 1869. Sulla placenta e sulle nutrizioni die

fetti nell' utero. Bologna.
GROSSER, O., 1909. Vergleichende Anatomie und Entwick-

lungsgeschichte der Eihaute und der Placenta.

Vienna.
HOFBAUER, J., 1905. Grundziige einer Biologie der mensch-

lichen Plazenta. Vienna und Leipzig.
MELISSENOS, K., 1906. Ueber die Fettkornchen und ihre

Bildung in der Placenta bei den Nagern und der

Katze. Archiv f. mikr. Anat., vol. 67, p. 267.
WISLOCKI, G. B., 1920. Experimental studies on fetal

absorption. Contributions to Embryology,

vol. 11, Carnegie Inst. Wash. Pub. No. 274.

EXPLANATIONS OF FIGURES.

FIG. 1. Section of the placenta of the pig showing the apposition of the chorion to the uterine mucosa. The chorionic

epithelium and uterine mucosa contain numerous mitochondria.
FIG. 2. Section of a uterine gland from the pig's placenta showing mitochondria.
FIG. 3. Section of the placenta from a cat, nearly half term, showing the mitochondria in the endothelium of the

maternal blood-vessels and in the chorionic epithelium.

FIG. 4. Section of the placenta of the guinea-pig, showing abundance of mitochondria in the interlobular syncytium.
FIG. 5. Section of the placenta of the guinea-pig, showing mitochondria in the syncytium of the labyrinth.
FIG. 6. Section of villus from human placenta showing mitochondria in the syncytium.

CH. EP., Chorionic epithelium.

EMB., Embryotrophe.

F. B. V., Fetal blood-vessel.

F. C. T., Fetal connective tissue.

G. C., Giant cell.

GL. EP., Glandular epithelium.

I. SY. Interlobular syncytium.

ABBREVIATIONS.

L. GL.,
M. B. S.,
M. B. V.
M. C. T..
BY.,
U. EP.,

Lumen of gland.
Maternal blood-space.
Maternal blood-vessel.
Maternal connective tissue.
Syncytium.
Uterine epithelium.

WISLOCKI AND KEY

PL ATE I

-.<-.'
' ' * * T ,

'V-- '

•*• - T "•- • \ <v ; .

& i -if:

V 7

*

I* 1 ,* A .£- . -•» -

4

-.- .Tj*fS

^^t^^^

— •» *^V *^' ••' ': ' ''* ,: .'^'.'i^

{•/ ^py ^^15^

I.5Y- ^i|r^% i «^

. 'i^ ^». ,.^

'"' \

--.^t?4^ I"SY.

^aa »./:

3

M B.V.

> A-:

5.C.

£

. -

CONTRIBUTIONS TO EMBRYOLOGY, No. 64.

CYCLIC CHANGES IN THE OVAEIES AND UTERUS OF THE SOW,
AND THEIR RELATION TO THE MECHANISM OF IMPLANTATION,

BY GEORGE W. CORNER,
From the Anatomical Laboratory of the Johns Hopkins Medical School.

With four plates and two figures.

117

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW,
AND THEIR RELATION TO THE MECHANISM OF IMPLANTATION,

1. GENERAL.

INTRODUCTION.

The undertaking set forth in these pages is simple in plan, namely, to follow
in one mammal (the sow) the anatomical changes in uterus and ovary which under-
lie the reproductive cycle. The prospective value of such a study may deserve a
word of explanation; it arises from the fact that in different organisms there are
varying forms of expression of the cyclic sexual processes. Even within the order
of mammals, the reproductive cycle of the female is obscured by as yet unresolved
dissimilarities and no satisfactory correlation has been made between phenomena
so unlike as the urgent rutting of deer, the inconspicuous oestrus of rodents, and the
menstrual cycle of the catarrhine apes and man. In our own species we have only
conflicting conjectures as to the time of ovulation and the meaning of menstruation,
although exact knowledge of this subject would be of value not only for its own sake,
but also for the calculations of the embryologist and the gynecologist. The need,
therefore, is for a detailed study of individual species, including the determination
of the oestrous period and the time relations of ovulation, growth, and retrogression
of the corpus luteum, tissue changes in the uterus, and the progress of the ova,
whether fertilized or unfertilized.

Some such account has been pieced together for five species, though not with
equal completeness: a marsupial, Dasyurus viverrinus (Hill and O'Donoghue, 1914);
the rabbit (Niskoubina, 1909; Ancel and Bouin, 1910; etc.); the guinea-pig (Loeb,
1911, 1914; Stockard and Papanicolaou, 1917; Ishii, 1920); the rat (Long and
Evans, 1920-1921)1; and the dog (Marshall and Jolly, 1906; Marshall and Hainan,
1917; Keller, 1909). Something is known of the cycle of ovulation but not of the
uterine changes in several other mammals, while in man we know much about the
histological cycle of the uterus and almost nothing of the ovarian sequence.

Choice of the domestic pig as subject for a similar study was determined by
a combination of circumstances which seemed to outweigh the disadvantages of
large size and commercial restrictions upon the collection of material. In
this animal oestrus is regularly periodic, frequent, and outspoken; ovulation is
spontaneous; study of the ova and embryos is facilitated by the large litters;
the ovaries of the ungulates are notably uncomplicated, as contrasted, for instance,
with those of the rodents; and the uterine mucosa is of the non-deciduate type,

1 1 am indebted to Professors Long and Evans for the opportunity to study in manuscript their forthcoming definitive
account of the oestrous cycle of the rat, based on their preliminary studies referred to in the appended bibliography.

119

120 CYCLIC CHANGES IN THE OVAKIES AND UTERUS OF THE SOW, ETC.

with a simple, diffuse placenta. In short, the problem is here reduced to its lowest
terms, and the solution promises, therefore, to be the more useful as a contribution
to the general theory of the reproductive cycle.

EXTERNAL MANIFESTATIONS OF THE REPRODUCTIVE CYCLE IN SWINE.

Investigations in the physiological anatomy of the reproductive system must
depend upon material collected at accurately determined periods of the ovarian-
uterine cycle. For this reason it will be necessary to review briefly the well-known
facts of the sexual manifestations of swine.

Sexual maturity is attained before the age of one year, sometimes as early as 4
months, often before the uterus has attained its full adult dimensions. Maturity
is characterized by the recurrence of periods of sexual activity ("heat," or oestrus)
at intervals of 2 to 4 weeks, the usual interval being 21 days. An exact study of
the duration of the oestrous cycle has been made by Struve (1911), whose results
agree closely with the figures just given; his computations give a mean interval of
20. 66 ±0.205 days, with a standard deviation of 2.36. The curve of frequency
distribution of observed cases shows the shortest interval to be 15 days, the longest
30 days, but 75 per cent of the animals fall within the limits of 18 to 23 days. The
duration of oestrus is commonly 3 days, during which time the rutting sow exhibits
excitement in the presence of the male, with ready acceptance of coitus. If no male
is present the sow will follow the other females about the pen, sniffing at their
genitals and frequently going through an imitation of the sexual act by "riding"
upon the others; or she may at times be the recipient rather than the donor of these
attentions. In a large herd without boars the females in heat will often be found in
a separate group apart from the others, where for hours at a time the exhibition of
the cestral urge continues, interrupted only by siesta or feeding time, until at the end
of 3 days, or occasionally longer, it subsides, to be followed by an interval during
which sexual activities are in abeyance.

At the height of oestrus the vulva is often slightly everted, swollen, and
reddened, and there is sometimes a slight serous or mucous vaginal discharge.
Occasionally the discharged fluid is flecked or stained with blood, but from internal
examination of the reproductive tract it is apparent that in these cases the bleeding
is of external origin, 'caused no doubt by trauma to the vulva; in this detail, as in
all others of the description just given, the external manifestations of the sexual
cycle are in complete contrast to those displayed by the human and other primate
races.

The 21-day cycle appears to continue regularly throughout the year unless
interrupted by pregnancy. In pregnant animals received at the packing-houses,
fetuses are found in all stages of development without much regard to the time of
year. Pregnancy can begin only at an cestral period, since then alone is copulation
permitted. The span of gestation is 16 to 17 weeks, usually 116 to 120 days. As
with other mammals, the cestral periods do not occur during pregnancy, but
according to Struve (1911), oestrus recurs about 4 to 9 days after parturition. -

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 121

The number of young in a litter varies from 1 to 23, the mean being from about
7 to 9 in different American breeds, according to Surface (1909), who has made a
biometric analysis of the data upon this subject.

COLLECTION OF THE MATERIAL.

The first step in correlating the organic changes with the outwardly visible
functional events which have been recounted is, necessarily, to determine at what
point of the cycle ovulation takes place. The results of this part of the investiga-
tion, and of a study of the origin of the corpus luteum and its alterations during
pregnancy, have already been described in previous publications by the writer
(1915, 1917a, 19176, 1919) and need only be summarized here.

It was found, as had been expected by analogy with previously known species,
that the period of oestrus is the time of ovulation. In the stockyards the condition
of heat was observed in about 30 animals, which were suitably marked and then
were followed through the processes of the abattoir until the genitalia were ob-
tained from the butcher's hands. In such of these animals as were killed during
the three days following the onset of cestrus the ovaries contained either mature
or recently collapsed Graafian follicles. This much had in fact already been
proved by Lewis (1911) in a publication at the time unknown to the writer, but
we were able to strengthen the evidence by actual recovery of the ova from the
Fallopian tubes. By the use of simple technical methods, which will be described
later, a series of segmentation stages and also of unfertilized ova was obtained
for study and correlation with the corpora lutea. By good fortune the packing-
house where these first steps were taken was not at the time under great stress of
production; a few of the swine were retained in the corrals as long as 11 days after
the onset of cestrus, and it was therefore possible to follow the development of the
corpus luteum up to about the tenth day. It was not until the adoption of a rapid
method of locating ova and blastodermic vesicles in the large uterine chambers of
the sow that the fate of the ova after their exit from the Fallopian tubes could be
continuously followed; but the writer already had in hand the ovaries of sows in
all stages of pregnancy from the third week until parturition, to the number of
about 140. Upon the sum of this material two previous contributions were based
(1915, 1919), covering, therefore, the corpus luteum of pregnancy from the day of
ovulation until after parturition and the corpus luteum of unfertilized ovulation
until the tenth day of its development.

It was, of course, impossible at the stockyards to follow animals into the latter
half of the cycle. Such material, however, was made available by the generous
interest of Mr. Walter N. Cooper, manager of the American Feeding Company,
of Baltimore, who extended every facility for the undertaking. At the establish-
ment of the company, about 20 miles from Baltimore, large numbers of pigs were
kept from an early age until they attained a profitable weight by the consumption
of table refuse and other edible garbage collected in the city. The writer made a
series of trips to the piggery farm on alternate days throughout a period of 3 weeks,

122 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

and at each visit selected, with the aid of Mr. Cooper and his assistants, 3 sows
which were evidently in active heat. These were marked by ear-tags and isolated
from the herds in a special pen. The date of cessation of cestrus was noted by
report of the piggery laborers and by personal observation on the days of visit.

At the end of 3 weeks 22 animals had been thus isolated, forming a series of all
stages of the cycle. These were then purchased by the University and resold to a
packer near the laboratory, by whom all of them were killed on the same day,
and the internal genitalia were thus recovered for study. The sows were all in
good condition, readily passing government inspection. They were of various
breeds and crosses, but were nearly uniform in age (about 10 to 12 months) averag-
ing over 180 pounds in weight, indicative of fairly mature state.

These 22 animals have served as a basis for all statements in the following pages
as to the time relations of the reproductive cycle, but, owing to the variability of
the intercestral period, it chanced that none of them was killed within the first
3 or 4 days following ovulation. This gap in the series of uteri might have been
filled by specimens from the first collection of animals mentioned above, but un-
fortunately all the uterine preparations of those sows had been put permanently
beyond the reach of study by the departure for the war of a pupil to whom they had
been intrusted. For this reason a further series of 30 specimens was collected from
the slaughter-house. These were not obtained from animals that had been ob-
served during life, but were at least approximately dated by the fact that early
corpora lutea were present in all and that the ova or early embryos were recovered
from each specimen. By the microscopic structure of the corpora, as worked out
from the previous specimens, and by the condition of the ova, it was possible to
rank them in a series and to discover from them the earlier uterine changes fol-
lowing ovulation.

It is obvious that there is a limit to the chronological accuracy of data obtained
in the various ways described above. Students of the human ovary and uterus
may perhaps think enviously of the opportunity afforded in the packing-house
and stockyard to collect fresh organs, datable more or less closely, from young,
healthy, mature animals to the number of twelve score and to see similar but
undated specimens totaling almost 12,000; on the other hand, no such exactness is
possible with the large animals of commerce as with animals like the rat, where the
method of Stockard, as applied by Long and Evans, permits prediction of the time
of ovulation within one hour. In this series of swine it has not been possible, in
most cases, to observe cestrus from start to finish, nor do we yet know more than
approximately the relation of the moment of ovulation to the usual 3 days of cestrus.
Moreover, we are dealing with variable quantities in the duration of the inter-
cestral period and the rate of development of the fertilized ova, both of which fac-
tors have been used in establishing correlations. However, the reader will prob-
ably not be led astray if we estimate that all references to a given day, assuming
a 21-day cycle as a working basis, imply a possible error of =±= 1 day, perhaps even
of ± 2 days as we approach the latter days of the intercestral period.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

123

METHODS OF PREPARATION.

Since recovery of the ova and early embryos by means of serial section of the
tubes and uteri is out of the question in so large an animal as the sow, a procedure
was devised which grew out of a suggestion by Professor Evans based upon the
practice of Martin Barry (1839), one of the earlier workers on the problems of
o vulation, who obtained the tubal ova of rabbits by stroking the tubes with a rod in
order to express their contents into a dish. Our improvement, which we have
since found had already been used by Sobotta (1897) and others, consists simply
in washing out the Fallopian tubes with a stream of isotonic salt solution.

Uterine cornu. 72 cm.

^4

Body ol tube 17cm

Funnel

4cm I

Corpus uferi 5cm. /s

Tube
Liq.Laium

F£G. 1. — Diagram showing form and dimensions of the uterus and Fallopian tubes of the sow, drawn
from an average specimen taken from a young mature animal.

Reference to figure 1 will make the following description clear. The Fallopian
tube is freed from the mesosalpinx by cutting the latter off close to the tubal border
and is detached from the uterus at the junction of the tube and cornu. The now
straightened tube is suspended by one hand over a Syracuse dish and is filled, by
means of a pipette, with sufficient normal saline solution to distend the lumen
moderately. The narrow uterine end of the tube prevents the fluid from draining
into the dish until .the tube is gently "milked" out with the fingers. If ova are
present they usually pass out with the first drops of fluid, but in some cases four or
five washings are necessary to recover all that are present in the tube. The ova
are located in the dish by inspection of the washings under a low-power microscope.

This method is rather awkward when applied to the whole uterus because of the
large quantity of fluid required to distend the canal, amounting often to 300 c.c. For
this reason recourse has been had to a plan by which a small amount of fluid (10 to
30 c.c.) is caused to distend successive short portions of the uterus. The uterine
cornu having been uncoiled by cutting it from the mesometrium, a stout clamp is
applied to it a few centimeters below the ovarian extremity and the distal portion is
dilated with the salt solution; another clamp is now applied a few centimeters
below the first, which is removed, and the fluid is forced by gravity and by digital
pressure into the second section of the uterus, which is thus in turn distended, and so
on. In this manner any ova or unattached embryos which are present are picked up
and carried along until the fluid has passed through the whole length of the uterus,
and are then deposited with the washings in one or two Syracuse dishes.

124 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

The washing-out method is very satisfactory for recovering ova and vesicular
embryos from the uterus, but if it is inadvertently applied to a specimen containing
embryos of the latter part of the second week or of the third week, which have
undergone the extraordinary extension in length of the chorion which is character-
istic of the pig and some other ungulate species, the result will be an almost hopeless
tangle of chorionic membranes. Therefore, it is a wise precaution, when expecting
early embryos of uncertain age, to slit up the cervical end of the uterus for a distance
of several centimeters, watching for the delicate, glutinous, threadlike appearance
which marks the early chorion. If none is found, then any still earlier embryos
which may be present are presumably grouped somewhere in the remaining part of
the uterus and may be recovered by washing.

Embryos of the third week and older have been removed by the usual method
employed by embryologists, namely, by cautious opening of the uterus under salt
solution in a well-illuminated glass dish against a dark background. • The salt
solution used in the work has generally been prepared according to the formula of
Locke, except for the omission of glucose.

The large size of the mature Graafian follicle also presents difficulty when it is
desired to locate the undischarged ovum. The use of serial sections of the whole
follicle is sometimes necessary, in which case much labor can often be saved by
embedding in celloidin and employing an assistant to cut the sections, which the
observer examines in turn as they are cut, until the cumulus oophorus begins to
appear in the sections, after which still greater care is used until the ovum itself is
reached. Another method is to lay the hardened ovary into serial slices of about 1
mm. thickness and then to search the follicle walls in the slices under the binocular
microscope until the cumuli are found. The ovum-bearing portion of the follicle is
cut out of the slice and sectioned in paraffin. As there are usually several follicles
in each ovary, the writer's practice has been to try the latter method, reserving
enough of the follicles to assure success by the former plan if the latter fails.

Histological preparations have been made by the usual methods. Bouin's
picro-acetic-formol has been the chief fixing fluid, supplemented, of course, by a
variety of others for special purposes. The stains used will be specified in con-
nection with the illustrations. When it was necessary to wash out the uterus, a
block for histological preservation was first cut from the middle of one cornu in
order to spare the tissues any harm resulting from the fluid and from the pressure;
the portions remaining were then separately washed out for the ova.

As it is difficult to cut thin sections of the sow's uterus except from very small
blocks, complete transverse sections of each uterus were made by the celloidin
technique to serve as reference for the small blocks used for study of cytological and
microchemical details.

The writer's thanks are due to the American Feeding Company and to Hohman
& Sons of Baltimore, to the Western Meat Company of San Francisco, and to
Joseph Stern & Company of New York City for opportunity to collect this
material; and to the Department of Embryology of the Carnegie Institution for
technical assistance.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 125

GROWTH OF THE GRAAFIAN FOLLICLE AND MATURATION OF THE OVUM.

The ovaries of the large animals are not well suited to the study of the various
problems connected with the growth of the Graafian follicle, on account of the
obvious physical difficulties which prevent systematic examination of large numbers
of follicles and ova. Sufficient information has been gained, however, with respect
to the pig, to supply an outline of follicular growth and to correlate its events roughly
with those of the oestrous cycle. It appears that normal follicles of maximum size
are found only during the period of a few days previous to ovulation. At all other
times there is what might be termed a reserve stock of smaller follicles in the ovary,
forming a series of all sizes from the microscopic primordial stage to a diameter of
about 5 mm. Judging from the evidence of the 22 mature sows mentioned above,
it seems that practically all actively functioning ovaries contain a number of follicles
of 3.5 to 5 mm. diameter, which are in readiness for the final enlargement which
they are destined to undergo just before rupture. Follicles of larger size are usually
found to be atretic, although in a few cases there are normal follicles of 6 and even 7
mm. diameter during the intercestral period.

At a time not as yet accurately determined, but which can not be more than 2
or 3 days before the onset of oestrus, there begins a rapid enlargement of those
follicles which are to discharge their ova, bringing them to a diameter of 7, 8, or
even 10 mm.; and at the same time there is a series of histological changes, which,
in the smaller animals, have long been known to be connected with the maturation
of the ovum. These consist of growth of the theca interna by enlargement of its
cellular elements and of a partial dissolution of the cumulus oophorus by separation
of its cells, so that the ovum is finally almost freed from its originally firm anchorage
to the follicular wall. A full description of these changes will be found in the
author's paper of 1919. The ovum itself goes through the first stages of the process
of maturation, its nucleus moving toward the periphery of the cell, to form first the
typical germinative vesicle, and then to undergo mitosis and extrusion of the first
polar body.

The study of new material has not modified the opinion already expressed in a
brief note (1917) that the process of maturation of the ovum follows in the pig
the same course as in other mammals. At the moment of rupture of the follicle the
first polar body has been discharged and the second polar spindle has been formed,
but completion of the second polar body does not take place unless the ovum is
fertilized.

Rupture of all the follicles seems to take place simultaneously or at least
within a brief space of time, for in a careful examination of perhaps 200 pairs of
ovaries containing recently ruptured follicles there has been only one case in which
there were also normal, mature, uncollapsed f ollicles. A few other specimens which
appeared to be of this sort proved upon section to be atretic. This conclusion is
contrary to that of Ktipfer (1920), who thinks that an appreciable space of time
may be required for the rupture of a group of follicles, but the specimens upon which
he bases this statement have apparently not been subjected to microscopic exami-

126 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

nation. Corner and Amsbaugh (1917) concluded, from a study of 10 animals,
that ovulation probably occurred on the first or second day of osstrus; but the
exact time of onset of heat in these animals could not be stated with great exactness,
since the pens were visited but twice daily, and in some cases less often. A some-
what different conclusion is drawn by Lewis (1911) from the results of his experi-
ments, unfortunately not known to us at the time of our work. His material con-
sisted of 23 sows, in which the onset of heat and the time of copulation were noted
with exactness, the animals being killed at varying times thereafter. In those
killed before 30 hours after onset of cestrus the follicles were not ruptured (with
one exception), but in most of those killed between 30 and 48 hours the follicles
had collapsed. In one case ovulation had not occurred at 70 hours, but as no
microscopic examination is mentioned, we can not exclude the possibility that this
last case is one in which the follicles had undergone atresia instead of rupture.
Lewis's interpretation is that "the ovum (in hogs) is not liberated from the -ovary
until the last part of the period of heat, " but from his table one is forced, rather,
to conclude that ovulation usually occurs during the second day of the period of
restrus.

FATE OF THE OVUM: PASSAGE AND ATTACHMENT OF THE EMBRYOS.

The time necessary for the passage of the ovum through the Fallopian tube
has been set by Assheton (1898) at 3 days, and the present writer's findings are in
accord with this statement. This journey of 18 or 20 cm. is therefore traversed in
the same time as the far shorter distance in a small mammal like the mouse. If
copulation has occurred, the conjugation of ova and spermatozoa takes place in the
Fallopian tube and segmentation reaches the stage of 2, 4, or 6 blastomeres by the
time the uterine end of the tube is reached. About the fourth day the ova enter
the uterus.

Much interest attaches to the fate of the ova when copulation does not occur.
Degenerating ova are but rarely found in the tubes, and one is therefore forced to
suppose that all the ova pass into the uterus, whether or not the sow has been
impregnated. This conclusion is strengthened by the fact that if from the run of
sows at the abattoir one selects those whose ovaries contain fresh-looking solid or
nearly solid corpora lutea (i. e., 5 to 7 or more days after ovulation), a considerable
proportion of these will be found, upon careful search, to contain either normal-
looking or degenerating unfertilized ova in the uterus. Scores of such obser-
vations, confirmed in numerous cases by microscopic estimation of the age of the
corresponding corpora lutea, make it seem clear that the unfertilized ova regularly
pass into the uterus and degenerate there, not disappearing completely, however,
before at least the seventh day after ovulation.

Some information as to the probable time of their final dissolution can be
obtained from the following table of cases in which the ova were sought in the uteri
of animals whose last cestrous period had been observed, and which had not been
impregnated.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

Degeneration of ova in non-preqnant animals.

127

Serial
No.

Calculated
days after onset.

Age of corpus
luteum estimated

State of uterine ova.

of oestrus.

from structure.

1

3 to 5

4 to 6

Present, not degenerated.

2

3 5

6 7

Not found.

3

6 8

7

Present, not degenerated.

4

6 8

7

Present, not degenerated.

5

8 10

7 about.

Not found.

6

8 10

7 about.

Not found.

7

6 8

7

Present, degenerating.

8

6 8

7 or more.

Not found.

9

8 10

7 or more.

Present, degenerating.

No ova were found in animals of later date.

We shall not be far wrong, therefore, in assuming that the unfertilized ovum of
the sow disappears by degeneration in utero at about the seventh or eighth day after
its discharge from the follicle. Degeneration of the unfertilized ovum is charac-
terized by division of the cytoplasm into rounded masses, which are usually not
uniform in size, but sometimes are sufficiently regular to simulate segmentation.
These masses ultimately contract so as to present a shrunken appearance within a
zona pellucida which is often swollen but at the same time less refractive. The
zona pellucida frequently loses its spherical shape and at the last may present only
a vague, flattened halo, containing minute punctate granulations, around a few
shrunken masses of cytoplasm.

Of still more importance, in affording a basis for correlating the anatomical
events of the reproductive cycle, is the question of the rate of growth and the time
of implantation of the fertilized ovum. Fortunately for our purposes, pig embryos
of known ages have now been seen at all stages of early development. The present
writer, with Amsbaugh.(1917), had the opportunity of completing the series by
the observation of ova immediately after fertilization, before conjugation of the
pronuclei. The careful studies of Assheton (1898) on about 100 embryos carry the
history from the stage of two blastomeres until the tenth day, and 30 more speci-
mens of the ninth, tenth, and eleventh days have been described by Weysse (1894).
From the fourteenth to about the twenty-fifth day we have the detailed data of
^KeibePs Normentafeln (1897). The following account of the early embryology
of the pig is based upon the contributions just mentioned and upon personal study
of a series of specimens of all stages.

Segmentation begins in the tube and continues after passage of the ova into
the uterus; no specimen beyond the stage of 6 blastomeres has yet been observed
in the tube. Assfeeton found that various sows killed on the sixth day contained
uterine embryos from 6 blastomeres to fairly well-developed blastodermic vesicles.
By the seventh day the zona pellucida has disappeared and two layers, epiblast and
hypoblast, may be distinguished in the inner cell-mass of the vesicles. By the
eighth or ninth day the vesicles begin to be wrinkled and then to undergo the
remarkable elongation which is so characteristic of early pig embryos. By the

128 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

twelfth day the vesicles are 10 to 12 mm. in length by about 3 mm. in diameter
at their widest part; and by the fourteenth day they are 20 to 30 cm. long, with
embryonic areas showing the neural groove well developed and possessing 1 to 5
somites. At 17 or 18 days the uterine cavity is completely filled by the embryonic
envelopes, which, when their accordion-like folds are eliminated by tension, are
found to have reached their full length of 30 to 40 cm.

It is important to remember that the relations between maternal and fetal
tissues are of the simplest character in this species. No decidua is formed from
the uterine mucosa, which retains during gestation a structure not widely different
from that of the non-pregnant period, and the chorion simply becomes applied,
over practically its whole surface, to the epithelium of the uterus (fig. 33, pi. 4).
Moreover, the chorion itself remains a simple one-layered membrane, against the
inner surface of which, after the third week, the vascular allantois is intimately
applied. Nutritive substances passing from mother to embryo must then traverse
both the uterine epithelium and the columnar chorionic epithelium in order to enter
the fetal (allantoic) vessels. At all periods of pregnancy the chorion is readily
detached by gentle traction and at parturition the fetal membranes separate from
the uterine wall without lesion or hemorrhage, leaving an intact surface — an
arrangement which seems excellently adapted to minimize the strain of giving
birth to litters sometimes as numerous as 20 or more.

In view of the diffuse non-deciduate placentation, it is not possible to define
sharply the time of implantation. We can say only that after about the tenth day
growth of the vesicle is so extensive that it must then be considered dependent upon
maternal nutrition rather than upon its own resources; by the thirteenth day, if
not earlier, its position is fixed ; by the fifteenth the chorion has come into relation
with a large part of the uterine mucosa; while during the third week the allantois
begins to line the inner surface of the chorion and hence brings fetal and maternal
tissues into their definite relationship. It is the period from the tenth to the
fifteenth day that is most nearly comparable to the much more precisely limitable
act of implantation which occurs in animals like the rabbit and guinea-pig, and
presumably in man.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 129

II. OVARY.
ORIGIN AND COMPLETED STRUCTURE OF THE CORPUS LUTEUM.

The author's views as to the origin of the corpus luteum have been fully con-
firmed by the examination of new material obtained in connection with the study
of the uterine cycle, amounting to about 35 animals taken in the first 8 clays after
ovulation. For a detailed description of the changes undergone by the discharged
follicle during its conversion into the corpus luteum, the reader is referred to the
author's article of 1919; at the present time a brief summary will suffice as intro-
duction to an account of the subsequent history and degeneration of the corpus
luteum.

When the Graafian follicles collapse at the time of rupture, the extrusion of the
follicular fluid and the contraction of smooth-muscle cells in their walls reduce their
diameter of 8.5 to 10 mm. to 4 to 6 mm. By the end of one week's growth, however,
the corpora lutea are again 8 to 9 mm. in diameter, and if pregnancy ensues there
is a further growth until, after 2 or 3 weeks, the maximum size of 10 to 11 mm. is
reached. The membrana granulosa is retained intact, except for loss of the cumulus
oophorus, after rupture of the follicle. Its cells increase in size without division,
their cytoplasm becomes laden with lipoid substances, and they become the larger
elements, commonly knowfi as "lutein cells," in the fully formed corpus luteum.
During this process the membrana granulosa is invaded by blood-capillaries from
the theca interna, which ramify to" form an extensive vascular plexus throughout the
new structures. The large lipoid-latlen cells of the theca interna of the Graafian
follicle are increased in number by mitotic divisions and pass into the corpus luteum
to become lodged between the granulosa cells throughout the whole structure.
There is no evidence that the cells of the theca interna are ever converted into
fibroblasts of the usual spindle-cell type or that they lay down the fibrils of the
close-meshed reticulum which is present in the corpus luteum.

The time relations of these changes may be given as/ollows': ovulation probably
occurs on the second day of oestrus ; invasion of the granulosa by blood-vessels and
theca interna cells begins on the third or fourth day after on^et of osstrus and is
completed about the sixth or seventh day; by the seventh day the corpus luteum is
usually solid and its cells have become fully differentiated.

Kiipfer (1920) has given a valuable series of colored plates representing the
gross appearance of the follicles and of the corpora lutea at various stages of develop-
ment and retrogression. Unfortunately, the figures were not made from specimens
of known relation to the cestrous cycle, and for this reason we have provided (plate 2)
drawings of ovaries of 3 sows killed at known periods : during astrus (fig. 9) ; at
8 days after ovulation, with completely solidified corpora lutea (fig. 10); about 17
days after ovulation, with degenerating corpora lutea and a new group of follicles
beginning their precestrous enlargement (fig. 11).

The corpora lutea of a sow at the tenth day after ovulation are very conspic-
uous objects by reason of their size, reaching 8 to 9 mm. in diameter and pro-
jecting nearly all their bulk from the ovary, so that when numerous corpora are

130 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

present the remainder of the ovarian substance is dwarfed in comparison. At this
stage the corpora are nearly always solid, though at times one or two in an ovary
may remain slightly cystic. The cut surface bulges from the capsule and presents
a somewhat velvety texture of pink color without trace of the yellow and orange
pigments which are so characteristic of the corpus luteum in the bovine and human
ovaries. The consistence of the corpus luteum at this time is not unlike that of the
pig's liver.

In microscopic sections (fig. 3, pi. 1; fig. 12, pi. 2) the granulosa lutein cells
form the most conspicuous element of the tissue, since they are distinguished both
by their large size (reaching the diameter of 30 to 40 micra) and by the elaborate
cytoplasmic patterns which they contain after fixation in slow aqueous fixatives
like formol, Bouin's picro-formol-acetic fluid, etc. As previously described (1919),
the granulosa lutein cells of certain species contain a large amount of a lipoid or
mixture of lipoids, probably associated with proteins, which is sufficiently oily to
round up into droplets when the tissue is submitted to the action of water. The
droplets thus produced usually surround the pre-existing globules of neutral fat,
which are also present in considerable numbers in the granulosa lutein cells of swine,
but these latter are readily removed by the alcohol, ether, and xylol of the usual
histological procedures, and hence the final appearance is usually such as shown in
figure 12 (gr. I. c.), in which the lipoid droplets are seen as hollow spheres (ap-
pearing as rings in thin sections) more or less retracted from the surrounding cyto-
plasm. The bodies in question are not seen in fresh tissues, nor in cells fixed very
promptly in rapid coagulants like osmium tetroxide, which presumably precipitate
the proteids before the oil droplets round up. The appearances described, there-
fore, are simply the result of methods of fixation which do not preserve certain
obscure lipoids in their natural diffused state ; but the artifact has proved useful in
several ways, notably in tracing the history of the granulosa derivatives.

Besides the granulosa lutein cells and the blood-vessels, the corpus luteum
contains cells of another type, whose origin has been traced by the author to the
theca interna of the Graafian follicle. Corpora of the eighth to the tenth day
always contain a few distinct clumps of theca interna cells in their original position
about the periphery of the corpus luteum and along the vessel-bearing septa which
pass radially inward where the follicle- wall was infolded by collapse at ovulation.
Many others of the theca cells, however, have wandered among the granulosa
lutein cells, from which they can usually be distinguished by smaller size (diameters
of 10 to 25 micra), by a more deeply staining cytoplasm, which is often densely
packed with minute regular vacuoles, giving a characteristic foamy appearance,
and (in osmium preparations) by the presence of plentiful fat globules which vary
greatly in number and size (fig. 13, pi. 2, th. I. c.). The theca lutein cells often
have squared or irregular outlines, fitting into the interstices between the swollen
rounded surfaces of the larger granulosa lutein cells (fig. 12, pi. 2, th. I. c.).

The tissue is held together by a framework of reticular connective fibrils
which, as illustrated in figure 4, plate 1, from a preparation by Bielschowsky's
method, form a dense network about all the lutein cells. In view of the apparent

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 131

scantiness of fibroblasts of the usual type, the presence of so complete a reticulum
gave cause for surprise until it was found (Corner, 1920) that here, as in a number
of other organs, the capillary endothelial lining is the source of the reticulum. As
the capillary bed is so complex that every cell in the gland is in contact with a
blood-vessel, it is not difficult to understand the great density of the reticulum.

RETROGRESSION OF THE CORPUS LUTEUM.

It is a fact of the greatest importance that the changes, just summarized,
which lead up to the formation of the corpus luteum, take place with equal complete-
ness, whether or not the sow has been impregnated. It is not possible to distinguish
the corpus luteum of pregnancy from that of unfertilized ovulation during the first 2
weeks after discharge of the ova, nor can the observer determine, from the appear-
ance of two specimens of equal age, which of them was destined to further growth
in size and a 4 months' span of activity, and which to immediate retrogression.

The time of degeneration of the corpus luteum of unfertilized ovulation can be
determined with fair accuracy, for the data, though few, are in complete accord
with each other. Seven animals of our series were taken 15 or more days after an
observed oestrus but before another ovulation had occurred, and all of these con-
tained retrogressing corpora lutea, while all those killed earlier showed no sign of
degeneration; 2 of the 7 had begun a second cestral period and thus gave a further
check upon the time relations. It is upon the fourteenth or fifteenth day, then,
that a sudden change overtakes the structure which has been so elaborately erected.
Within 2 or 3 days the diameter of the corpus has decreased from 8.5 or 10 mm.
to 6 mm.; its color has changed from the pink of an active capillary circulation to
the whitish tone of scar-tissue, and its texture has much increased in firmness and
toughness. Microscopic examination (fig. 6, pi. 1; fig. 14, pi. 2) shows the change
to have been so rapid as almost to defy comparison with the previously existent
conditions, and exact analysis of what has happened is thus rendered difficult. An
example of the suddenness of the break-down is given by one of the animals, killed
on the fifteenth day, in which some of the corpora were degenerated while others
were intact, but in one corpus there was a patch of advanced degeneration sur-
rounded by unchanged tissue.

The most striking feature of the change is the disappearance of the granulosa
lutein cells. In the sows whose ovaries show the least degeneration many of these
cells can still be distinguished in various stages, from a fair state of preservation
to almost complete degeneration (fig. 14, pi. 2), but in other animals of this
period nothing remains but vague vacuolar spaces containing many pycnotic
nuclei and nuclear fragments, where but a few days before were closely packed
masses of cells, among the most imposing of the animal body. Since these cells
in number and bulk formed by far the greater part of the corpus luteum, it is obvious
that their breakdown alone will account for the decrease in size and for the relatively
greater density of the retrograding corpus. Another result of the degeneration is
seen at the periphery of the corpus luteum, where the former sharp demarcation

132 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

between lutein tissue and the surrounding capsule of connective tissue is now
obscured (fig. 6, pi. 1). In one or two of the animals there are great numbers of
polymorphonuclear leucocytes in the tissues and also mononuclear wander-cells of
the macrophage type, no doubt to serve the purpose of clearing away cellular
debris.

The fate of the blood-capillaries is a matter of special interest in view of what
has already been said about the capillary endothelium as the source of reticular
connective-tissue fibrils. Deprived of much of their circulatory function by
abolition of the granulosa lutein cells, the capillaries for the most part collapse, with
nuclear degeneration in some cells, but remain in situ with the reticulum. There
is at first no increase in the total amount of connective tissue, though the decrease
in bulk of the corpus luteum crowds the fibrils into smaller space ; but as retrogres-
sion proceeds (fig. 5, pi. 1) the reticulum gradually becomes denser and thicker
until it gains an appearance like that of collaginous tissue, which is so characteristic
of the corpora albicantia (fig. 7, pi. 1). Whether in this change the endothelial cells
are gradually replaced by in wandering fibroblasts, or themselves retain the fibro-
blastic function, can not be said with certitude, but there seems to be no a priori
reason to doubt the latter possibility, in view of what we know of cellular dedif-
ferentiation in general. It may at least be definitely stated that the fibrous tissue
of the retrogressing corpus luteum is laid down in situ, for the Bielschowsky prepa-
rations make it clear that there is no direct growth of fibers from the surrounding
ovarian stroma into the corpus.

It remains to discuss the fate of the cells of the third type, which the writer has
described as theca lutein cells. These seem, surprisingly enough, to survive the
blow which destroys the granulosa lutein cells; for among the debris and the col-
lapsed capillaries, and also in occasional clumps at the periphery, are found
numerous cells which, by reason of their usually angular or elongated shape, foamy
cytoplasm, and wealth of osmium-staining fatty material, can hardly be deemed
other than the theca lutein cells (figs. 15 and 16, pi. 2, th. I. c.). In one or two
animals the osmium preparations show numerous globules of blackened fatty
substance as large as 30 to 40 micra (represented, of course, in ordinary sections by
vacuoles) which seem to he in some of these same cells, suggesting that they, too,
are temporarily affected by the process of degeneration, to the extent of partial
fatty degeneration. When after a few days the nuclear fragments and the vacuolar
spaces left by the degenerated granulosa derivatives have disappeared (fig. 15, pi. 2),
the lipoid-laden cells are more clearly seen, enmeshed in the scar-tissue, where they
persist for weeks, acquiring even denser stores of yellow-pigmented fat (fig. 8, pi. 1) .

By the time of a new ovulation the corpora lutea have diminished to 6 mm.
diameter, at the mid-intercestral period to 4 mm., and by the second ovulation
(i. e., when their own age is about 6 weeks) to 3 or 2 mm. Klipfer (1920) also finds
that they can be traced through a second interoestral period. After this they can
still be recognized in sections for at least another cycle of ovulation, but finally
become so obscure that they can no longer be certainly distinguished from atretic
follicles. It is interesting to note that the two specimens of retrogressing corpora

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 133

lutea of pregnancy (known to be of the seventh and tenth days after parturition)
which were briefly mentioned in the previous paper (Corner, 1919) are quite similar
in microscopic appearance to those of the non-pregnant sow of the same relative
period of retrogression; it is very likely, therefore, that there is no difference in the
mode of degeneration of the two types of corpora lutea.

What is given above is the first attempt, except for that of Leo Loeb (191 la),
dealing with the guinea-pig, to describe the histological details of retrogression of
the corpus luteum from specimens of known age, and in view of the somewhat
unorthodox result one is disinclined to enter upon a discussion of its general bearing,
except in a tentative way. It may fairly be said, however, that should the writer's
views of the structure of the corpus luteum, especially with regard to the fate of the
theca interna, be borne out by other work, they will afford a means of reconciling
one of the old disagreements of ovarian histology. It has often been pointed out
that in cellular arrangement the theca interna of atretic follicles rather closely
resembles the corpus luteum. Both tissues consist of large lipoid-laden cells sup-
ported by a reticular framework, with a rich capillary blood-supply. In the later
stages of retrogression the two tissues are indeed confusingly alike, and, in the
writings of Paladino (1879), Clark (1898), and many others, this fact has been one
of the mainstays of the theca-origin theory of the corpus luteum. This idea (in
connection with the view now held by nearly all investigators, that the so-called
"interstitial cells" found in some mammalian ovaries are derived from the theca
interna of atretic follicles) has been the basis of various attempts to correlate these
three elements of the ovary, such, for instance, as that of Ancel and Bouin (1909).

Sobotta (1896), however, with his clear-cut demonstration, now generally
accepted, that the granulosa layer of the follicle takes a very important part in
corpus-luteum formation, could not agree to any statement of resemblance between
the two type's of follicle-derived tissue, and so expressed himself, when participating
in a general discussion of the fate of the corpus luteum at the 1908 meeting of
the Anatomische Gesellschaft. According to his description of events, in the
corpus luteum the granulosa persists and the theca interna is used up in the pro-
duction of connective tissue, while in atresia, of course, the granulosa breaks down
and the theca proliferates.

Following, on the other hand, the present writer's account of formation of the
corpus luteum, and assuming that atresia folliculi, which has not yet been ade-
quately studied in the pig, is the same here as in other species, it will be seen that
the two processes differ in their course, but not greatly in their end-stages. The
granulosa cells, which degenerate in atresia, degenerate also in retrogression of the
corpus luteum, after their temporary metamorphosis into granulosa lutein cells;
while the theca interna cells, which by our account do not revert to fibroblasts
when they enter the developing corpus luteum, persist alike in the degenerating
corpus and in the atretic follicle. In this way the cells of the membrana granulosa,
derived presumably from the "germinal epithelium," are associated with the
ovum rather than with the soma, and those of each follicle run their course and
disappear when their own particular ovum leaves the body, either by degeneration

134 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

or by parturition; whereas the theca interna cells, presumably derived from the
ovarian stroma, behave as if morphologically part of the somatic ovarian tissues,
only temporarily bound up with the fortunes of the ovum. It is to be hoped that
this conception may be tested by work on other species.

Text-figure 2 gives, in diagrammatic self-explanatory form, a picture of the
ovarian cycle as we have described it in the foregoing pages.

OESTRUS

OESTRUS

10 12 14 16 18 20 22 24 26 28 JO 32 34 36 36 40 42

OESTRUS

II-
10
9-
8-
7-
6-
5-
4-
3-
Z-

0 F

P R E G NAN C V

ALLANTOIS EXPANDED

I i >•("("(• i i — i — i — i — i — i i i i — i — i — j— i—i — I — I — r— i — r— I — I — r— r-i — I — r— i — I i i — i — i i i I — i—i — i— I — i —

DAY5 0 Z 4 6 B 10 12 14 16 18 20 12 24 26 28 30 i2 34 36 16 40 42, ere.

Fio. 2. — Diagram showing in graphic form the relations between oestrus, ovulation, the development

of the. corpus luteum, and the progress of the ova in the aow.
Above: events of the average cycle of 21 days in the non-pregnant sow.
Below: events of the first weeks of pregnancy.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OP THE SOW, ETC. 135

III. UTERUS.
GENERAL DESCRIPTION OF THE SOW'S UTERUS.

During the period of sexual maturity the sow's uterus undergoes regular alter-
nations of structure, correlated with the ovarian cycle, which are continuously in
progress and are varied in their course only by the occurrence of pregnancy. These
alternations involve changes in the degree of surface folding of the mucosa, in the
morphology and dimensions of the surface epithelium, in the dimensions of the cells
of the superficial gland tubules, in the division rate of all the epithelial elements,
in the number and kind of wander-cells in the subepithelial portions of the stroma,
and in the amount of fluids in the interstices of the stroma. In the following pages
these changes will be followed in detail.

The general structure of the cornual portions of the sow's uterus is illustrated
in plates 3 and 4. There is a lining epithelium resting upon a stroma of con-
nective tissue, and this in turn is seated directly upon the muscularis, which, as
usual in tubular forms of the uterus, is disposed in two layers, the inner circular and
the outer longitudinal. Into the stroma pass a large number of glands having
openings in the epithelium.

The epithelium is fundamentally of a simple one-layered type, but (as we shall
see) the columnar arrangement is greatly modified at one period of the cycle, so
that some histologists (for instance Schmaltz, 1911) have even described it as
"many-layered." We may suspect that there is a good deal of incompleteness in
the descriptions of the uterine epithelia of other mammals, due, as in this case, to
lack of an understanding of the cyclic changes.

The glands open into the uterine lumen by means of tubules which run more
or less directly, with but little branching and slight tortuosity, until they are well
into the stroma, when they begin to branch rather freely and to exhibit marked
tortuosity. In this way the mucosa is divided into two fairly well marked layers,
the superficial zone containing but few glands and the basal zone densely packed
with glands so twisted and involved that in sections they are cut across in every
imaginable way. There is a slight difference between the cells lining the glands of
the two zones; those of the superficial tubules are higher, measuring usually 15 to
25 micra in height as against 12 to 18 micra in the glands of the basal zone. There
is thus a greater proportion of cytoplasm to nucleus in the cells of the superficial
tubules, and they look larger and clearer than those of the smaller and deeper-lying
glands. Furthermore, the lumina of the superficial tubules are usually about
twice as wide as those of the basal glands, as might be expected from the probability
that many of the latter are drained by a few of the superficial tubules. There are no
permanent shorter glands or crypts, as in the uterus of the dog and other carnivores.

There are no ciliated cells in the surface epithelium of the sow's uterus, but
cilia begin to appear immediately within the necks of the glands (fig. 17, pi. 2),
and may be found throughout the glands. They are better observed in the larger
superficial tubules, not only because they are more easily seen in the wider lumina,
but also because they are larger, stouter, and more numerous here. One may

136 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

roughly guess that in this region one-fourth of the gland cells are ciliated. In the
basal zone the ciliated cells seem fewer and the cilia are slighter and less obvious
because packed into a narrow lumen (fig. 18, pi. 2, a) ; but when by chance a dilated
gland is found they can be seen to good advantage (fig. 18, pi. 2, 6). If a small
piece of the fresh uterine mucosa is snipped off and flattened under a cover-slip, the
cilia may be seen beating within the glands. It was by this method, and in
fact with tissue from a sow, that the cilia of the uterine glands were first discovered
by Nylander in Leydig's classroom (Leydig, 1852).

There is no evidence to indicate that there is other than a serous secretion
from either glands or surface epithelium. There are no goblet cells in the uterine
epithelia of the sow, and careful microchemical tests for mucin in smaller masses
give negative results. Equally negative were the results of an effort to demonstrate
unsaturated neutral fats by means of osmium tetroxide. The statements of
Wegelin (1911), that there is a cyclic variation in the amount of glycogen demon-
strable in the human uterine mucosa, suggested a similar study of the sow's uterus,
especially since we had already observed glycogen in the fetal membranes in early
pregnancy; but no glycogen could be demonstrated, either by Best's carmine method
or by iodine after alcohol fixation, in any part of the uterine mucosa of the non-
pregnant sow at any stage of the cycle.

The stroma of the uterine mucosa is no more nor less than a rather gelatinous
or fluid-infiltrated areolar connective-tissue, through which course the glands,
blood-vessels, lymph-vessels, and nerves, and which contains in the meshes of its
fibers the various cells of areolar tissue, namely, fibroblasts, macrophages ("clas-
matocytes"), plasma cells (often very numerous, but varying without reference
to the cycle), and the various leucocytes. There is a slight condensation of the
stroma just under the surface epithelium, forming a narrow subepithelial connec-
tive-tissue zone, of which the most superficial fibroblasts are flattened against the
epithelium to form a basement membrane.

The blood-vessels (fig. 19, pi. 3) pass through the muscularis, giving off branches
to plexuses in and between the muscular layers, and then form a network of large
channels near the base of the stroma. From this network long loops ascend toward
the lumen, to end in a delicate capillary plexus in the subepithelial tissue, just
below the epithelial cells.

THE UTERUS DURING OESTRUS.

(Figure 20, plate 3; 25, 26, plate 4.)

During the days of oestrus the uterine epithelium has a total thickness of 25 to

30 micra. As shown in the figures, it is not obviously columnar, but presents an

arrangement which is deceptively suggestive of stratification.

Ova maturing in the The cells are so closely compressed together laterally, and at

Graafitm follicles, or dis- , . , . , , . ,

charged and in passage the same time have attained so low a form, that they are
through the tubes. Ear- rather irregularly packed and the nuclei thus appear to be

liest stages of the corpora 11 /-,

lutea. arranged in several layers. On careful study many of the cells

appear to extend from base to surface of the epithelium, but

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 137

others seem to be impeded by their neighbors from reaching the free surface. The
cells are small, since there is a relatively low proportion of cytoplasm to nucleus;
so that, taking all these criteria together, a histologist coming upon such a tissue
for the first time might perhaps class it among epithelia like those of the urinary
bladder rather than among the secretory types of epithelial tissues.

Three details are especially to be noted at this stage: First, mitotic figures
are very numerous in the epithelium, in some specimens occurring as often as 1 in
every 50 nuclei. The epithelium, therefore, is actively proliferating. Second, a
contrary process is also indicated by the presence here and there of phenomena of
degeneration; at the base of the epithelium there are points where two or three
cells have become vague of outline, with chromatolysis of the nuclei, so that a small
vacuole is formed in which lie a few nondescript nuclear fragments or granules
(fig. 26, pi. 4, v. d.~). This degeneration is, as we shall see, merely the latter stage
of a phenomenon which sets in slightly earlier than the period of oestrus. The
same may be said of the third fact of especial interest, namely, the presence, in
large numbers, of neutrophilic polymorphonuclear leucocytes in the subepithelial
connective-tissue, and even of a few which are embedded in the epithelium, pre-
sumably making their way into the lumen (figs. 25, 26, pi. 4, p. m.).

During oestrus the stroma of the uterus, in mature animals, is very edematous,
so that the interspersed cells are widely spaced (fig. 20, pi. 3).

There are a few mitoses in the cells of the superficial gland tubules, but none
in those of the deep glands. A curious feature is the presence, in some of the
gland cells, of highly chromatic extra-nuclear granules, reaching a diameter of 1 to
2 micra. The nuclei of cells possessing these granules are usually of normal ap-
pearance; but I have tentatively considered the granulation as a degeneration
phenomenon affecting a few cells of the glands. It is seen only during the cestrous

period.

TRANSITIONS DURING FIRST WEEK AFTER OVULATION.

(Figure 27, plate 4.)

During the first week after ovulation we see the onset of changes which finally

effect a striking alteration in the form of the surface epithelium. The individual

cells at first merely grow larger, so that the layer which they

Ova in transit through . , . , , .

the Fallopian tubes; then form is further piled up to a thickness reaching ,35 to 50 micra.
in the uterus, where they -phe yacuolar spaces indicative of degeneration disappear

degenerate about the .

seventh or eighth day. by the end of oestrus, and by the time the eggs have passed
Had they been fertilized • t ^ uterus there are no more polymorphonuclear neutrophil

the ova would now be .

passing through the stages leucocytes to be seen in the epithelium or subepithelial stroma.
"he Mitotic divisions of the surface epithelium continue to be very

corpora lutea are in pro- numerous until the end of this period, when they cease alto-
gether, not to be seen again until just before the next oestrus.
About the time of their cessation there is a further change in the morphology of the
surface epithelial cells, which have been growing in height and which at last are
somewhat suddenly ranged into a simple, high columnar epithelium, which will be
described in the next section.

138 CYCLIC CHANGES IN THE OVARIES AND UTERUS OP THE SOW, ETC.

There is also a wave of mitotic division in the gland cells, which does not begin,
however, until the ova are about to pass into the uterus (3 or 4 days after ovulation) .
In the superficial gland tubules the mitoses cease simultaneously with those of the
surface epithelium, about the sixth or seventh day, but in the basal glands a few
mitotic divisions may be seen a day or two longer, even after degeneration of the
ova and establishment of the high columnar surface epithelium.

Eosinophil poly morphonu clear leucocytes, which are always present in small
numbers in the uterine stroma, especially in the more superficial portions, undergo
a great increase in number during the first week after ovulation. Unlike the neu-
trophils of the cestrous stage, the eosinophils do not invade the epithelium, but
remain in groups about the vessels and glands of the superficial zone.

There is a marked reduction of the edema of the stroma at the conclusion
of cestrus, so that the connective tissue and the structures passing through it are
again condensed into somewhat smaller compass.

STAGE OF EIGHT TO TEN DAYS AFTER OVULATION.

(Figures 21, 24, plate 3; 28, plate 4.)

The surface epithelium now presents a striking contrast to its former appear-
ance, for its cellular elements are tall and narrow and ranged in simple columnar
form. The relative proportion of cytoplasm to nucleus is
The ova have degen- grea% increased and the nuclei generally occupy a central
erated and disappeared, position. The cells vary in height from 20 to as much as

Had they been fertilized , e j j • i j.i_ j_i e

they would now be large 4^ nucra, and are arranged in such a way that the surface
biastodermic vesicles, not formed by their free ends is not smooth but wavy, giving rise to

yet fixed to the uterine ,. , , .„ , , ...

mucosa. Corpora lutea little hillocks, between which are depressions or pseudocrypts
fully organized. (fig- 21, pi. 3). As we shall see, this arrangement aids in

giving the mucosa, in the gross specimen, a soft and velvety
texture. The whole appearance is now that of an actively secreting epithelium.

At this stage one notes particularly well a detail which has often been com-
mented upon by observers of other species, including the human, and which is seen
at all stages of the cycle, namely, the presence between the normal epithelial cells
of others which are compressed laterally, have darker-staining cytoplasm, and
usually compressed pycnotic nuclei (fig. 28, pi. 4, i. c.}. They have been called
"Stiftchenzellen," "cells with pycnotic nuclei," "intercalar cells," etc., and have
been variously interpreted. I can see no reason to doubt that view which holds
them to be degenerated cells in the act of extinction. Here and there one also sees
small cells of uncertain provenance, with dense, round nuclei, which seem to have
intruded themselves into angles at the bases of the tall epithelial cells upon the
basement membrane.

At this time the invasion of the more superficial parts of the stroma by eosinophil
leucocytes (described in last section) is at its height, as shown in figure 24, plate 3.

The cells of the superficial gland tubules now undergo a slight enlargement.
From the seventh to the tenth day of the cycle they measure from 18 to 30 micra, at
all other times from 15 to 20 micra. Likewise, the cells of the basal glands gain
a little in height, exceeding their usual limit of 10 to 15 micra and gaining a

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 139

dimension of 15 to 20 micra. This enlargement of the gland cells is due to an
increase in the relative amount of cytoplasm, so that they also gain an appearance
as of active serious secretion. It should be mentioned here that at no time are
there any marked changes in the morphology of the glands like those described in
the human uterus by Hitschmann and Adler (1908) and others.

STAGE OF THE TENTH TO THE FIFTEENTH DAY.

(Figure 22, plate 3; 29, plate 4.)

The stage of high columnar epithelium is succeeded about the tenth day after

ovulation by another phase, which is characterized by a further modification of the

surface epithelium. This consists of a reduction of the cells

No ova in the uterus; to a ^ow columnar form, measuring from 15 to 20 micra, and

had they been fertilized by the extrusion (from the surface of each cell) of cytoplasmic

gtL^aCrintTthe processes ranging from 3 to 8 micra in height, which will be

uterus. Corpora lutea in more readily comprehended by study of the illustrations than

a fully developed stage ,. i_ i j • j. • T v j • ji

until degeneration sets in from Verbal descriptions. In SOme preparations the appear-
on the fifteenth day. ance js as if the free ends of the cells were frayed and eroded;
in others the processes are rounded or pointed, and sometimes
they might almost be taken for poorly fixed or agglutinated cilia, except for the
absence of the basal granules which mark all true cilia in the uterine glands. (Com-
pare fig. 29, pi. 4, with fig. 17, pi. 2.)

We shall have occasion to return, in a later section, to the consideration of this
very peculiar change of the epithelium. There appear to be no similar observa-
tions on record except those of Hitschmann and Adler (1908), who describe (in
the late-interval and premenstrual stage of the human uterus) epithelial cells
with frayed-out surfaces, and of Geist (1913), who illustrates, also in the human
premenstrual stage, rounded protuberances very much like those of the pig. Geist
considers that the protuberances consist of a secreted substance passing out of the
cells into the lumen; but as far as our specimens are concerned there seems to be
no reason for thinking them other than a modification of the superficial cytoplasm.

Coincident with the lowering of the height of the epithelial cells, there is a
disappearance of the complex hillocky arrangement of the surface. The gland
epithelium now returns to its usual size and the excess of eosinophil leucocytes
disappears from the subepithelial stroma.

STAGE OF THE FIFTEENTH TO THE TWENTIETH DAY.
(Figure 23, plate 3; 30, plate 4.)

During the few days previous to a new ovulation the surface epithelial cells

become so low (15 to 20 micra) that in some preparations they are

During the latter part °f cuboidal form, hardly higher than their own nuclei. The sur-

of this stage a new group face protuberances diminish in height and finally disappear as

of follicles is prepared for ,, ,, , . , j-j. j J.-G J

ovulation. The corpora the cells once more begin to be arranged into a pseudostratmed
lutea are degenerating. epithelium. This stage marks what might be called the comple-
tion of the cycle; in the pig it is of brief duration, but in the sheep,
where there is a long anoestrous period, the uterine mucosa apparently remains in
such a resting stage during the 10 months or more of interval.

140 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

During the last days before the onset of oestrus, however, a few mitotic figures
are found in the surface epithelium and at the same time there are numerous points
of vacuolar degeneration of the epithelial cells. The most remarkable event of this
stage is the assembling of large numbers of neutrophil leucocytes in the subepi-
thelial stroma. In some of the specimens (fig. 23, pi. 3), the leucocytes are seen
swarming out of every arteriole in the superficial stroma and passing toward the
epithelium; a little later they are arranged along the bases of the epithelial cells
and a few are to be found passing through toward the lumen. An cedematous
state of the stroma also sets in toward the end of this period, and thus, by a combi-
nation of all these changes, the uterine mucosa is brought up to the cestrous stage,
at which we began the description.

To conclude this cursory description of the uterine cycle, it may be added that
the histological changes are not without effect on the general appearance and tex-
ture of the uterine mucosa. An opened uterus taken during the cestrous period is
paler than at other times, with a firmer and at times slightly gelatinous inner sur-
face (due to the oedema) ; while during the rest of the cycle, and perhaps especially
from the ninth to the tenth days, the mucosa is pink or red, soft, and velvety.
There is also a periodic change in the external dimension of the uterus, for, by
plotting the circumferences (at the mid-points of the cornua) of the uteri of our
22 mature sows of similar age, in the order of their cyclic stages, a significant curve
is obtained which is directly correlated with the degree of oedema of the stroma.
In other words, the uterus is slightly larger just before and during the period of heat,
because the stroma is then thickened by cedema. Caution must be used, however,
in the interpretation of measurements and gross appearances as seen in specimens
obtained at random at the slaughter-house, for there are numerous possibilities of
uncontrollable variation, especially in the amount of bleeding of the carcases and
in the ages of the animals. Some are taken very young, even before the first
oestrus, and therefore long before the uterus has attained full development. A
marked state of oestrous cedema does not seem to occur at the earliest heat periods.

For the same reason I have refrained from dogmatic statements on a point of
some interest — namely, whether there is a postcestrous hypertrophy of the glands,
such as has been described, for instance, by Keller (1909) in the dog. Study of
our 22 animals of known age and comparable size, though suggesting a positive
answer, does not suffice to settle this question, since there are wide variations in the
number of glands, and one is further confused by the difference in fluid-content of
the stroma. It is difficult to compare specimens in one of which the glands are
widely spaced and the stroma thick, while in the other the glands are densely
packed in a narrow space.

Another uncertain matter concerns the numerical relations of the epithelial
cells. There is a period of very active proliferation of both surface and glandular
epithelium, but no time of widespread destruction. One is forced to assume for the
present that the postcestrous wave of mitosis is compensated for by the sporadic
degeneration of epithelial elements, which must be very frequent if the compressed
cells (page 138) are, as suggested, in a moribund or degenerated state, and by the
loss of cells due to the vacuolar degeneration of the precestrous and cestrous periods.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 141

THE UTERINE MUCOSA DURING THE EARLIEST WEEKS OF PREGNANCY.

(Figures 31, 32, 33, plate 4.)

It now becomes a matter of great interest to compare the histological state of
the uterus during the earliest weeks of pregnancy with the cyclic alterations just
described in the non-pregnant animal. The only available information on this
point is given in brief form by Assheton (1906) in his description of the ungulate
placenta. It has therefore seemed so important to make renewed observations on
this subject that a series of early pregnant uteri has been assembled (by the methods
outlined above, pages 123-124), including ova undergoing segmentation in the tubes
and uterus, morulse and early blastocysts, and implanting embryos in the shield and
earliest somite stages. These have been obtained from the butcher and are there-
fore without exact data as to age; they have been arranged into a series by com-
parison of the corpora lutea and the stage of the embryos. The probable ages in
days have been estimated from the data of Assheton (1898) on the rate of develop-
ment of the pig embryo and from our own studies of the development of the corpus
luteum (Corner, 1919).

Examination of these very early pregnant uteri gives a result which is as im-
portant as it is simple. The same histological changes are found during the first
15 days of pregnancy as during the 15 days following an oestrus without copulation.
By the seventh day the cestrous epithelium has passed into the high columnar stage,
with hillocky arrangement of the surface; mitoses have ceased in the epithelium
but are numerous in the glands; eosiniphil leucocytes are present in large numbers
in the subepithelial stroma. Thus by the time the ova have developed into large
spherical blastocysts (eighth and ninth days) the uterus can in no wise be dis-
tinguished from a non-pregnant uterus 8 to 10 days after ovulation. (Compare
fig. 28 with fig. 31, pi. 4.) Two uteri containing wrinkled vesicles of age estimated
at 11 days show low columnar epithelium with smooth surface, diminution of the
hillocky arrangement of the epithelium, and few eosinophils. By the fourteenth or
fifteenth day, when the chorion is in contact with a large part of the mucosa, the
low columnar or cuboidal cells are covered with the curious frayed or rounded
protuberances which we have seen to be characteristic of the non-pregnant uterus
at the same time after ovulation (fig. 32, pi. 4). These changes of course affect
the more distant crevices and angles of the mucosa as well as those areas where the
chorion is applied to the uterine surface. Where the chorion has been separated
from its attachment to the epithelium, as inevitably occurs over large areas during
fixation, it may be seen that the trophoblast is pitted and roughened by contact
with the irregular surface of the epithelium (fig. 33, pi. 4).

After this stage the identity of the histological processes in pregnant and non-
pregnant uteri is at an end. At a time when the non-pregnant uterus is again
undergoing precestrous changes (18 to 20 days) the pregnant uterus presents even
lower epithelial cells and greater roughening of the cellular surface, so that Assheton
speaks even of a degeneration of the epithelium at this time (fig. 33, pi. 4) . The
cells do not die off, however, but on the contrary again become of medium or high
columnar type, and so persist, as is well known, throughout pregnancy. The stage

142 CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

of low epithelium with surface protuberances or roughening may quite as plausibly
be considered not a degenerative phase, but one of physiological significance, in
some way. (See page 143.)

PREVIOUS ACCOUNTS OF THE UTERINE MUCOSA OF THE SOW.

Certain previous contributions to the cyclic anatomy of the pig's uterus de-
serve consideration at this point. Givkovitch and Ferry (1912) attempt a direct
comparison between the changes of the human menstrual cycle and those of the pig's
uterus. They describe, without giving the time relations, four stages of the corpus
luteum: formation, full development, early and advanced involution; and they
correlate with these stages four steps in the condition of the uterine mucosa, which
they call prehyperamic, hypersemic, posthypersemic, and interval. Such a division
does not disagree with our present description, although we have not seen all the
histological changes mentioned by Givkovitch and Ferry. It is to be regretted
that their preliminary note has not been followed by a definitive account.

Stegu (1912) interested himself chiefly in the question of distribution of the
uterine cilia. He had 60 animals at various stages of the cycle and made studies
of the fresh mucosa in all of these with fixed and sectioned preparations of 15 of
them. As a result of this work he has given the best account of the external fea-
tures of cestrus that the present writer has seen; he speaks of ovulation during
heat ; hints at the oestrous oedema of the uterine mucosa, and mentions the degenera-
tion of certain epithelial cells at this time. Being anxious to contrast oestrous with
definitely non-cestrous uteri, he had but one animal killed between the end of
cestrus and the tenth day thereafter; in this specimen he observed the hillocky
arrangement of the epithelium, with crypt-like depressions intervening, which we
have found to be characteristic of this stage.

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC. 143

IV. SUMMARY AND DISCUSSION.

It remains to consider how far our results have justified the hope, with which we
began, that the reproductive cycle of the sow might prove illuminating in propor-
tion to its simplicity. We have found, first, that there is a regular periodic chain
of events in the ovary; beginning with rupture of the follicle during oestrus, the
corpora lutea attain complete organization about the seventh day and hold their
full development until the fourteenth or fifteenth day. We have seen that this
interval is just long enough to cover the period during which the embryos are
becoming attached, should pregnancy result from the ovulation. If no pregnancy
occurs, atrophy of the corpora lutea begins about the fifteenth day.

In the uterus we have seen a correlated cyclic alternation. During oestrus
there is a characteristic state of the uterine mucosa very similar to that which has
been described by Loeb (1914) and by Stockard and Papanicolaou (1917) in the
guinea-pig, and very fully and accurately by Long and Evans in their forthcoming
monograph on the rat. During the time of formation of the corpora lutea the
epithelium, glands, and stroma undergo a series of changes which attain their
height about the same time that the corpora become solid. From the eighth to
the tenth day the uterine epithelium presents the appearance of active serous secre-
tion, and then its cells lose height, subside to a low columnar or cuboidal form, and
become marked by cytoplasmic protrusions or roughenings on their free ends. After
the fifteenth day, when the corpora lutea begin retrogression, there is a slow rever-
sion to the cestrous type of structure. A key to the understanding of these changes
is found in the fact that an exactly similar histological progression occurs during
the first two weeks of pregnancy, no doubt serving as a mechanism to permit
attachment of the embryos. Recalling the elementary type of implantation in this
species, one is tempted to form the simple, perhaps crude hypothesis that the
uterine epithelium by its serous-secretory stage provides for flotation of the delicate
embryonic vesicles and thus facilitates their migration and spacing in the relatively
extensive uterine cavities (see Corner, 1921); but during the period of attachment
the mucosa is, on the other hand, rendered glutinous by the peculiar surface rough-
ening of the epithelial cells, in order to assist in attachment of the chorions.

Whatever the functional value of these histological details, the reader will
agree that the processes described in this paper strongly suggest that the underlying
fact of the uterine cycle of the sow is an upbuilding of the mucosa, presumably
under control of the corpora lutea, for the purpose of successful implantation.
Each act of ovulation is thus accompanied and followed by uterine changes, which
either go on to placenta formation or (in the absence of embryos) subside once more,
as do the corpora lutea, in preparation for a new ovulation.

Such a general conception has long since been suggested by the work of a num-
ber of investigators, including that of Hitschmann and Adler (1908) on the human
uterus, L. Loeb (19116, 1914) on the guinea-pig, Keller (1909) on the dog, Ancel
and Bouin (1910) on the rabbit, and Hill and O'Donoghue (1914) on Dasyurus
viverrinus. These writers did not see the changes exactly as we have seen them in

144

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

the pig, nor do they agree in all particulars among themselves, though most of them
describe a wave of cell-division in the epithelium and considerable growth of the
glands and stroma in the days following ovulation. The discrepancies of detail
are no doubt partly due to the fact that each of the animals used for these studies
is a law unto itself, both as to anatomy of the reproductive organs and as to outward
expression of the cycle. Preparation of the pig's uterus for the diffuse non-deciduate
placentation of this species may well involve changes somewhat different from
preparation for the elaborate implantation of the guinea-pig and man. Clearly,
much remains to be done before satisfactory generalizations can be established.

Tabular synopsis of the ovarian and uterine cycle.

Ovary, non-pregnant
animal.

Ovary, pregnant
animal.

Unfertilized
ova.

Fertilized
ova.

Non-pregnant
uterus.

Pregnant
uterus.

CEstrus

Follicles enlarging,

Maturing,

Maturing in

CEstrous type of epithelium; epithelial

then ovulation,

in follicles

follicles

vacuolar degeneration (later stage);

then earliest stages of the corpus

then in

then in tubes.

mitoses; migration of leucocytes;

luteum,

tubes.

Conjugation.

oedema of stroma.

(For illustrations, see Corner, 1919.)

Early segmen-

(PI. 3, fig. 20; pi. 4, figs. 25 and 20.)

tation.

Fourth to

Developing corpora lutea.

In

Pass into

Transitions to high columnar epitheli-

seventh

(For illustrations see Corner, 1919.)

uterus ;

uterus;

um; continued mitosis in epithelium;

day.

degenerate

segmentation

invasion of stroma by eosinophil

about

and early

leucocytes.

7th to 8th

blastocyst

(PI. 4, fig. 27.)

day.

stages.

Eighth to

Full development of corpora lutea.

Unattached

High columnar epithelium giving im-

tenth

(PI. 1, figs. Sand 4; pi. 2, figs. 10, 12,

blastocysts

pression of active serous secretion

day.

and 13.)

on surface and in glands; no mitosis

in epithelium; numerous eosinophils.

(PI. 3, figs. 21 and 24; pi. 4, figs. 28 and

31.)

Tenth to

Full development of corpora utea.

Gaining

Low columnar epithelium with altered

fifteenth

attachment

surface of the cells.

day.

to uterus

(PI. 3, fig. 22; pi. 4, figs. 29 and 32.)

Fifteenth

Retrogression of

Continued

Implanted;

Very low epithelium,

Columnar epi-

to

corpora lutea.

growth and

expansion of

becoming "pseudo-

thelium, to

twentieth

Enlargement of new

persistence

allantois

stratified" with re-

which chorion

day.

follicles.

of corpora lutea.

and

newed degeneration

is applied.

vascularization

tion, mitosis, and

(PI. I, figs. 5 and 6;

(For illustra-

of

migration of leuco-

(PI. 4, fig. 33.)

pi. 2, figs. 11, 14, IS

tions see

membranes.

cytes.

and!6;laterstages

Corner, 1915.)

(PI. 3, fig. 23; pi. 4,

of retrogression, pi.

fig. 30.)

1, figs. 7 and 8)

CYCLIC CHANGES IN THE OVARIES AND UTERUS OF THE SOW, ETC.

145

BIBLIOGRAPHY.

ANCEL, P., and P. BOUIN, 1909. Sur les homologies et la
signification des glandes a secretion interne cle
1'ovaire. Compt. Rend. Soc. Biol. Paris, vol. 67,
pp. 464-466, 497-498.

1910. Recherches sur lea fonctions du corps jaune

gestatif. I. Sur le d6terminisme de la prdparation
de 1'uterus a la fixation de 1'ceuf. Jour, de Physiol.
et de Path, sen., vol. 12, pp. 1-16.

ASSHETON, R., 1898-9. The development of the pig during
the first ten days. Quart. Jour. Micros. Sci., n. 3.
vol. 41, pp. 329-359.

- 1906. The morphology of the ungulate placenta.
Phil. Trans. Roy. Soc. London, Series B, vol. 198, pp.
143-220.

BARRY, M., 1839. Researches in embryology. Phil. Trans.
Roy. Soc. London, vol. 129, part 2, pp 307-380.

CLAKK, J. G., 1898. Ursprung, Wachstum, und Ende des
Corpus luteum, nach Bemerkungen am Ovarien
des Schweines und des Menschen. Arch f. Anat. u.
Physiol., Anat. Abt., pp. 95-134.

CORNER, G. W., 1915. The corpus luteum of pregnancy as
it is in swine. Carnegie Inst. Wash. Pub. No. 222
f Contributions to Embryology), pp. 69-94.

— 1917. Maturation of the ovum in swine. Anat.
Rec., vol. 13, pp. 109-112.

- 1919. On the origin of the corpus luteum of the sow
from both granulosa and theca interna. Amer.
Jour. Anat., vol. 26, pp. 117-183.

1920. On the widespread occurrence of reticular

fibrils produced by capillary endot helium. Carnegie
Inst. Wash. Pub. No. 272 (Contributions to Em-
bryology), pp. 85-93.

1921. Internal migration of the ovum. Johns

Hopkins Hospital Bulletin, vol. 32, pp. 78-83.

— and A. E. AMSBAUGH, 1917. CEstrus and ovulation
in swine. Anat. Rec., vol. 12, pp. 287-291.

GEIST, S. H., 1913. Untersuchungen liber die Histologie der
Uterusschleimhaut. Arch. f. mikr. Anat., vol. 81,
I Abt., pp. 196-219.

GIVKOVITCH, J., and G. FEHRY, 1912. Sur les rapports de
1'ovulation et de la menstruation (note preliminaire).
Compt. Rend. Soc. Biol. Paris, vol. 72, pp. 624-626.

HILL, J. P., and C. H. O'DONOGHUE, 1914. The repro-
ductive cycle in the marsupial Dasyurus vivemnus.
Quart. Jour. Micros. Sci., u. a., vol. 59, pp. 133-174.

HITSCHMANN, F., and L. ABLER, 1908. Der Bau der
Uterusschleimhaut des geschlechtsreifen Weibes mit
besonderer Berucksichtigung der Menstruation.
Monatschr. f . Geburtsh. u. Gynakol., vol. 27, pp. 1-82.

ISHII, O., 1920. Observations on the sexual cycle of the
guinea pig. Biol. Bull., Woods Hole, vol. 38, pp.
237-250.

KEIBEL, F., 1897. Normentafeln zur Entwickelungsge-
schichte des Schweines. Jena.

KELLER, K., 1909. Ueber den Bau des Endometriums beim
Hunde, mit besonderer Berucksichtigung der
cyclischen Veranderungen an den Uterindrusen.
Anat. Hefte, vol. 39, pp. 307-391.

KtJPFER, M., 1920. Beitrage zur Morphologic der weib-
lichen Geschlechtsorgane bei den Saugetieren. Ueber
das Auftreten gelber Korper am Ovarium des
domestizierten Rindes und Schweines. Vierteljahrs-
schrift d. naturf. Gesellsch. in Zurich, vol. 65, pp.
377 433.

LEWIS, L. L., 1911. The vitality of reproductive cells. Bull.
No. 96, Agric. Exp. Station of Oklahoma.

LEYDIG, F., 1852. Ueber Flimmerbewegung in den Uterin-
drusen des Schweines. Arch. f. Anat. u. Physiol.,
pp. 375-378.

LOEB, L., 191 la. The cyclic changes in the ovary of the

guinea-pig. Jour. Morph., vol. 22, pp. 37-70.
- 191 Ib. Ueber die Bedeutung des Corpus luteum filr
die Periodizitat des sexuellen Zyklus beim weiblicheu
Saugetierorganismus. Deutsch. med. Wochnschr.,
37Jahrg.,pp. 17-21.

1914. The correlation between the cyclic changes

in the uterus and the ovaries in the guinea-pig.
Biol. Bull., Woods Hole, vol. 27, pp. 1-44.

LONG, J. A., and H. M. EVANS, 1920. The cestrus cycle of
the rat, etc. Proc. Am. Assoc. Anat., 36th session,
Anat. Rec., vol. 18, pp. 241-249.

• 1921. Further studies on the physiology of repro-
duction. Proc. Am. Assn. Auat., 37th session, Anat.
Rec., vol. 21, pp. 56-63.

MARSHALL, F. H. A., and W. A. JOLLY, 1906. Contri-
butions to the physiology of mammalian reproduc-
tion. Phil. Trans. Roy. Soc. London, series B,
vol. 198, pp. 99-141.

MARSHALL, F. H. A., and E. T. HALNAN, 1917. On the post-
cestrous changes occurring in the generative organs
and mammary glands of the nonpregnant dog. Proc.
Roy. Soc. London, vol. S9, series B, pp. 546-659.

NISKOUBINA, N., 1909. Recherches sur la fonction du corps
jauue de la grossesse. Thesis, Nancy.

PALADINO, G., 1879. Studio sulla fisiologia del 1'ovaja.
Struttura, genesi, e significazione del corpo luteo.
Napoli.

SCHMALTZ, R., 1911. Die Struktur der Geschlechtsorgane
der Haussaugetiere mit anatomische Bemerkungen.
Berlin.

SOBOTTA, J., 1896. Ueber die Bildung des Corpus luteum bei
der Maus. Arch. f. mikr. Anat., vol. 47, pp. 261-308.

1897. Ueber die Bildung des Corpus luteum beim

Kaninchen. Anat. Hefte, vol. 8, pp. 469-524.

STEGU, J., 1912. Untersuchungen am Endometrium des
Schweines, mit besonderer Berucksichtigung des
Flimmerepithels und der Brunsterscheinungen.
Inaug. Diss., Wien; also in Osterr. Wchnschr. f.
Tierheilk., pp. 399-443.

STOCKARD, C. R., and G. N. PAPANICOLAOU, 1917. The
existence of a typical cestrous cycle in the guinea-pig,
with a study of its histological and physiological
changes. Am. Jour. Anat., vol. 22, pp. 225-265.

STRUVE, J., 1911. Die Perioden der Brunst bei Rindern,
Schweinen, und Pferden. Fuhling's Landwirt-
schaft. Ztg., Jahrgang 60, pp. 832-838.

SURFACE, F. M., 1908-09. Fecundity of swine. Biome-
trika, vol. 6, pp. 433-436.

WEGELIN, C., 1911. Der Glykogengehalt der menschlichen
Uterusschleimhaut. Ceutralbl. f. allegem. Path. u.
path. Anat. vol. 22, pp. 1-8.

WEYSSE, A. W., 1894. The blastodermic vesicle of Sus scrofa
domesticus. Proc. Am. Acad. Arta and Sci., vol. 30,
283-321.

146 CYCLIC CHANGES IN THE OVARIES AND UTERUS OP THE SOW, ETC.

DESCRIPTIONS OF PLATES.

Comparison of the plates will be facilitated by use of the tabular synopsis, page 144.

PLATE 1.

FIG. 3. Section of the corpus luteum about 10 days after ovulation, showing fully developed granulosa lutein cells. Bouin's
fluid, Mallory's connective-tissue stain. Photograph. X210. (Figure 12, plate ?, shows higher mag-
nification of the same preparation.)

Fio. 4. Corpus luteum about 13 days after ovulation, showing reticulum stained by Bielschowsky's silver-impregnation
method. Photograph. X210.

Fio. 5. Corpus luteum (degenerating), about 17 days after ovulation, showing increase in reticulum. Bielsehowsky.
Photograph. X210. (Same animal as fig. 15, plate 2.)

Fio. 6. Corpus luteum, degenerating, about 20 days after ovulation. Bouin, iron hematoxylin. Photograph. X25.
a, blurring of delimitation between lutein tissue and the capsule, as described in text.

Flo. 7. Corpus luteum in well-advanced degeneration, about 4 weeks after ovulation, illustrating complete disappearance
of the granulosa lutein cells. Bouin, iron hematoxylin. Drawing. X80.

Fio. 8. Corpus luteum (degenerate) about 7 weeks after ovulation, showing numerous fat-bearing cells presumably derived
from the theca interna. Osmium tetroxide fixation without other stain. Photograph. X 150.

PLATE 2.

Fio. 9. Ovary of a sow in cestrua. Natural size, m.f., mature follicles.
FIG. 10. Ovary of a sow about 8 days after ovulatiob. Natural size.

c. 1., corpora lutea of the latest ovulation (8 days old). (See fig. 3, plate 1, and figs. 12, 13, plate 2.) •
r. c. 1., retrogressing corpora lutea of the next previous ovulation (8+21 days), i. /., immature follicles.
FIG. 11. Ovary of a sow about 17 days after ovulation. Natural size.

c. I., corpora lutea of the latest ovulation (17 days old), now retrogressing. (See fig. 14.)
/., Graafian follicles enlarging preparatory to the next ovulation, due in about 4 days.
Fio. 12. Corpus luteum about 10 days after ovulation. Bouin, Mallory's connective- tissue stain. X700.
gr. c. I., granulosa lutein cell.
th. I. c., theca lutein cell.
Fio. 13. Corpus luteum about 10 days after ovulation. Osmium tetroxide fixation without additional staining, to show fat.

X700. gr. 1. c., granulosa-lutein cell. th. I. c., theca lutein cell.
Fio. 14. Corpus luteum in early degeneration about 17 days after ovulation, showing collapsed capillaries and degenerating

lutein cells. Bouin, Mallory's stain. X700.
cap., collapsed capillary blood-vessels.
Fio. 15. Corpus luteum in slightly later degeneration about 17 days after ovulation (not the same animal as fig. 14). Most

of the cellular debris has disappeared. Bouin, Mallory's stain. X700.
th. I. c., persistent cell, presumably a theca lutein cell.
Fio. 16. Corpus luteum in early degeneration about 17 days after ovulation (same animal as fig. 14). Osmium tetroxide for

fats. X700.

th. I. c., persistent fat-bearing cells, presumably theca lutein cells.

Fio. 17. Uterus, showing mouth of gland, to demonstrate that cilia are present in the glands but not on the surface epithe-
lium. Bouin, iron hematoxylin. X330.

Fio. 18. Uterus, showing ciliated epithelium in glands lying at the base of the mucosa, near the muscularis. Bouin, iron
hematoxylin. X330.

PLATE 3.

Flo. 19. Section of uterus showing blood-vessels. Ink injection, carmine stain. Photograph. X10.

Fio. 20. Uterus during o?strus. Bouin, iron hematoxylin. Photograph. X10.

FIG. 21. Uterus about 8 days after ovulation. Bouin, iron hematoxylin. Photograph. X10.

FIG. 22. Uterus about 15 days after ovulation. Bouin, iron hematoxylin. Photograph. X10.

FIG. 23. Portion of uterine epithelium with subjacent arteriole about 17 days after ovulation, showing migration of poly-

morphonuclear neutrophil leucocytes from the blood-vessel. Bouin, carmine. Photograph. X500.
pm., polymorphonuclear leucocytes.
FIG. 24. Portion of uterus about 7 days after ovulation, showing accumulation of eosinophtil leucocytes in the stroma.

Photograph. XI 20.
eos., eosinophil leucocytes.

PLATE 4.

Sections of sow's uterus at various periods of the cycle. Fixed in Bouin's fluid, cut at 5 micra,

and stained with iron hematoxylin. Drawn at uniform magnification of GOO diameters.
FIG. 25. Uterus during oestrus.

p. m., polymorphonuclear leucocyte en route through the epithelium.

mil., mitotic figure in dividing cell.
FIG. 26. Uterus during a-strus.

v. d.t vacuolar degeneration.

p. m., polymorphonuclear leucocytes at base of epithelium.

Fio. 27. Uterus (non-pregnant) about 6 days after ovulation. mil., mitotic figure in dividing cell.
FIG. 28. Uterus (non-pregnant) about 8 days after ovulation. p. «., pseudocrypt. i. c., intercalar cell.
FIG. 29. Uterus (non-pregnant) about 11 days after ovulation.
FIG. 30. Uterus (non-pregnant) about 16 days after ovulation.

mil., mitotic figure in dividing cell. v. d., vacuolar degeneration.
FIG. 31. Uterus (pregnant) about 8 days after ovulation.
Fio. 32. Uterus (pregnant) 14 to 15 days after ovulation.
Fio. 33. Uterus (pregnant) 18 to 20 days after ovulation. cho., chorion of embryo, partly detached.

G. W. CORNER

PLATE 1

.-•.--;.- --.

:{:.:;

-t , ., •'. - , ; - .

>•

.- . ;.': :'•' ' '';. -'•'.- ' '••' . :'•'.'•';

'-i'v ':.''.>. V.f-V". !-.''./'•;'; . >^SP^S^

*. * ,

PLATE 2

;

•>xVi- • . ' -•-•'.'• W T

..''•• <:•• -,'".-" '

! | ;

®^wr< ; , •

-»:-'<' fe> -

•mMf '-^

/V) . ,''-^^r, v • • •

14

:i &&&m

• :--t T&-.S.: L

r-^^T- v ; :
.s '• 't?-: •'-.'-.>

'•^^Xt-1^:^^

|OBE>i

^r^ » v v^ % «s m - «

,--^^~i^^^ MB- ftv'

^ }%;,

^^

J. F. Didusch lith.

A. Hoen & Co. imp.

G. W. CORNER

PLATE 3

,j '.$ -. *J,iS>) I . y <<^>x: ~ - - \ .-'

t • V" ^^-i&^/L y&*& ~*

V , /'^A^^-v > SCT Xir- - kT

?"««»:

••--

^M?;f5ftf*f

it-

J

•^s»2jtT" ^-t-'Jm -t .

' • • •-/•$&•*#'
• v I'sUB'? ^& «^»**

.^>v— • — -^_ _-*-,,.Jt" •. -xKl--.i-i

£os.
« '/•-*

Eos.

tta

,

.

-• ." • :•

frs

.

fe! •:

24

G. W. CORNER
p. m.

m
i

PLATE 4

f^*iK m iS'l^VM— '"• :K^rf'< -&t; 'A r^ 'Pi P '"

*e^^i»^^.:M IlilfiiM

s SS'ffiis ; - ;

mil. i

- 1 '•

«^ F\ r~i

* .'sS7®*?

B1BS1S'

29

<J5->-

31

J. F. Didusch Nth.

\W$ -W. &

^%^S> y ^^n^^^:

%#1--r?r^- -s^- - ^

•T " ?- * 'K' -i^7\ ^

32

~ ' ' cho

'rf ?

Dt-$

• f* 7'

"•

GSS ,r

<c,_ . _. ^^

VKO ~^k. (i*X

.<-, X> :. ^O

<-^^'

«

33

«>

A. Hoen & Co. imp

CONTRIBUTIONS TO EMBRYOLOGY

VOLUME XIII, Nos. 57-64

PUBLISHED BY THE CARNEGIK INSTITUTION OF WASHINGTON
WASHINGTON, 1921